mm m ! I' THE WINGS OF INSECTS X o H o fc The Wings of Insects An Exposition of the Uniform Terminology of the Wing- Veins OF Insects and a Discussion of the More General Characteristics of the Wings of THE Several Orders of Insects By JOHN HENRY COMSTOCK Professor of Entomology and General Invertebrate Zoology, Emeritus, in Cornell University ITHACA, NEW YORK THE COMSTOCK PUBLISHING COMPANY 1918 COPYRIGHT I918 BY THE COMSTOCK PUBLISHING COMPANY PRESS OF \V. F. Hl'MPHREY GENEVA, N. Y. 5". 7 TO ANDREW DICKSON WHITE In grateful recognition of the encoui'agement extended to me in the beginning of my scientific work and continued to the present time, this volume is affectionately dedicated. PREFACE DURING the last forty years, much attention has been given to the development of a terminology of the wing-veins of insects that should be available in the descriptions of insects of all orders, and that should replace the many systems that have been devised by specialists, who have restricted their studies, in each case, to a single order. There has resulted from these efforts the perfection of a uniform terminology, which is being very generally adopted, and which will doubtless supersede the many systems still in use. Many investigators have contributed to the attainment of this result. A brief review of their more important publications on this subject is given in Chapter I of this work. As these publications are widely scattered through various journals that are to be found only in the larger libraries, most of them are inaccessible to many who wish to obtain an understanding of the uniform terminology^ and of the reasons that have led to its adoption. There is, therefore, a demand for a comprehensive exposition of this terminolog}' , and for a statement of the facts upon which it is based. It is to meet this demand that this book is offered to the public. I undertook the preparation of this treatise upon the earnest solicitation of many co-workers in entomology, and with the feeling that the great amount of work, extending over a long period, that I have devoted to this subject made it appropriate for me to do so. It is now more than thirty years since I began a special study of the homologies of the wing-veins of insects, a subject in which my interest had been awakened a decade before by my teacher. Dr. Hagen. My first effort to solve the problem was tmdertaken in the course of the preparation of an article on the H>Tnenoptera published in "The Standard Natural History" in 1884. Much time was devoted to the subject, but without the attainment of any results that seemed worthy of incorporation in that article. With our present knowledge of the subject, it is easy to see how hopeless was the attempt to solve the problem by beginning with a study of the highly specialized wings of the HvTtienoptera. A few years later I attacked the problem again ; this time by a study of the wings of the Lepidoptera. This was a more fortunate starting point ; and considerable progress was made. The results of these studies were published in 1892 and 1893. The results obtained by further studies of the wings of the Lepidoptera and by a study of the wings of the Diptera and Hymenoptera were included in a text book published in 1895. Soon after the publication of this text book, I\Ir. J. G. Needham and I began the investigation of the development of the wings of representatives (vii) viii Preface of all of the orders of insects of which we could obtain living nymphs or pupae. The results of this investigation were published in a series of articles, entitled "The Wings of Insects," which appeared in the "American Naturalist" during the years 1898 and 1899. Since the appearance of this series of articles, I have watched the prepar- ation of many theses on the wings of insects, by students in the entomologi- cal laboratory of Cornell University. The subject, therefore, has been almost constantly in my mind throughout the greater part of my scientific life. The series of articles published jointly by Dr. Needham and myself is the most comprehensive discussion of the wings of insects in which I have taken part ; and, consequently, in the writing of the following pages I have copied freely from them. I am also under obligation to Dr. Needham for continued assistance throughout the preparation of this treatise. Acknowledgments of assistance from other colleagues and students and from published papers are given in the following chapters, where use has been made of it. In this place I will refer only to some of the more impor- tant of these contributions. The papers that had a bearing on the development of the uniform terminology of the wing-veins that were published before the appearance of the series of articles by Comstock and Needham in 1898 and 1899 are enumerated in Chapter I, to which the reader is referred. The most extended use that has been made of the uniform tenninology of the wing-veins is that of Handlirsch in his great work on fossil insects ('o6-'o8). This work was the chief source from which I drew my con- clusions regarding the paleontological evidence concerning the primitive form of insect wings and the methods of their specialization that are given in Chapter IV. Among the more important of the more special contributions to the application of the uniform terminology of the wing-veins of insects are those of Professor J. G. Needham on the Odonata, Miss Anna H. Morgan on the Ephemerida, Professor W. D. Funl<:houser on the Membracidas, Professor Z. P. Metcalf on the Jassidae, Fulgoridas, and Cercopidae, Miss Edith Patch on the Psyllida;, Aphididae, Aleurodidas, and Coccidas, Pro- fessor A. D. MacGillivray on the suborder Chalastogastra of the Hymenop- tera, and Professor J. Chester Bradley on the Evaniidae. In addition to this published paper by Professor Bradley, I have been permitted to study a very extended manuscript on the venation of the wings of the H^-menop- tera written by him. I have also made use of an unpublished thesis by Mrs. Ruby G. Smith on the "Evolution of the Venation in the Anal Area of the Wings of Insects." This thesis embodies the results of an investigation made under my direc- tion, to test the correctness of the conclusions reached by Comstock and Preface ix Needham regarding the venation of the anal area. In the course of this investigation the wings of more than six hundred species of insects, repre- senting eight orders were examined. Nearly all of the figures illustrating this work are reproductions of photographs of wings. In preparing these figures blue prints were made from the negatives and the lines in the prints were inked with waterproof India ink; then the blue was bleached from the figures by the use of a solution consisting of one ounce of potassiimi oxalate dissolved in six ounces of water. This results in producing figures suitable for reproduction by photo-engraving. By this method very accurate figures can be made easily, even of complicated wings like those of the Odonata. John Henry CoiMstock. Entomological Laboratory. Cornell University, November, 191 7. TABLE OF CONTENTS CHAPTER I PAGE The Genesis of the Uniform Terminology of the Wings by Insects i CHAPTER n The Tracheation of the Wings of Insects Method of study of the tracheation of wings 12 The development of wing- veins 12 The hypothetical primitive type of insect wings 15 The tracheation of the wings of larvas and pupas 19 Variations in the extent of the tracheation of the wings of nymphs and of pupse 20 Illustrations of the simpler type of the tracheation of the wings 21 Illustrations of increased tracheation of wings 21 Illustrations of reduced tracheation of wings 22 Eccentric tracheation of wings 23 The basal connections of the tracheae of the wings 25 The Basal Connections of the Trachea of the Wings of Insects. By Royal Norton Chapman 27 CHAPTER III The More General Features of the Wings of Insects The presence or absence of wings in insects 52 The fundamental structure of the wings of insects 53 The different types of insect wings 53 The fan-like wings 53 The elytra 54 The hemelytra 54 The tegmina 54 The halteres 54 The pseudo-halteres 54 The margins of wings ' 54 The angles of wings 54 The tegula 54 The axillar}^ cord 54 The axillary membrane 55 The alula 55 The articulation of the wings 55 The thoracic supports of the wings 55 The axillary sclerites 56 The costal sclerite 57 The tuberosities of the base of the wing 57 The cubito-anal sulcus 57 The corrugations of the wings 57 The cubito-anal fold 57 (xi) xii Table of Contents PAGE The furrows of the wing 58 The anal furrow 58 The median furrow 59 The nodal furrow 59 The axillary furrow 59 The axillary excision 60 The posterior lobe 60 The methods of uniting the two wings of each side 60 The hamuli 60 The frenulum and the frenulum hook 61 The jugum 61 The fibula 61 An expanded humeral angle of the hind wings 63 The clothing of the wings 64 The principal wing-veins 64 The chief branches of the wing-veins of the prcanal area 65 The veins of the anal area 66 The reduction of the number of wing- veins 67 Serial veins ■ 69 The increase of the number of wing-veins 70 The secondary longitudinal veins 71 The accessory veins 71 The marginal accessory veins 72 The definitive accessory veins 72 The intercalary veins 74 The adventitious veins 76 The anastomosis of veins 76 The cross- veins 7^ The humeral cross- vein 78 The radial cross-vein 78 The sectoral cross- vein 78 The radio-medial cross-vein 78 The medial cross-vein 78 The medio-cubital cross-vein 78 The arculus 78 The costal cross- veins 78 Other cross- veins in many-veined wings 79 The terminology of the cells of the wing 79 Additional definitions The cubito-anal excision 8] Convex and coricave veins 81 The ambient vein The humeral veins The pterostigma or stigma The bulte 81 The appcndiculate cell 81 The epipleurae 82 The transverse cord 82 The discal cell and the discal veins 82 Chart of Geologic Time and Formations 84 Table of Contents xiii CmVPTER IV PAGE The Paleontological Data Bearing on the Development and the Specialization of the Wings of Insects (a) Earl}' views as to the paleontological evidence 85 (6) On the origin of wings 87 The paleontological date on the origin of wings 88 (c) On the course of the evolution of each of the principal wing-veins 91 The evolution of the costa 92 The evolution of the subcosta 92 The evolution of the radius 95 The evolution of the media 99 The evolution of the cubitus 1 04 The evolution of the anal veins 1 07 Summary 1 09 CHAPTER V The Development of the Wings of Insects (c) The development of the wings of nymphs no First appearance, position, and growth of the wings of nymphs no The tracheation of the wings of nymphs 113 {b) The development of the wings of larvae and pupa; 114 The development of the wdng-buds 114 The development of the tracheation of the wings 115 CHAPTER VI The Steps in the Specialization of Wings 118 (a) The development of the wings 119 {b) The basal connections of the tracheae that precede the wing-tracheae .... 119 (c) The cross-veins 119 (d) The longitudinal veins 120 Table of the methods of specialization of the wings characteristic of the orders of insects 120 List of orders 122 CHAPTER VII The Wings of the Orthoptera (a) The more general features of the wings of the Orthoptera 123 {b) The wings of the Blattida; 123 (c) The wings of the vSaltatorial Orthoptera i -7 CHAPTER VIII The Wings of the Isoptera (a) The more general features of the wings of the Isoptera 132 {b) The tracheation and the venation of the wings of the Isoptera 135 R^sum6 ... '43 CHAPTER IX The Wings of tiii-; Xkuroptera (a) The more general features of the wings of the Neuroptera {b) The special features of the wings of the Neuroptera 1 ' xiv Table of Contents PAGE The accessory veins 146 The suppression of the dichotomy of the radial sector 147 The radial cuneate area 1 62 The secondary cubital fork 165 Gradate veins 166 The recurrent vein 1 66 The coalescence of veins Sc and Ri 166 Marginal dots or dashes 167 The first radio-medial cross-vein of the hind wings .' 167 (c) The wings of the Sialida3 168 The wings of the Sialinse 168 The wings of the Corydalinae 170 (d) The wings of the Raphidiidas 171 ie) The wings of the Mantispidae 174 (/) The wings of the Ithonidce 175 (g) The wings of the Sisyridse 178 (h) The wings of the Sympherobiidge 178 {i) The wings of the Hemerobiidse 1 80 (7) The wings of the Dilaridae 1 84 (k) The wings of the Berothidas 186 (/) The wings of the Polystnechotidag 187 {m) The wings of the Psychopsidaj 188 (w) The wings of the Chrysopidae 1 89 (0) The wings of the Osmylidse 1 92 ip) The wings of the Myiodactylidse 193 iq) The wings of the Nymphids 195 (r) The wings of the IMyrmeleonidse 196 (5) The wings of the Ascalaphidae 206 (/) The wings of the Nemopteridag 208 (w) The wings of the Apochrysidse 210 (i') The wings of the Coniopterygidse 212 CHAPTER X The Wings of the Ephemerida (o) The more general features of the wings of the Ephemerida 214 (b) The tracheation of the wings of the Ephemerida 214 (c) The homologies of the wing- veins of the fore wings of the Ephemerida . . 217 (d) The corrugations of the wings of the Ephemerida 219 Table of the wing- veins of the Ephemerida 220 (e) The homologies of the hind wings of the Ephemerida 222 CHAPTER XI The Wings of the Odonata (a) The more general features of the wings of the Odonata 224 Table of the wing-veins of the Odonata 228 (b) The special features of the wings of the Odonata 229 The nodus 230 The subnodus 230 The oblique vein 230 The bridge 230 Table of Contents xv PAGE The formation of the bridge 230 The intercalary veins 232 The supplements 234 The antenodal cross-veins 235 The postnodal cross-veins 236 The arculus 236 The triangle of the Anisoptcra 236 The supertriangle 237 The anal veins 237 The anal crossing 237 The secondary anal veins 239 The basal anal area > 239 The cubital area 239 The anal triangle 239 The membranule 239 The chief cubito-anal cross- vein 239 The subtriangle 239 The anal loop 239 The supplemental anal loop 240 The cubito-anal or foot-shaped loop 240 The cubital supplement 240 The quadrangle of the Zygoptera 240 The subquadrangle of the Zygoptera 241 (c) The methods of specialization of the wings of the Odonata 241 {d) Comparison of terminologies of the wing-veins of the Odonata 242 CHAPTER XII The Wings of the Plecoptera (a) The more general features of the wings of the Plecoptera 243 {b) The tracheation of the wings of the Plecoptera 243 (c) The classification of the Plecoptera 244 (d) The special features of the wings of the Plecoptera 245 The anastomosis or the transverse cord 247 The pterostigma 247 The basal anal cell 247 (e) The primitive plecopterous type of wings 248 (/) The methods of specialization of the wings of the Plecoptera 254 Specialization by reduction 254 Specialization by addition 255 CHAPTER XIII The Wings of the Corrodentia 258 CHAPTER XIV The Wings of the Embiidina 262 CHAPTER XV The Wings of the Thvsanopteka 267 ■ xvi Table of Contents CHAPTER XVI PAGE The Wings of the Homoptera 269 (a) The more general features of the wings of the Homoptera 269 (b) The wings of a cicada 271 (c) The wings of the Membracida 274 (d) The wings of the Jassidas 280 (e) The wings of the PsyUid£e 283 (/) The wings of the Aphidae 285 (g) The wings of the Aleurodidse 289 (h) The wings of the Coccidae 290 (i) Supplemental Note 291 CHAPTER XVn The Wings of the Heteroptera 292 (a) The more general features of the wings of the Heteroptera 292 (b) The tracheation of the wings of the Heteroptera 292 (c) The veins and furrows of an adult wing 294 CHAPTER XVni The Wings of the Dermaptera 295 CHAPTER XIX The Wings of the Coleoptera 297 CHAPTER XX The Wings of the Strepsiptera ' 301 CHAPTER XXI The Wings of the Mecoptera (a) The more general features of the wings of the Mecoptera 302 (b) The venation of the wings of the typical Mecoptera 302 (c) The aberrant Mecoptera 303 CHAPTER XXII The Wings of the Trichoptera suborder phryganeina , 307 (a) The more general features of the wings of the Phryganeina 307 (b) The tracheation of the wings of the Phryganeina 308 (c) The phryganeid type of wing- venation 308 (d) The more general features in the specialization of the wings of the Phryganeina 311 (e) The methods of uniting the two wings of each side 312 SURORDER MICROPTERYGINA 313 (a) The venation of the wings of the Microptcrygina 314 (6) The tracheation of the wings of the Microptcrygina 315 (c) The zoological position of the Microptcrygina 317 Table of Contents xvii CHAPTER XXIII PAGE The Wings of the Lepidoptera (c) The more general features of the wings of the Lepidoptera 319 (b) The clothing of the wings of the Lepidoptera 319 The clothing of scales 319 The clothing of fixed hairs 323 (c) Methodsof specialization of the wings of the Lepidoptera 324 (d) The primary divisions of the Lepidoptera indicated by the structure of the wings 325 Suborder Jugatae 325 Suborder Frenatae 326 (e) The wings of the Jugatae 327 The jugum 327 The tracheation of the wings of pupae 327 The venation of the wings of hepialids 328 (/) The wings of the Frenatae 330 The frenulum and the frenulum hook 330 The loss of the frenulum in certain of the Frenatae 331 The reduction of the radius of the hind wings 332 The reduction of media to a three-branched condition 334 The atrophy of the main stem of media 337 The transfer of the branches of media to adjacent veins 338 The first anal vein and the anal furrow 340 The reduction of the anal area 340 The anastomosis of veins 342 The costa of the hind wings 342 The humeral veins 343 The splitting of the radial sector in butterflies 343 (g) Comparison of terminologies of the wing- veins of the Lepidoptera 345 CHAPTER XXIV The Wings of the Diptera (a) The more general features of the wings of the Diptera 347 (b) The methods of specialization of the wings of the Diptera 348 Ordinal specializations 348 The loss of vein M4 348 Specializations within the order 350 The costa 350 The subcosta 350 The radius 350 The media 353 The coalescence of veins M3 and Cui 353 The coalescence of veins Cu2 and the second anal vein 355 The reduction of the number of cells in a wing 355 The anal veins 357 The arculus 358 (c) Comparison of terminologies of the wing- veins of the Diptera 358 (d) Comparison of terminologies of the cells of the wings of Diptera 360 xviii Table oj Contents CHAPTER XXV page The Wings of the Hymenoptera (a) The general features of the wings of the Hymenoptera 362 {h) The venation of the wings of the more generahzed Hymenoptera 362 (c) The tracheation of the wings of the Hymenoptera 368 {d) Methods of modification of the primitive hymenopterous type of wing- venation 370 Reduction by atrophy 371 Reduction by coalescence 374 Modification of the course of a vein by the coalescence at base 375 Modification of the course of a vein by the coalescence of its tip with an adjacent vein 375 Serial veins 376 (e) An illustration of the reduction of the wing- venation in the Hymenoptera 379 CHAPTER XXVI The Teaching of the Uniform Terminology of the Wing-Veins of Insects Suggestions to teachers 382 Outline of laboratory work in the study of the venation of the wings of insects 387 Bibliography 417 Index 424 THE WINGS OF INSECTS THE WINGS OF INSECTS CHAPTER I THE GENESIS OF THE UNIFORM TERMINOLOGY OF THE WING-VEINS OF INSECTS From the earliest days of systematic entomology to the present, much use has been made of the wings in the classification of insects. The names of the Linnean orders were all suggested by the nature of the wings except one, Aptera, and that by the absence of wings; many of the established families of insects are most easily recognized by peculiarities in the form of the wings; and hundreds of genera have been distinguished by details of wing-venation. A reason for the extensive use of characteristics of wing-structure in the classification of insects is the ease with which they can be observed. Owing to the broadly expanded form of the wings, it is more easy to see variations in their structure than it is to see variations in the form of other parts of the body. The wings are open pages upon which are to be read, with compara- tive ease, the results attained by the processes of evolution. Although the wing-characters have been used in the classification of insects from a verv'' early period, a full appreciation of their importance for this pur]Dose is a matter of comparatively recent growth. It was not until it was demonstrated that the wings of all winged insects are modifications of a single primitive type that it was possible to fully understand the taxo- nomic value of these organs. The early entomologists who worked without the aid of the modem conception of organic evolution, only vaguely attempted to recognize those homologies of the principal wing-veins that we now know to exist. The result was the establishment of many different systems of terminolog}'-, the students of each order having a system of their own, and in some cases several different systems were used by the students of a single order. It would require too much space to explain all of these systems. The student who uses the older books will need to leani the particular system employed by the author he is studying. In later chapters, under the dis- cussion of the venation of the wings of the separate orders, a table is given, when practicable, in which one or more of the older s>-stems is compared with that of Redtenbacher and with the one adopted in this work. (1) 2 THE GENESIS OF THE TERMINOLOGY It seems probable that the short paper published by Dr. Hagen in 1870, entitled " Ueber rationelle Benennung des Geaders in den Fliigeln der Insekten" was the beginning of the series of efforts to establish a system of terminology of the wing-veins of insects applicable to all orders of insects. For the author of the "Bibliotheca Entomologica" was very familiar with the literature of entomology then published, and had there been earlier efforts in that direction he would have quoted them. It is with much pleasure that I remember that my own interest in this subject was first awakened by a lecture Dr. Hagen delivered to me, his first American student, in the summer of 1872. In this lecture he called my attention to the paper referred to above, then only two years old, and urged the importance of investigations in this field. Dr. Hagen recognized six longitudinal veins, four principal veins and a branch of each of two of these. His paper is accompanied by a plate, which is reproduced here (Fig. i) ; but no explanation of the plate is given, and there is no reference to it in his text. It is evident, however, from his descriptions of the veins that he recognizes, that the significance of the lettering of the plate is that indicated below. The six longitudinal veins recognized by him are named as follows: subcosta (a), mediana (h), hind branch of the mediana (6^), front branch of the submediana (c^, sub- mediana (c), and postcosta {d). Dr. Hagen' s view regarding the practicability of establishing a unifonn terminology of the wing-veins met with little if any favor among the leaders in entomology. For example, Dr. Graber, in his " Die Insekten" published in 1877 (Vol. I, p. 196), in referring to the effort to establish a imiform tenninology says that, as a matter of course, no scientific significance should be attached to such attempts. Even as late as 1895 Professor David Sharp, in his most excellent text book of entomology (Camb. Nat. Hist. Vol. 5, p. 107-8), writes as follows: — "Various efforts have been made to establish a system of nomenclature that shall be uniform throughout the different Orders, but at present suc- cess has not attended these efforts, and it is probable that no real homology exists between the nervures of the different Orders of Insects." This statement by Dr. Sharp was made in spite of the fact that the classic contribution of Josef Redtenbacher, " Vergleichende Studien iiber das Fliigelgeader der Insecten" (1886) had appeared nearly a decade before; in fact, it was probably this paper that suggested it; for the obscurity into which Dr. Hagen's paper had fallen would have saved it, standing alone, from comment. It was not till the appearance of Redtenbacher' s paper that any great progress was made in the establishment of a uniform terminology of the wing- veins. This paper with its numerous illustrations drawn from nearly Fig. I. — The plate illustrating Dr. Hagen's paper "Ueber rationelle Benenung des Geaders in den Fliigeln der Insekten." (3) 4 THE GENESIS OF THE TERMINOLOGY all orders of winged insects, is really the starting point in the actual solution of the problem. Before referring farther to Redtenbacher's work, it is desirable to discuss briefly some publications that had a profound influence on it. These treat of the relation of the tracheation of the wings to the wing-venation and of the supposed existence of two morphologically distinct types of wing-veins, the convex and the concave. The earliest reference that I have found to the relation of the trachea- tion of the wings to the wing- venation is that of Carl Semper ('57) who stated that in the development of the wing the direction and course of the wing-veins are determined by the tracheas the principal branches of which indicate the wing-veins when of the latter no trace is to be seen. Semper, however, did not obtain a correct understanding of the method of development of the wing-veins. He represented a cylinder of epithelial cells, which he states secrete later on their inner surface the chitin that forms the vein. It is evident that he considered the wing-vein as a separate structure from the membrane of the wing. It is an interesting fact that Semper represented a nerve within the wing- vein. Such nerves are now well-known and the connection of their branches with scales and setae has been figured by several writers. Landois ('71) figures the larval wings of a Vmiessa and calls atten- tion to the now well-known fact that the larval tracheoles of the wing, the "geknauelten Tracheen," occupy exactly the position occupied later by the wing-veins. He also describes the degeneration of these tracheoles and the development of the permanent tracheae in the same position. He overlooked the fact that the permanent trachese are developed during the last larval stadium, observing them first in the pupal instar. A more serious error on the part of Landois was the description of elastic cords covered with an epithelium and closely parallel with the permanent tracheas, in the pupal wing. These he termed the "Fliigel- rippen" and states that they lie close upon the lower membrane of the wing. According to his view a completed wing-vein consists of one of these "Fliigelrippen sensu strictiori," tracheae, and the united upper and lower membranes of the wing enclosing these parts and a space in which the blood flows. I can not imagine what Landois saw and described as the "Fliigel- rippen sensu strictiori." I have been unable to find any such structure in any of the wings that I have studied. Spuler ('92) however, gives a figure of a cross-section of a wing-vein in which he represents it, and refers to it in his text as the so-called "Ri])pe" of Semper, but gives no description of it. In 1879 the "Ueber Insectenflugel" by G. Ernst Adolph appeared. The publication of this work was unfortunate ; for it added little of value to our THE GENESIS OF THE TERMINOLOGY 5 knowledge and did much to retard the progress of the solution of the problem of which it treated. In this work Adolph elaborated a theory of alternating convex and concave veins. This was based on the fact, already pointed out by several writers, Selys-Longchamps, Saussure, Hagen, and others, that as a result of a corrugation of the wings some of the veins extend along raised lines and others along stniken ones. He also made use of the results of Landois concerning the relation of the trachese of the wings to the wing-veins. According to this theory the wing trachea; determine the position of the concave veins, which represent the foundation of the vein system. Only later is each tracheal tube inclosed by a mass of chitine and thereby trans- formed into a vein. At the same time, according to Adolph, the tracheae force the two wing-plates apart and cause a thinning of the membrane, which shows itself, among other ways, by the fact that the wing regularly tears under pressure or strain along the concave veins. Between these primary, or concave veins appear later thickenings of the wing-membrane in the form of chitine lines with which the tracheal tubes and blood channels are finally associated, and which form the secondary, or convex veins. The two sorts of veins, accordingly, stand in direct opposition to each other; since the first is produced by a thinning, and the last by a thickening of the wing-membrane; and since, in the former, the tracheal tube, in the latter, the chitinous lines represent the primary foundation. Subsequent investigations have shown that there is little to support this remarkable theory. There is only one method of formation of the principal wing-veins; and the differences in position of the veins, that is along raised lines or along sunken ones, is merely the result of a secondary coiTugating of the wings. Redtenbacher was profoundly influenced by the theory of alternating convex and concave veins elaborated by Adolph and made his system of terminology conform to it absolutely. He recognized five fields in the wing each traversed by a convex vein. These veins he designated by the consecutive odd Roman characters: I, HI, V, VH, and IX. He also applied to these veins the names costa, radius, media, ctibitus, and anal vein. These names he selected from the many names then in use except in the case of media, which was proposed by him for the vein to which that name is now applied; and which previously had been generally regarded as belonging partly to the radius and partly to the cubitus; although it had been recognized as a principal vein by Edward Doubleday long before. See his "Remarks on the Genus Argynnis" (Trans. Linn. Soc. vol. XIX, 1845)- To the concave veins, which Redtenbacher, following Adolph, believed to alternate with the convex veins, be applied the consecutive even Roman 6 THE GENESIS OF THE TERMINOLOGY characters: II, IV, VI, VIII, X. In the case of wings with an expanded anal area the numbers XII and XIII are also used. The only so-called concave vein to which he applied a name is the subcosta. In subsequent modifications of his system the six names used by him, costa, subcosta, radius, media, cubitus, and anal veins have been adopted. The branches of the principal veins were designated by adding Arabic numerals to the Roman numeral designating the principal vein. Thus the branches of vein III are designated successively as IIIi, III2, III3, III4, and Ills. It was unfortunate that Redtenbacher was misled by the erroneous theory of alternating convex and concave veins elaborated by Adolph. The result was that, although Redtenbacher recognized the homologies of the main stems of the principal veins, he, in his efforts to apply this theory, was led into many serious errors. The result is that the terminology now adopted differs much in detail from that of Redtenbacher upon which it was based. These differences are indicated in later chapters. Redtenbacher was also misled by the then commonly accepted view of the paleontological data bearing on the evolution of wings. Stating this view he writes (page 155) as follows: — • The geologically older Orthoptera and Neuroptera show a much richer venation than the Coleoptera, Lepidoptera, Hymenoptera, and Diptera; likewise among the Rhyncota, the oldest forms, the Cicadidse and the Fulgoridffi, possess much more numerous veins than the Hemiptera. There is apparently, then no doubt, that the oldest insect forms were provided to a certain extent with a superfluity of veins, and that, in the course of develop- ment, all the superfluous veins disappear by reduction, and, in this way a simple system of venation was brought about. The acceptance of these two erroneous views, the theory of alternating concave and convex veins and the belief that the first winged insects had many wing-veins, did much to retard the progress of the efforts to establish a uniform terminology of the wing-veins. Redtenbacher's adherence to the theory of Adolph was short lived, however; for two years after the publication of his larger paper one appeared, of which he was the junior author (Brauer und Redtenbacher, 1888), in which the theory of different origins of the concave and convex veins is discredited. But there is no suggestion in this paper of a modifica- tion of the Redtenbacher system of terminology of the wing-veins. With the appearance in 1891, 1892, and 1893 of papers by Dr. Erich Haase and by Dr. Arnold Spuler, there began a series of modifications of the Redtenbacher terminology. The first of these three papers was a brief abstract by Haase ('91) of his conclusions regarding the development of the wing-veins of Papilio THE GENESIS OF THE TERMINOLOGY 7 Machaon. Then followed the more general work of Spuler ('92) entitled "Zt:r Phylogenie und Ontogenie des Fltigelgeaders der Schmetterlinge." And last of the three was the completed work of Haase ('93), his "Unter- suchungcn iiber die Mimicry auf Grundlage eines naturlichen Systems der Papilioniden." In this paper Haase gave excellent figures of the tracheation of the wings of the pupa of Papilio Machaon and showed the relation of the trachea? to the wing-veins of the adults; and he designated the wing- veins in his figures, by abbreviations of the names that he adopted for them instead of by numbers. Like Spuler and Bauer and Redtenbacher, he regarded costa as merely the thickened margin of the wing and not a vein; and as the first anal trachea coalesces with the cubital trachea at the base in this species, he believed cubitus to be three-branched. The second and third anal veins he designated as the first and second branches of the "dorsal vein." The veins that he recognized are subcostalis, radialis wnth five branches, mediana with three branches, cubitalis with three branches, and dorsalis with two branches. Spuler, in his paper which followed the brief abstract published by Haase but which appeared before the publication of the completed work by Haase, recognized two areas in the wing: an expanded area, which he termed the "Spreitentheil;" and a folded area, the "Faltentheil." The veins in the former, he numbered with Roman numerals ; those in the latter, with Greek letters. He began his numbering with the subcosta, which he designated as vein I; for, as he failed to find a trachea in the costa, he did not regard it as a vein. He also omitted from his series of veins, and very correctly, the supposed concave veins IV and VI of the Redtenbacher system. The result was that he recognized five veins in the expanded area of the wing, which he designated as veins I, II, III, IV, and V respectively, and two veins in the folded area, vein a and vein g.. The veins that he recognized in the expanded area are those now known as the subcosta, radius, media, cubitus, and ist anal; those of the folded area are the 2d and 3d anal veins. Probably the most important result of Spuler' s investigations was the determination of the type of the lepidopterous wings by a study of the tracheation of the wings of many lepidopterous pupa? and of the venation of the wings of a large series of adults. Packard ('95) published an extended review of Spuler's paper and ftirnished some additional matter and figiires. He adopted the terminology of Spuler except that he numbered the anal veins V, VI, and VII. He followed Spuler in not regarding the costa as a vein and applied the name costa and the number I to the subcosta and the name subcosta and the 8 THE GENESIS OF THE TERMINOLOGY niimber II to the radius, following Heinemann and other of the older writers on the Lepidoptera as regards the names of these veins. Before the publication of Spuler's paper the writer had made an extended study of the evolution of the wings of the Lepidoptera. His more important conclusions were presented in a lecture before the California Zoological Club, Jan. 30, 1892, an abstract of which was published in Zoe. vol. Ill, pp. 84-86. This paper includes the suggestion that the Lepidoptera should be divided into two suborders, the Jugatse and the Frenatas. Later in the same year the writer, to whom the works of Haase and of Spuler were still unknown, read a paper before the American Association for the Advancement of Science (see Proc. A. A. A. S., 41st Meeting, Aug. 1892) entitled "The Descent of the Lepidoptera. An Application of the Theory of Natural Selection to Taxonomy." In the following year (1893) the results of the studies briefly outlined in the two preceding communications were published in a paper entitled "Evolution and Taxonomy. An Essay on the Application of the Theory of Natural Selection in the Classification of Animals and Plants, Illustrated by a Study of the Evolution of the Wings of Insects, and by a Contribution to the Classification of the Lepidoptera" (The Wilder Quarter-Century Book.) Although no changes in the Redtenbacher system of terminology of the wing-veins of insects was suggested in these communications, they are quoted here because there is no doubt that the later consistent application of the method of study indicated in them hastened the putting of the uniform terminology of the wing-veins on a firm basis. The method is briefly outlined in the following quotation from the abstract published in the Proceedings of the American Association (/. c.) : "There is indicated in this paper a method of applying the theory of natural selection to taxonomy somewhat more fully than seems to have been done before. "The method consists in beginning with the study of a single organ possessed by the group of organisms to be classified. The variations in form of this organ are observed ; the function of the organ is studied ; and an effort is made to trace out the phylogenetic development of the organ, keeping constantly in mind the relation of the changes in form of the organ to its function. In other words, the record of the action of natural selection upon the group of organisms is read as it is recorded in a single organ. This gives data for a provisional classification of the group. Then another organ is selected and its history worked up in the same way. Then the results obtained in the two investigations are compared; and where they differ there is indi- cated the need of renewed study. For if rightly understood the different records of the action of natural selection will not contradict each other. The investigation is continued by the study of other organs and a correlating of results obtained until a consistent history of the group has been worked out. "This method differs from that commonly employed, in being a constant effort to determine the action of natural selection in the modification of the form of organisms, THE GENESIS OF THE TERMINOLOGY 9 in order to better adapt their parts to perform their functions. By the old method a search is made for characters by which a group can be divided and subdivided with but little regard to the meaning of these characters. In fact we rarely see in purely taxono- mic works any reference to the functional significance of the characters used. "As illustrating the proposed method of study, the structure and function of the wings of the Lepidoptera are discussed and conclusions arc drawn from this study regarding the phylogenetic development of the order." The next step in advance in the development of the nniform nomen- clature of the wing-veins was made by the writer, who worked out the homologies of the wing-veins of the Lepidoptera, Diptera, and Hymehop- tera and published his results in "A Manual for the Study of Insects" (1895). In the preface of the book he said: "The principal features of notation of wing- veins proposed by Josef Redtenbacher have been adopted. But as the writer's views regarding the structure of the wings of primitive insects are very different from those of Redtenbacher, the nomenclature proposed in this book is to a great extent original. The chief point of difference arises from the belief by the present writer that veins IV and VI do not exist in the Lepidoptera, Diptera, and Hymenoptera; and that, in those orders where they do exist, they are secondary developments." I had already reached the conclusion that the few-veined wings of Hepialus represented the primitive type more closely than the fan-like wing of the Ephemerida, as was commonly believed. But as I had not made, at that time, careful studies of the wings of insects with many wing- veins I was willing to admit that the veins IV and VI of the Redtenbacher system might exist as secondary developments in those orders with fan-like wings. The investigations of the writer referred to above were based on examinations of the wings of adult insects. Soon after the publication of the volume just cited, Mr. J. G. Needham, then a graduate student in the entomological laboratory of Cornell University, and the writer began the investigation of the development of the wings of representatives of all of the orders of insects of which we could obtain living n^nnphs or pupae. The results of these ontogenetic studies were published in a series of articles entitled "The Wings of Insects," which appeared in the American Naturalist during the years 1898 and 1899. This series of articles established on a firm basis the uniform terminology of the wing-veins. The results of our investigations are given in detail in subsequent chapters. In recording the results of our studies we found it necessary to introduce an important modification of the Redtenbacher system. As the non existence of veins IV and VI of the Redtenbacher system was demon- strated, and as it was desirable that the numbering of the principal veins should be continuous, a modification of the Redtenbacher system was obviously desirable. But we were met by another difficulty due to the fact that Spulcr and others did not regard the costa as a vein and began their 10 THE GENESIS OF THE TERMINOLOGY numbering of the veins with the subcosta which they designated as vein I. These facts led us to make the following statement : "In designating the wing- veins they may be either named or numbered. The simplest method is, doubtless, to number them; and had the system which was proposed by Redtenbacher been based on a correct understanding of the primitive type, nothing better could be desired. But it was not; and, as several modifications of the Redten- bacher system are already in use, it seems doubtful if uniformity in numbering them could be soon brought about. "From the great mass of names that had been proposed for the principal wing-veins, Redtenbacher selected a set of terms, to the acceptance of which no objection has been urged. It seems, therefore, that the surest way to bring about uniformity of nomen- clature is to give up the attempt to apply a set of numbers to the wing-veins, and to use the names adopted by Redtenbacher. These names and the abbreviations of them which we shall use in our text as well as in the figures illustrating it, are as follows: Costa, C Media, M. Subcosta, Sc. Cubitus, Cu. Radius, R. Anal veins, A. "In designating the branches of the forked veins we have adopted the principle of numbering them proposed by Redtenbacher and combine the numbers with the abbreviations of the names of the veins. Thus, the first branch of radius is designated as radius-one; and for this term the abbreviation R\ is used." The terminology of the wing-veins elaborated by Comstock and Needham in the series of articles referred to above is now commonly designated as the Comstock-Needham system. But it should be remem- bered that this system is merely a modification of that proposed by Red- tenbacher; and the fundamental work of this author should not be over- looked. Redtenbacher' s system, being based upon the erroneous theory of Adolph was not available without modification. This had been shown by Brauer and Redtenbacher, Haase, and Spuler. But as each of these authors maintained that costa is not a vein, an incorrect starting point for the numbering of the veins was indicated. Comstock and Needham demonstrated that costa is preceded by a trachea and is therefore a true vein. Previous to the publication of their articles Spuler had worked out the lepidopterous type of venation (except as regards costa) , and Comstock had also worked out that of the Lepidop- tera, and in addition that of the Diptera, and Hymenoptera. Taking up the work at this point, Comstock and Needham extended it by a study of the development of the wings of representatives of nearly all of the orders of winged insects, and presented a hypothetical type of primitive wing- venation from which they believe that' the venation of the wings of all orders have been evolved. They also pointed out the various methods of speciahzation by which the primitive type has been modified. This placed the uniform terminology on a broad and firm basis. THE GENESIS OF THE TERMINOLOGY 11 Although the unifonn terminology adopted here is based on that of Redtenbacher, the modifications of the older system that have been found necessary have resulted in marked differences in detail between the two systems. These differences are indicated later in the discussions of the venation of the wings of the several orders. Since the appearance of the series of articles by Comstock and Needham, the homologies of the wing-veins of several groups of insects have been studied by various authors. The more important of the contributions on this subject are referred to and utilized in later chapters; and are listed in the bibliography at the end of this essay. The work that remains to be done to perfect the uniform terminology of the wing-veins is the application of this system in the descriptions of those highly specialized groups of insects where the homologies of the wing-veins are difficult to determine. This can be well done in each case only by a specialist who makes a thorough study of the wings of the group, family or order in question, and compares these wings with those of more generalized forms. This work will doubtless reqtiire many years for its completion; but the value of the results in determining the relationships of the groups studied will be worth the effort. CHAPTER II THE TRACHEATION OF THE WINGS OF INSECTS A Study of the developing wings of nymphs and of pupse has shown that in the more generahzed orders of insects the principal, longitudinal wing- veins are formed about tracheae, which extend into the wing early in its development and before the beginning of the fonnation of the wing-veins. As these trachea are very constant in number and position, they furnish the most available data for determining the homologies of the wing-veins that are later developed about them. An account of the earlier publications on the relation of the tracheation of the wings to the wing-venation is given in the preceding chapter. A comparative study of the tracheation of the wings of the several orders of insects for the purpose of determining the homologies of the wing-veins was first made by Comstock and Needham ('gS-'gg). That work of this kind had not been undertaken previously was doubtless due to difficulties that stood in its way. The tracheae of the wings of pupae and of nymphs are often very delicate ; and, when filled with the mediiun in which a wing is mounted for micro- scopic study, they are usually invisible. It is not strange, therefore, that the study of them was so long delayed. Method of study of the tracheation of wings. — In the course of our investigations Dr. Needham and I devised a method of study of the wings of immature insects, which renders the observation of the tracheee in them a simple matter. A description of this method, compiled from our joint account of it, with a few changes made necessary by subsequently obtained information is given in the outline of laboratory work included in Chapter XXVI of this vohmie. By this method most beautiful objects can be prepared, which will show the minutest ramifications of the tracheae. Plate II is a half-tone reproduction of a photograph of an object prepared in this way. This figure represents a small portion of a wing of a pupa of Corydalus cornutus. The development of wing-veins. — "Not only can the tracheae that precede the wdng-veins be studied in this manner, but if the wing be taken at the right stage, pale bands can be seen which indicate the position of channels about which the veins are later developed; this is shown in the figure of a wing of a pupa of Sialis (Fig. 2). An examination of a cross- section of a developing wing (Fig. 3) will show the nature of these channels, or vein-cavities, as they may be termed. "In all insect wings the two plates of the hypodermis constituting the wing fold are at first separate, i. e. not fused internally. At the time when (12) I PLATE II Part of a wint; of a pupa, of Corydalus corn ut lis. iProm C. and N'.) THE TRACHEATION OF WINGS 13 the tracheas enter the fold the two layers become approximated along lines midway between the trachece, resulting in an actual fusion of the internal ends of the cells. Then follows a gradual lateral extension of the areas in Fig. 2. — A wing of a pupa of Sialis 0\ftcr C. & N.)- which the cells are fused to delimit definitely the channels through which the tracheas pass. This process is illustrated by the figures showing two stages in the development of the wings of Psocus (Figs. 4 and 5). "The pale color of the bands, indicating the extent of the vein cavities when viewed by transmitted light, is doubtless due to the fact that the haemolymph filling these cavities is more translucent than the hypodermal tissue, which completely fills the wing elsewhere. "As an illustration of the nature of the data that can be obtained by the method described above, a half-tone reproduction of one of our photographs is printed here (Plate I Frontispiece). This figure represents a wing of a nearly mature m^mph of a Nemoura, one of the genera of the Plecoptera. In making the preparations it was impracticable to remo\'e all of the dirt adhering to the wing without danger of injuring it ; this is often the case in preparing mounts of the wings of aquatic nymphs. The irregular blotches Fig. 3. — Cross section of a fore wing (in part omitted) of a nymph, two-thirds grown, and recently molted, of Anax Junius (After C. Tnphs and of pupae of the several orders for the purpose of determining what features were common to all; for, as it is evident that features common to all must have been inherited from a 16 THE TRACHEATION OF WINGS common ancestor, it was believed that such a study would reveal the more important features of the tracheation of the wings of 'the primitive winged insects, and, consequently, of the wing-venation of the same. S<^^ Sc, Fig. 6. — Hypothetical tracheation of a wing of the primitive nymph (After C. & N.). The result of our studies enabled us to construct a diagram representing the hypothetical tracheation of a wing of the primitive nymph (Fig. 6). Since there is no doubt that the tracheae of the fore wings and of the hind wings are homodynamous, a single diagram serves to represent the trachea- tion of either the fore wing or the hind wing of the hypothetical primitive nymph. In order to avoid unnecessary repetition the data tipon which our conclusions regarding the probable primitive type of wing-venation were based are not given here. In the discussions of the wing-venation of the different orders of insects, given later, the fundamental type of the wing- venation of each order is described and its con^espondence with the hypothe- tical type demonstrated ; and in Chapter IV the correlation of the paleon- tological data with that drawn from the study of living insects, as regards the typical form of each of the wing-veins is indicated. In Figure 6, and in other figures representing the tracheation of wings, the tracheae are designated by the abbreviations of the names of the veins with which they correspond. In referring to trachese in the text they are designated in one of two ways. Thus the trachea that precedes the radius, for example, may be designated as the radial trachea or it may be termed trachea R. The trachea that precedes the first branch of the radius is designated as trachea R\ ; and other branches of the branched tracheae are designated in a corresponding manner. It was found that the basal connections of the wing tracheas afford characters of considerable taxonomic importance. In what we regard as the more generalized of winged insects, i. e., in the Plecoptera and in certain THE TRACIIEATION OF WINGS 17 members of the family Blattidae of the Orthoptera, there are two distinct groups of tracheae that enter the wing; these we designated as the costo- radial group and the cuhito-anal group respectively. In these insects the medial trachea is a member of the costo-radial group (Fig. 6). In all other forms that we studied, except the Ephemerida, a transverse trachea connects these two groups of wing trachea;; the position of this transverse basal trachea of the wing is indicated in the figure (Fig. 6) by dotted lines.* Frequently the transverse basal trachea is indistinguishable from the two main trunks which it connects, the three forming a single, continuous, transverse trachea, from which arise all of the wing tracheae (Fig. 7). All of the stages of this development were found by us within the Orthoptera. We also found that when a transverse basal trachea is formed, the medial trachea tends to migrate along it toward the cubito-anal group of Fig. 7. — \Yings of an acridid nymph (After C. & N.). tracheae, and often becomes united with that group. This is well shown in certain Orthoptera and in the cicada. In some cases the base of the radial *More recently it has been found that in certain Homoptera, the Membracidae and the Jassidas, the same generalized condition of the basal connections of the wing tracheae exists as in the Plecoptera. 18 THE TRACHEATION OF WINGS trachea tends to follow that of the medial trachea in its migration along the transverse basal trachea towards the cubito-anal group (Fig. 7).* As the diagram representing the hypothetical type of the tracheation of the wings wih serve also to represent the supposed primitive type of wing- venation, there may be indicated in it the position of a few cross-veins, which are so constant in their position, and which occur in so many widely separated groups, that they are regarded as being comparatively primitive. This is done in Figure. 10. None of the cross-veins are primitive in the sense in which the longi- tudinal veins are primitive; for in the wings of the older Paleozoic insects Fig. 8. — Stenodictya lobata (After Handlirsch). there were no distinct cross- veins, but an irregular network of thickened lines; this is weU-shown in the wings of Stenodictya (Fig. 8), which is described in Chapter IV. In the older insect wings in which there are distinct cross-veins there are many of them, and the few cross-veins that are referred to above, and *Enderlein ('02) has worked out with great care the tracheation of the wings of Antheraea pernyi, a saturniid moth, and the connections of the tracheae of the wings with the longitudinal tracheas of the body; and his account is illustrated by inost excellent figures. In this highly specialized moth, media has completed its migration along the trans- verse basal trachea and becpme a member of the cubito-anal group of tracheae. This fact has probably misled Enderlein, for he argues at length that the veins of each wing are separated "in zwei genetisch vollig verschiedene Systeme, das radiate und das mediane Adersystem" (/. c. p. 28). He quotes the separation of the veins of the wing, by Comstock and Needham, into the costo-radial group and the cubito-anal group, but evidently overlooked our account of the migration of media; for he proposes the division of the wing-tracheaj into a radial and a medial grotip, without any reference to the facts that led us to make the division indicated above. THE TRACHEATION OF WINGS 19 that we have named, can not be identified ; one of the oldest of these wings is that of Eitrythmopteryx (Fig. g), which is also discussed in Chapter IV. ''^ '^ Oil 1 Fig. 9. — Wing of Eurythmoptcryx (After Handlirsch). In the course of the evolution of the few-veined type of wing nearly all of the cross- veins were lost ; but five or six were retained and are still pre- served in several orders of insects. These we have named as follows : the humeral cross-vein (Fig. 10 h)\ the radial cross-vein (Fig. 10, r); the sectorial cross-vein (Fig. 10, s); the radio-medial cross-vein (Fig. 10, r-m)\ the medio-cubital cross-vein (Fig. 10, m-cii); the medial cross-vein (Fig. ID, m)\ and the posterior arculus (Fig. 11, p. a.) The cross-veins are described more fully in the next chapter. The tracheation of the wings of larvae and pupae. — The development of the tracheation of the wings in insects with a complete metamorphosis, where the wings reach an advanced stage of development within the body, differs remarkably from that of the wnngs of insects with a gradual or with an incomplete metamorphose. With the latter the wings of the njTnphs Fig. 10. — The hypothetical primitive type of wing venation with the named cross-veins added. are developed as outward projecting appendages of the body, into which the tracheae penetrate early; and there is a gradual, direct development of the tracheation in the successive stadia. But in insects with a complete 20 THE TRACHEATION OF WINGS metamorphosis the development of the tracheation of the wings presents some of the more remarkable features of this highly specialized mode of development. As an understanding of the more general features of the internal develop- ment of wings is essential to an understanding of the development of the RJrM^ Fig. II. — Diagram of an arculus of a dragon-fly. tracheation of the wings of larvse and pupae a discussion of this subject is omitted here, and will be taken up in Chapter V. Variations in the extent of the tracheation of the wings of nymphs and of pupae. — While the study of the tracheation of the wings of njTiiphs and of pupag has yielded most valuable data for determining the homologies of the wing-veins of adult insects, an extended investigation in this field has revealed remarkable differences in the extent of the tracheation of the Fig. 12. — The wings of a nymph of Nemonra (After C. & N.). wings of immature forms and in the degree of its correspondence with the venation of the wings of adults. In some of the orders of insects, the tracheation of the developing wings is of the greatest importance in deter- THE TRACHEA TION OF WINGS 21 mining the homologies of the wing-veins; in others, it is of little or no value for this purpose. And, as a rule, there is throughout each order a marked uniformity in this respect. The variations in the extent of the tracheation of developing wings can be grouped under a few general heads : in certain orders the tracheation is comparatively simple; in others, there has been developed a greatly increased tracheation of the wings ; while in still others the tracheation has been greatly reduced; and in one order at least, the tracheation is eccentric. Examples of each of these classes are given below. Illustrations of the simpler type of the tracheation of the wings. — Excel- lent illustrations of the simpler t}'pe of the tracheation of the wings of n}TTiphs are presented by members of the Plecoptera, of which Nemoura (Fig. 12) may be taken as an example. The figure represents the wings of a ^a^ Fig. 13. — A wing of a pupa of Chaidiodes (After C. &' N.). grown nymph in which the developing veins appear as pale bands. In each of the principal veins and in the branches of these there is a prominent trachea; while the cross-veins are without tracheae.* Wings in which the tracheation is of this type present few if any difficul- ties in the determination of the homologies of the wing- veins. This is especially true as the cross-veins are sharply differentiated from the princi- pal veins by the absence of tracheae in them. Of the other orders that exhibit a simple type of tracheation of the ^vings there may be cited : the Corrodentia, of which Psociis (Fig. 5) wall serve as an example; the Homoptera, illustrated by a cicada, the wings of which are figured later; and the Lcpidoptera, which is discussed in detail in another chapter. Illustrations of increased tracheation of wings. — In the simpler type of tracheation of the wings only the primitive wing-veins, those represented in our h^T^othetical primitive type, are represented by prominent tracheae. *In the specimen figured the costal trachea was not observed; but this trachea was found in other,' closely allied, forms. 22 THE TRACHEATION OF WINGS What may be regarded as the first step in the direction of an increased tracheation of the wings is illustrated by the Neuroptera, of which Chaulio- des (Fig. 13) will serve as an example. Here the primitive wing- veins and the secondarily developed accessory veins are preceded by tracheae, while the cross- veins are not. Although many fine tracheae branch from the principal tracheae and ramify throughout the wing sac they do not follow the courses of the cross-veins. The extreme limit of increased tracheation is reached by the Odonata, where not only the principal veins and the intercalary veins are traversed by tracheae, but there is also a distinct segregation of small tracheae and tracheoles in the cross-veins; this is shown in Plate V. In this order the study of the tracheation of the wings affords no help in distinguishing the cross-veins from other veins. Illustrations of reduced tracheation of wings. — In several of the orders of insects a remarkable reduction of the tracheation of the wings has taken place; and in some cases, at least, associated with this reduction is a retarda- tion in the development of those tracheae that are retained. In what is regarded as the more normal relation between the tracheation and the venation of wings, the tracheas are developed early in the growth of the wing, before the beginning of the development of the wing- veins, which are later formed about them. Here the courses of the tracheae determine to a great extent the courses of the wing-veins. But in those orders where there is a reduction of the tracheation of the wings, the tracheae, in those cases that we have studied, do not penetrate the wing-sac until after the vein cavities are foi-med, and then they follow the most available channels, which in a wing where the venation is highly specialized, as in the Hymenoptera, may be very different from the primitive course of the tracheae. A great reduction of the tracheation of the wings has taken place in the Trichoptera. If a wing of a pupa of a caddice-fiy be examined at that stage when the forming wing-veins appear as pale bands (Fig. 14). it will be seen that the tracheation bears but little relation to the wing-veins. Usually only two or three main tracheae are present; and although these may coincide with forming veins, their branches bear no relation to veins. In the Diptcra an equally great reduction of the tracheation of the wings has taken place in most families; and even when most of the Fig. 14. -A wing of a pupa of a caddice-fly (After C. & N.). THE TRA CREATION OF WINGS 23 tracheae are retained, as in certain asilids, they have not retained their primitive position. A similar condition exists in the Hymenoptera. In the more general- ized Hymenoptera, as in Tretnex (Fig. 15), the main stems of the principal Fig. 15. — The wings of a pupa of Tremex (After C. & N.). tracheag are retained, and occupy very nearly the normal positions; but the courses of the branches of these trachea? bear little if any relation to their primitive courses. In the more specialized families, there is a greater reduction of the tracheation and a wider deviation from the primitive type in the branches of the trachea that are retained. The wings of a young pupa of Apis (Fig. 16) illustrates this. In these wings the vein-cavities are already fonned and the tracheae are beginning to push out into them, follow- ing the most direct courses in the already formed vein-cavities. ^J;- Fig. 16. — The wings of a pupa of Apis (After C. & X.). Eccentric tracheation of wings. — In the Ephemerida there are many modifications of the tracheation of the wings that do not appear to result from any definite course of specialization. In many cases a principal trachea is reduced in length, so that it traverses only a part of the vein with which it corresponds; frequently a trachea follows its vein for a dis- 24 THE TRACHEATION OF WINGS tance and then changes its course suddenly and enters and follows another parallel vein; and there is no regularity in these aberations. In some forms there is a great reduction of the principal tracheae and correlated with this reduction a remarkable increase of fine tracheal twigs; so that the result is an increased tracheation of the wing. Examples of these devia- tions from the typical tracheation are figured in the chapter devoted to the wings of the Ephemerida. Limitations to the value of tracheae in determining the homologies of veins. — Much stress has been laid in the preceding pages upon the value of the tracheae of the wings as an aid to determining the homologies of the wing-veins. In fact the most conclusive proof of the uniformity of the fundamental type of the wing-venation of all orders of insects is drawn from studies of the tracheation of the wings of representatives of those orders in which the tracheation is well preserved. Fortunately in the case of those orders where the tracheation is reduced the venation of the wings of adults so closely resembles in its more general features that of the orders in which the tracheae are well preserved that there is no difficulty in recognizing the identity of the principal veins. There are cases, however, in which the evidence presented by the tracheation of the wings is misleading and can not be accepted. This does not imply that we are to accept the testimony of the tracheae if it suits our purpose and to reject it if it does not; but rather that we are to consider other evidence as well as that presented by the tracheation. Two illustrations will serve to make this point clear. In the suborder Anisoptera of the order Odonata studies of the trachea- tion of the wings have shown that in the adult wing of Gomphus, for exam- ple, the radial sector occupies a position between veins M2 and M:j. Not only does the basal connection of the radial sector trachea show this but the successive stages of the migration of the radial sector trachea from its normal position to this unusual one can be seen in the wings of a series of nymphs of different ages. Figures illustrating this are given in the chapter treating of the Odonata. If one studies the wings of a member of the suborder Zygoptera of this order and compares the venation with that of Gomphus there will be no difficulty in identifying the veins and in recognizing the fact that the radial sector occupies the same position as in the Anisoptera. But when one studies the tracheation of the wing of a nymph of one of the Zygoptera the radial trachea is found to be unbranched and the trachea that precedes what is doubtless the radial sector is a branch of media. It is obvious that in this case the evidence presented by the tracheation is misleading and that the evidence presented by the adult venation is more reliable. The explanation of the discrepancy is that in the Zygo])tera the radial sector THE TRACHEATION OF WINGS 25 trachea has become united with the medial trachea at the point where it crosses the medial trachea and has lost its earlier basal connection. The second illustration is found in the more specialized Lepidoptera where the base of media has been lost by atrophy and the first branch of media, vein Ri, has become attached to the radial sector; correlated with this shifting of the base of vein Mi to the radial sector trachea Ri appears to be a branch of the radial sector trachea. In the more generalized Lepidop- tera trachea Mi has preserved its primitive connection with the other branches of the medial trachea. In the latter case the evidence presented by the tracheation is complete in itself; in the former case there is other evidence to show that the tracheation has been modified secondarily. In general it may be said that the evidence presented by the tracheation of the wings of the more generalized insects is reliable, but in the case of the more specialized insects where extensive modifications of the venation have taken place the evidence presented by the tracheation should be carefully considered before it is accepted. The basal connections of the tracheae of the wings. — It has been shown on an earlier page that the tracheae of the wings are branches of two large tracheae, one of which enters the wing near the humeral angle of the wing, the other, in the region of the base of the anal area. In the Plecoptera and in certain other groups of insects these two large trachese are distinct, but in most insects they become connected by what has been termed the transverse basal trachea. In the diagram representing the hypothetical primiti\-e tracheation of the wings (Fig. 6) the connections of the tracheae of the wings with these two large tracheae are represented; and the position in which the transverse basal trachea is found when it exists is indicated by two dotted lines. The two groups of wing-tracheae arising from these two large tracheae have been designated as the costo-radial group and the cubito-anal group, respectively. And it has been shown that although the medial trachea belongs to the costo-radial group of tracheae in those forms in which there is no connection between the two groups of tracheae, when the two groups are connected by a transverse basal trachea the base of the medial trachea tends to migrate along the transverse basal trachea towards the cubito-anal group of tracheae, which it reaches in the more specialized forms. These facts suggested the question : what are the causes that produce this result ? Regarding this Comstock and Needham ^\'rote as follows: "We have found no indication that the formation of a transverse basal trachea and the subsequent migration along it of the base of the medial trachea is influenced at all by the flight function of the wing, as the arrange- ment of the wing-veins does not appear to be modified by it. It should be remembered that the transverse basal trachea and the bases of the wing 26 THE TRACHEATION OF WINGS trachea are within the thorax of the adult insect, and are thus beyond the influence of the migrations of the wing- veins. "It is probable that these changes have to do with improving the air supply of the wing; but we have not sufficient data, as yet, to warrant a definite statement on this point. The important thing for the purpose of the present discussion is that one must know of this tendency on the part of the medial trachea to migrate along the transverse basal trachea in order to be able to recognize it in its various positions." In taking up the subject again, in the course of the preparation of the present essay, it occurred to me that there are certain modifications of the wing venation in certain insects that may be due to the same cause as that which causes the migration of the base of the medial trachea along the transverse basal trachea. For example, if the large trachea that supplies the cubito-anal group of tracheae with air affords a better supply of air to the wing than does the trachea that supplies the costo-radial group of trachese, this may be the factor that determines in the wings of dragon-flies the invasion of the area of the radial sector by some of the branches of the media. It seemed worth while, therefore, to make a special study of the basal connections of the two large trachese that supply the wing with air. This investigation was undertaken, at my suggestion, by Mr. R. N. Chapman; and the work was done in the entomological laboratory of Cornell Univer- sity, where I had the pleasure of following it step by step. The investigation was a very difficult one, involving the making of a large number of most delicate dissections; but it was prosecuted in a most successful manner. We now know the facts regarding the basal connec- tions of the trachea of the wings in most of the orders of insects. The results of this investigation have not shown that the air supply of the cubito-anal group of trachese is better than that of the costo-radial group. It does not seem probable, therefore, that an improvement of the air supply to the wing is the factor that has determined the migration of the base of the medial trachea, when it occurs, or the invasion of the area of the radial sector by branches of media in the Odonata and Ephemerida. In order that the data now at hand bearing on this problem may be kept together, I publish Mr. Chapman's account of the results of his investiga- tion as an appendix to this chapter. THE TRA CREATION OF WINGS 27 THE BASAL CONNECTIONS OF THE TRACHEA OF THE WINGS OF INSECTS By Royal Norton Chapman, M.A. This paper is the result of an investigation undertaken at the suggestion of Prof. J. H. Comstock, under whose direction the work has been done. The purpose of the work has been to determine the generahzed type of the basal connections of the tracheae of insect wings and to ascertain what the principal lines of modification have been, with the hope that the conditions of the basal connections of the tracheee may offer some explanation for some of the modifications of the tracheae in the wings themselves. It v/as found by Comstock and Needham {American Naturalist, xxxii, 1898, p. 88-89) that the tracheae that supply the wings with air arise from two distinct trunks, an anterior costo-radial trunk and a posterior cubito- anal trunl<. In certain generalized insects these two trunks are distinct; where this condition exists the medial trunk is a member of the costo-radial group of tracheae. In most insects there has been developed a transverse trachea connecting these two groups of tracheae, the transverse basal trachea (Fig. 1 7, tb).* When a transverse basal trachea is formed, the base of the medial trachea tends to migrate along it towards the cubito-anal group of trachccB, and often becomes united with that group. Comstock and Needham were unable to explain the cause of the forma- tion of a transverse basal trachea and the migration along it of the base of *Lettering of the Figures Iillistratixg the Basal Connections of the Trache.e of the Wings of Insects. a. c-r, accessory costo-radial trachea; a. cii-a, accessory cubito-anal trachea; a. dt, accessory longitudinal trachea; as, anterior stem of the leg trachea; c, costal trachea; c-r, costo-radial trachea; cu-a, cubito-anal trachea; d. It, dorsal longitudinal trachea; g, tracheal gill; at, gill trachea; I2, mesothoracic leg trachea ; /a, metathoraeic leg trachea; //, longitudinal thoracic trachea; mt, muscle trachea; p$, posterior stem of the leg trachea; r. spi, rudiment of the mesothoracic spiracle; r. sp'z, rudiment of the m.'tathoracic spiracle; r. 5/J3, rudiment of the first abdominal s])iracle; sp\, mesothoracic spiracle; sp-z, metathoraeic S]nracle; 5p3, first abdominal si)iracle; spi, second abdominal spiracle; tv, transverse basal trachea; u, union of the costo-radial and transver.se basal traclieae; vlt, ventral longitudinal trachea; V. sp2, vestige of the metathoraeic spiracular trunk; V. sp3, vestige of the first abdominal spiracular tnmk. 28 THE TRACHEATION OF WINGS the medial trachea when it is formed. Concerning this they made the following statement: "It is probable that these changes have to do with improving the air supply of the wing; but we have not sufficient data, as yet, to warrant a definite statement on this point." A desire on the part of Professor Comstock to have ascertained the exact nature of the basal connections of the trachea of the wings in different orders of insects, in the hope that this knowledge would throw light on this problem, led him to suggest the making of the investigation the results of which are given here. Another condition that suggested the same question is the fact that in the Odonata the branches of the medial trachea invade the region of the tracheae of the radial sector, which becomes greatly reduced. If the medial trachea has a better air supply than has the radial trachea, this invasion can be understood. The references to the basal connections of the wing tracheae which have been found in literature are few and fragmentary. So far as can be deter- mined no general study of the subject has been made previously. How- ever, various authors, while working on closely related subjects, have noted and described the conditions in certain forms. These descriptions may be best referred to in the parts of the paper relating to these various forms. However the work of Karel Sulc [Uber Respiration, Tracheensystem und Schaumproduction der Schaumcikadenlarven (Aphrophorinae- Homoptera) Zeitschrift fiir wissenschafliche Zoologie. 1911:99, p. 147-188] deserves special mention. This author described and figured the tracheal system of Philaenns lineatus L. with detailed drawings of the leg and wing tracheae in the various nymphal stages. The basal connections of the wing trachete are more generalized than those of the Hemiptera described and figured in this paper (Figs. 22, 23, and 24) and agree with the typical condition of the wing tracheae (Figs. 1 7a and 1 7b) except that the transverse basal trachea has been developed. The basal connections of the fore wing pad of a first stage nymph of Aphrophora salicis are described and figured in which the transverse basal trachea is not developed and the costo-radial and cubito-anal groups of tracheae are shown arising respectively from the anterior and posterior stems of the leg tracheae. Careful dissection was found to be the best means of studying the tracheae. The use of transmitted light to distinguish the opaque tracheae in the more or less transparent bodies, without dissection, was not satisfac- tory because it was not possible, in this way, to accurately determine the relationships of the tracheae which lie one above the other. Nymphs and pupae were dissected under water and care was taken not to break large trachejc and allow the air to escape. In this way the trachere were kept filled with air and were easilv studied. Several dissections were made of THE TRACHEATION OF WINGS 29 each species and the results were carefully checked over to make sure that no tracheae had been broken or overlooked. The dissections were drawn with a camera-lucida before the air had escaped from the tracheae and while the trunks retained their normal size. It is to be noted that no pretense has been made to include all of the muscle tracheae; only the larger muscle tracheae connected with the wing or leg tracheae have been shown. rt'.//-^ THE TYPICAL ARRANGEMENT OF THE BASAL CONNECTIONS OF THE TRACHEA OF THE WINGS Two diagrams of what seems to be the typical condition of the basal connections of the wing tracheae have been con- structed (Fig. 17). The condition of the tracheal branches to the wings and legs, as figured in these diagrams, does not differ in any of its essentials from the condition found in the more generalized insects. It will also be noticed that none of the insects studied present conditions which cannot be looked upon as modifications of this typi- cal condition. The con- struction of these diagrams is, therefore, necessary only to simplify the discussion of the modifications of the typical condition and not to represent an imagin- ary step in the develop- ment of any of the condi- tions found in any of the more specialized insects studied. One of the diagrams (A) is a side view of the thoracic tracheae in which the wing tracheae are represented as extending vertically above the thorax and those of the legs extending downward. The other diagram (B) is|a Fig. 17. — The typical condition of the basal con- nections of the tracheae of the wings of insects. A. Side view. B. Dorsal view. 30 THE TRACHEATION OF WINGS dorsal view of the thoracic tracheae and represents the tracheae of both the wings and legs extending laterad. It must be kept in mind that when the thorax is viewed from the dorsal side the wings are directly above the legs (Fig. 17, B). Since the conformation of some of the insects has made it necessary to draw the dissections from the side view and others from the dorsal view, the two diagrams are given to aid in the comparison of these two different aspects of the thorax. There are three spiracles directly connected with the respiration of the wings; one in each of the last two segments of the thorax and one in the first segment of the abdomen (5^1, spo, and sps). The anterior and posterior of these two spiracles (5^1 and 5^3) each contributes one tracheal branch to the appendages of the mesothorax and the metathorax respectively. The middle spiracle {sp^} contributes two branches; one to the mesothoracic appendages and one to the metathoracic appendages. In this way each wing and each leg receives two tracheal branches; one from the spiracle anterior to it {c-r to the wing, and as to the leg) and one from the spiracle posterior to it {cu-a to the wing, and ps to the leg.) This condition is retained, in its essentials, in all the insects studied. Even tho the meta- thoracic spiracle is absent in many of the more specialized insects, its position is indicated by the rudimentary spiracular tnmk (Fig. 21, r. spi), from which arise branches to the thoracic appendages, very much as in the more generalized insects, where the metathoracic spiracle is well developed. As has already been stated, these branches that arise from the trunks at the bases of the spiracles supply air to both the wings and the legs. The leg tracheae are Y-shaped; each leg trachea being formed by the union of two trunks. These trunlcs are designated as the anterior stem {as) and the posterior stem (ps) respectively. In the more generalized insects the branches to the wings arise from the stems of the leg tracheae, which form the arms of the Y. The figure of the nymphal cockroach (Fig. 18) shows that the tracheas going to the legs are much larger than those leading to the wings. The relations of the leg and wing tracheae are such that it seems very evident that the branches coming from the spiracles are primarily leg tracheae and that the branches to the wings are secondary. In some cases the wing and leg tracheae arise separately from the spiracles. Such a condition may be explained as the result of a divergence of the bases of the two tracheae which has been carried so far that their connections are entirely separate. However the more general relations between the tracheae to the wings and those to the legs have been retained, at least to some extent, in all of the insects studied. For this reason it would seem that a consideration of the tracheae to the wings cannot ignore the condition of the tracheae to the legs. If there is any preference as to the air supply from the different spiracles which results in a modification of the tracheae to the wings, some modification of the tracheae to the legs might THE TRACHEATION OF WINGS 31 also be expected. If it is found that any factor causes changes in the tracheal branches to the wings, it may be expected that a similar factor affecting the legs will cause similar changes in the trachea entering the legs and the converse should also be true. EXAMPLES OF THE MORE GENERALIZED ARRANGEMENT The basal connections of the wing trachecz in a cockroach nymph (Phylo- dromia germanica) , Order Orthoptera. — The condition of the tracheae in the cockroach is very similar to that shown in the diagrams. The connections are as shown in the drawing (Fig. 1 8) which is a dorsal view. The cubito- anal tracheee (cu-a) originate nearer the spiracles than the costo-radial tracheae (c-r) do, but they have the typical connections with the stems of Fig. 1 8. — The basal connections of the wing tracheas of a cockroach. Dorsal view. the leg tracheas {as and ps) which form the two arms of the Y-shaped leg tracheas. There are muscle tracheae (nit) which go to the coxal muscles but they do not continue into the legs as do the tracheae labeled k and k. In the nymphal cockroach there is a single slender dorsal longitudinal trachea (d. It) , on each side of the thorax, lying near the digestive tract and connecting the branches from each of the three spiracles (5^1, spo, and spa). The size and relationships of this trachea are such as to indicate that it may be only an enlarged anastomosis of the branches from the different spiracles. In adult cockroaches the intra-segmental portions of this longitudinal trachea become enlarged to form air sacs. The former statement that the wing tracheae of cockroaches are con- nected with the longitudinal tracheae of the thorax has not been confirmed. (Comstock and Needham, Am. Nat. xxxii, p. 88). As the drawing shows, the basal connections of the wing tracheas and the dorsal longitudinal tracheae are quite separate. 32 THE TRACHEATION OF WINGS The basal connections of the wing trachei Fig. 19. — The basal connections of the wing tracheae of Pteronarcys. Dorsal view. but the relationships are quite generalized. The typical Y-shaped leg tracheee are present in both cases, but in the case of both legs the posterior stem (ps) is connected with a trunk leading from the spiracle to the tracheal gill (gt) . The cubito-anal wing tracheee (cu-a) have diverged from the leg trachege until they have attained independent origins. The conditions of the anterior branches to the legs and wings have not been modified and the wing trachea (c-r) arises in common with the anterior stem of the leg trachea (as). The dorsal longitudinal trachea (d. It) is well developed but it will be noticed that it has no direct relationship to the wing tracheae. The basal connections of the wing trachecu of Chauliodes, Order Neur- optera. — The wing tra- cheae of Chauliodes (Fig. 20) represent no great deviation from the typical condition. It is interesting to note that the cubito-anal trachea (cu-a) and the posterior stem of the leg trachea (^5) are separated at their bases in both cases and that the costo- radial tracheae (c-r) and the anterior stems to the legs (as) are connected Fig. 20. — The basal connections of the wing tracheas of Chauliodes. Dorsal view. THE TRACHEATION OF WINGS 33 for only a short distance. It would seem that the tracheae to the wings are almost at the point of being separated from those of the legs, in this form. The longitudinal trachea {d. It) is well developed yet it does not seem to be in any wa}- connected with the wing tracheas. The basal connections of the wing trachecB of Antheraa roylei, Order Lepidop- tera. — Three pupse of the Lepidoptera were dissected, Aclias luna, Samia cecropcea, and Anthercea roylei. Any of the three might have been used equally well for the conditions of the tracheae are the same in all of them. In the micsothorax (Fig. 21) there seems to have been no modification from the typical condition, but in the metathorax the cubito-anal trachea (cu-a) Fig. 21. — The basal connections of the wing trachese of Anthercea. Side view. and the posterior stem of the leg trachea {ps) are just separate from each other at their bases. The dorsal longitudinal trachea id. It) is present but has no connection with the wing tracheae. The metathoracic spiracle is absent but is repre- sented by the trunk of the rudimentary spiracle (r. sp-i) with which the tracheae have retained their typical relationships. The thoracic tracheas of Anthercea perni have been described and figured by Enderlein {Zoologische Jahrhucher, 1902; vi.). This author figures trachese to the mesothoracic leg as a single stem arising from the spiracular trunk of the mesothorax and the leg trachea of the metathorax as also single and arising from the costo-radial trachea of the mesothorax. The conditions of the tracheae seem to be so constant in representatives of this order which have been studied that it would seem very improbable that there should be such a great difference between two members of the same genus as Enderlein's figure would indicate. 34 THE TRACHEATION OF WINGS The basal connections of the wing trachece of a notonectid nymph, Order Hemiptera. — The costo-radial tracheae (Fig. 22 c-r) arise from the anterior Fig. 22. — The basal connections of the wing tracheae of a notonectid nymph. Side view. stem of the leg tracheae {as) . The bases of the cubito-anal tracheae {cu-a) have diverged from the posterior stems of the leg tracheae (ps.) and have come to arise in common with muscle tracheae {mt) which pass to the median part of the body. The condi- tions of the wing tracheae are, however, so near the typical that there is no difficulty in vmder- standing the relationships. The leg tracheae are modified to a greater extent. In the meso-thorax the anterior and posterior stems to the leg (as and ps) unite to form the typical Y-shaped leg trachea and at their point of union give off a large muscle trachea (nit.). The posterior stem of the leg trachea in the metathorax (ps,) lies laterad of the anterior stem (as) at their point of union and the muscle trachea (mt) appears to be a continuation of the posterior stem of the leg trachea (ps). Fig. 23. — The basal connections of the wing tracheae of a corisid nymph. Side view. THE TRA CREATION OF WINGS 35 The basal connections of the wing trachece of a corisid nymph, Order Hemiptera. — The two tracheal stems to the wings (Fig. 23, c-r and cu-a) have retained their typical relationships in this form, but the cubito-anal trachea {cii-a) to the hind wing is so small at its base that it is followed with difficulty. The anterior and posterior stems of the leg tracheae {as and ps) have attained an almost horizontal position, giving the leg tracheae a T-shape. However, the relationships between the wing and leg tracheae have remained typical. TJie basal connections of the wing trachea of a Lcihoceriis nymph, Order Hemiptera. — The tracheae leading to the wings are, of themselves, rather typical save for the divergence of the costal trachea (Fig. 24, c) from the Fig. 24. — The basal connections of the wing tracheae of Lethocertis. Side view. costo-radial trachea {c-r) of the front wing. The relationships of the leg and wing tracheae at first seem quite puzzling but if conditions are com- pared with those found in notonectids and corisids the explanation seems rather simple. It is quite evident that a "T" shaped trachea similar to that found in the corisid nymph has been modified by the posterior migra- tion of the main trunk to the leg which has resulted in the lengthening of the anterior stem {as) and shortening of the posterior stem {ps) which results in the practical obliteration of the latter. The horizontal trunk leading from the mesothoracic spiracle {spi) to the metathoracic spiracle {sp-i) is, therefore, the anterior stem of the leg trachea {as) and the posterior stem of the leg trachea has disappeared. The costo- radial trachea {c-r) has its typical origin with the anterior stem of the leg trachea {as) altho it diverges from the latter some distance posterior to the tj^pical position, and the costal trachea {c) has an independent origin. The 36 THE TRACHEATION OF WINGS condition of the tracheas of the metathorax is similar to that of the meso- thorax except that the costal trachea (c) has retained its typical position and the cvibito-anal trachea {cu-a) has moved slightly forward and might almost be said to arise from the anterior stem of the leg trachea {as) in the absence of the posterior stem from which it typically arises. The basal connections of the wing trachea of Monohanimus, Order Coleop- tera. — The wing tracheae of the Coleoptera have retained their original relationships to the spiracular trunks but there is a modification of the leg trachccc. Both of the posterior stems of the leg tracheag (Fig. 25, ps.) are split nearly to their basal connections and they are entirely independent of Fig. 25. — The basal connections of the tracheae of the wings of Monohamus. Side view. the cubito-anal tracheae {cu-a). The anterior stem of the leg tracheae {as.) of the metathorax is very slender but it has retained its original connection with the costo-radial trachea {c-r). In the mesothorax the anterior stem of the leg trachea is entirely absent. The longitudinal trachea is well developed and the position of the meta- thoracic spiracle, which is absent, is indicated by the spiracular trunk {v. sp2). None of these conditions seem to have modified the original relationship of the wing tracheae to the spiracular tnmks. The basal connections of the wing trachece of Bittaconiorpha, Order Diptera. — The connections in this form are rather generalized and even tho the hind wing is absent the halter receives tracheae which have all the relationships that the tracheae to the wing would have (Fig. 26). There is a tendency THE TRACHEATION OF WINGS 37 toward the reduction of the cubito-anal tracheae which are ver>' small and are followed with difficulty in dissection. The tracheae to the legs are unmodified except for the shortening of the anterior stem {as) to the hind leg. EXAMPLES OF SPECIALIZATION BY REDUCTION The basal connections of the wing trachecB of Ryacophila, Order Trichop- tera. — The wing trachea) of Ryacophila are greatly reduced and the basal connections are correspondingly specialized. The thorax is specialized in that there are marked intersegmental depressions on the sides, within which the rudimentary spiracles are found. However, the connections to the legs have retained their primitive condition. Fig. 26. — The basal connections of the trachea; of the wings of Bittacomorpha. Side view. The wing tracheae of each wing appear to arise as two trunks which connect indirectly with the mesothoracic spiracular trunk (Fig. 27, s/?i) in the case of the front wing, and the metathoracic spiracular trunk {sp2) in the case of the hind wing. The anterior trunk to each wing arises in common with the anterior stem of the leg trachea and is undoubtedly the costo-radial trachea (c-r). The cubito-anal trachea of each wing {cu-a) arises in its t\T)ical position near the spiracle which nonnally supplies it with air. But owing to a cephalization of its air supply which results in this trachea receiving the greater part of its air by the way of the transverse basal trachea {tb) its course becomes greatly modified and the proximal portion of it greatly reduced. This results in that part of it beyond the point of union with the transverse basal trachea appearing to be a continuation of that trachea. There has been developed a small accessory longitudinal trachea {a-dt) which arises with the transverse basal trachea near the spiracular trunk 38 THE TRACHEATION OF WINGS and extends parallel to the dorsal longitudinal trachea {d.li). Its only connection with the wing trachea is at its point of origin. Fig. 27. — The basal connections of the tracheae of the wings of Ryacophila. vSide view. The basal connections of the wing trachecB of an unknown genus (Limno- philidcB), Order Trichoptera. — This form illustrates an interesting step in the reduction of the tracheae to the wings. (Fig. 28) . In the hind wing both Fig. 28. — The basal connections of the tracheae of the wings of unknown genus of Trichoptera. Side view. the costo-radial (c-r) and cubito-anal tracheae (cu-a) are present altho the latter is very small as is also the transverse-basal trachea {tb). In the front wing the cubito-anal trachea and the transverse basal trachea are THE TRA CREATION OF WINGS 39 absent. However the reduced cubito-anal trachea (cu-a) is present in the thorax with its typical connections except that it fails to reach the wing. The leg tracheae have the typical connections. The basal connections of the wing trachece of Stenophylax, Order Trichop- iera.— The cubito-anal tracheae (Fig. 29) do not reach the wing in either case in this form altho their vestiges are present in the thorax (cu-a) and both have their typical connections with the posterior stems of the leg tracheae (ps) . In both cases the costo-radial tracheae {c-r) have their typical connections except that the transverse basal tracheae are entirely absent. The basal connections of the wing trachece of a nymph of Epeorus. Order Ephemerida. — The wing tracheas of Ma\^ies are highly specialized and Fig. 29. — Tlie basal connections of the trachcse of the wings of Stenophylax. Side view. there is a marked cephalization of the flight function in this order. Morgan in a paper on the wings of May-flies (Morgan, Anna H. Homologies of the wing veins of the May-flies, Annals, Ent. Soc. Amer. 191 2, v. 5, p. 89-105), concluded that the wing tracheae of Epeorus represented one of the more generalized conditions. For this reason Epeorus has been used for the study of the basal connections of these tracheae. The anterior tracheae to each leg and wing have retained their primitive conditions (Fig. 30); the anterior stem of the leg tracheae {as) and the costo-radial trachea (c-r) to the wing have a common origin in each case. However the posterior connections to the wings and legs have been greatly reduced. The portion of the cubito-anal trachea (cu-a) proximal to its union with the transverse basal trachea (tb) is absent except for a short vestige in the front wing. This short vestige extends proximally from the 40 THE TRACHEATION OF WINGS 10 ^ Fig. 30. — The basal connections of the tracheae of the wings of Epeorus. Side view. union of the transverse basal trachea {th) and the cubito-anal trachea {cii-a) and connects with a small accessory trachea which leads anteriorly to the costo-radial trachea (c-r) . The remainder of the cubito-anal trachea which Morgan had suspected to be present, was not found in any of the specimens dissected. Even this small vestige was found to be absent in the hind wing. The posterior stems of the leg trachese {ps) are repre- sented by slender tracheae which enter the coxee in the typical position for the pos- terior stems of the leg tracheas but their connections with the anterior stems {as) are very slight and seem to be lacking in some cases. It is very evident that the posterior tracheae to the legs are at the point of being lost in this form. There is a saucer shaped disc, which has been indicated by a dotted line in the figure, between the cubito-anal and costo-radial groups of tracheae in each wing. There is also a well developed apodeme, represented by a dotted line, just posterior to the union of the cubito-anal trachea {cu-a) and the transverse basal trachea {th) in each wing. The influence of these structures upon the tracheae of the wings will be taken up in the latter part of this paper. The basal connections of the wing trachece of Chirotonetes nymph, Order Ephemerida. — The conditions of the basal connections of the wing tracheae of Chirotonetes are more easily understood after having studied the conditions in Epeorus. In Chirotonetes (Fig. 31) the anterior tracheal connections to the legs and wings have retained their primitive condition but the posterior connections are entirely absent. Small muscle tracheae, somewhat resembling the typical posterior connections, were found but their former connections have been entirely lost and the Fig. 31.— The basal connections of 1 r i-- r ^1 • 1 ^ i_i • the tracheae of the wings of cephahzation of the air supply to the wmgs Chirotonetes. Side view. is complete. The basal connections of the wing trachece of Apis nielli fica, Order Hymen- optera. — The pupa of a drone honey bee was used for the study of the trachea rilE TRACHEATION OF WINGS 41 of this form because of its large size. The highly specialized bee has specialized tracheal conditions but a study of the figure will recall the conditions in the more generalized insects. The Y-shape of the leg tracheae is as striking in the metathorax of the honey-bee as in the diagram of the typical condition (Fig. 32). In the mesothorax the anterior stem to the leg (as) is greatly elongated and extends dorsally to its connection with the costo-radial {c-r) branch to the wing and then passes ventrally into the leg. The posterior stem (ps.) to the mesothoracic leg {I.2) is united with the anterior stem (as) to the metathoracic leg a short distance from the spira- cular trunk. The spiractilar trunk leading to the vestige of the meta- thoracic spiracle (v. sp-i) is very long. The tracheae leading to the muscles of the coxae {mi) are well developed, and as shown in the figure, are united much like the leg tracheae with which they actually connect. But these muscle tracheae lie deeyj within the thorax and are relatively much smaller than the true leg trachese. The branches to the wings exhibit the greatest specialization. In the case of both wings the air supply comes entirely from the spiracle anterior to the wing, for there is no connection with the spiracle posterior to the wing. A short distance distad from the union of the two branches to the front wing there will be noticed a slight projection from the tracheae. This projection is in such a position that it is highly suggestive of a vestige of the former cubito-anal trachea which extended from this point to the base of the metathoracic spiracular trunk. In a similar position on the trachea leading to the hind wing there are a few tracheal branches which may have a significance similar to that which has been ascribed to the projection on the trachea leading to the front wing. If these structures represent the vestiges of the cubito-anal tracheae leading to the spiracular trunks, the part of the tracheae between these structures and the union of the tracheae (m) must represent the transverse basal trachea and the remainder of the posterior branch must represent the distal part of the cubito-anal trachea (cu-a). Fig. 32. — The basal connections of the tracheae of the wings of Apis. Side view. 42 THE TRACHEATION OF WINGS EXAMPLES OF SPECIALIZATION BY ADDITION The basal connections of the wing trachece of a nymph of Melanoplns sp., Order Orthoptera. — The dissection of an acridid nymph is rather difficult because of the many tracheal branches and air sacs in the thoracic region. The minor branches and the air sacs were removed in the course of the dissection and the figure is a camera-lucida drawing of the tracheas with which this paper is concerned together with the more important minor branches. The tracheal connections are modified to a certain extent but the condi- tions are easily explained (Fig. 33). The point of union of the costo-radial Fig- 33- — The basal connections of the trachefe of the wings of Melanoplus. Side view. trachea (c-r) and the anterior stem of the leg trachea (as) has moved dorsad in both cases. In the mesothorax the anterior stem to the leg arises from the spiracular tnuik (5^1) and goes dorsad to the point where the costo- radial trunk (c-r) is given off and then it passes ventrad to meet the posterior stem (^.s-) thus forming the usual Y-shaped trachea to the leg. It will be noticed that the posterior stem is the larger of the two. From the costo-radial trachea (c-r) of the front wing there is an acces- sory trachea (a. c-r) which connects with the saclike dilated dorsal longitu- dinal trachea ((i./O- The base of the cubito-anal trachea (cu-a) has fused with the anterior stem (as) to the metathoracic leg to fonn a common trunk extending dorsad for a short distance from the metathoracic spiracle. The cubito-anal trunk {cti-a) to the front wing has an accessory tracheal THE TKACHEATION OF WLXGS 43 connection (a. c-a) which passes mesad, just dorsad of the dorsal longi- tudinal trachea (d. It), to anastomose with its fellow of the opposite side. The changes in the costo-radial trachea to the hind wing have been similar to those to the front wing. The anterior tracheal stem to the leg (as) forms a common tmnk with the cubito-anal trachea of the front wing, as has just been described, and then it continues dorsad, in common with the costo-radial trunk (c-r) of the hind wing, until it becomes independent of the wing trachea and passes ventrad to unite with the posterior stem of the leg trachea (ps). The costo-radial trachea of the hind wing also has an accessory connec- tion (a. c-r) with the dilated dorsal longitudinal trachea {d. It) . The cubito- anal connection (cu-a) is unmodified except for a small accessory connection between it and the anterior stem of the leg trachea (as) , similar to a connection found in the mesothorax. The posterior stem (ps) of the trachea to the metathoracic leg (h) is small and has new relationships at its point of entrance into the specialized hind leg. The descriptions of the leg trachea in the metathorax, which have given rise to these new relationships involving the second abdominal spiracle, belong more properly to the last part of this paper and their dis- cussion will be postponed for the present. The basal connections of the wing trachece of Anax and Lestes, Order Odonota. — The tracheation of Plathemis lydia was described and figured by G. G. Scott {Biological Bulletin 1905, ix : 341-354). The author described the tracheae of the wing as arising from a loop, the anterior trachea of the loop arising from the dorsal longitudinal trachea and the posterior trachea, in the case of the mesothorax, arising from the short transverse connective in such a way that the two posterior tracheae of the fore wings originate side by side. The tracheation of the hind wings is described as similar to that of the fore wings except that the anterior trachea arises from the spiracular trunk near the dorsal longitudinal trachea and the posterior trachea is described as originating directly from the dorsal longitudinal trachea. R.J. Tillyard {Proceedings of the Linnean Society of New South Wales, 1914, xxxix, pt. I, p. 163-216, pi. xi-xiii) made a study of Odonata n^nnphs using transmitted light to distinguish the tracheae. His description is said to embody the results obtained from the study of both the Anisoptera and the Zygoptera and a diagram of the lateral view of the thorax is given to show the connections of the wing tracheae. This author describes one tracheal branch to each leg and one trachea, arising from the dorsal longi- tudinal trachea, as entering each wing pad from the anal end, forming a loop from which the wing tracheae arise, and passing out of the wing pad at the costal end as a fine threadlike trachea which connects with the leg trachea of the same segment. 44 THE TRACHEATION OF WINGS Tillyard's description and figure differ from those of Scott in that the former author has the anterior wing tracheae connecting with the leg tracheae and the posterior wing tracheae connecting with the longitudinal trachea, while the latter author has the anterior trachea of the fore wing connecting with the dor- sal longitudinal trachea and that of the hind wing connecting with the spiracular trunk near the dor- sal longitudinal trachea. The posterior trachea to the wings is said by him to connect with a transverse connective, in the case of the fore wing, and with Fig. 34-— The basal connections of the tracheae the dorsal longitudinal trachea in of the wings of Anax. Dorsal view. ,1 r ^.i 1 • j • the case 01 the hind wmg. The figures accompanying this paper are the result of numerous dissec- tions and show the conditions in a representative of each of the two sub- orders (Figs. 34 and 35). There are two tracheae entering the front of each wing pad, the usual costo-radial trunk (c-r) connecting with the anterior stem of the leg trachea (as) and accessory costo-radial trunk (a. c-r) connecting with the dorsal longitudinal trachea (d. It), in the case of the front wing, and with the trunk to the vestigial metathoracic spiracle, in the case of the hind wing. It is probable that Tillyard has seen the costo-radial trachea in studying the lateral view of the thorax while Scott has seen the accessory costo- radial trachea in his study of the dorsal view of the thorax, which he figures. There are likewise two posterior tracheae to each wing pad. The cubito-anal tracheae (cn-a) of both wings are vestigial and are made out with some difficulty but they have retained their original connections with the posterior stems of the leg Fig- 35- — The basal connections of the tracheae of the wings of Lestes. Side view. tracheae. The accessory cubito-anal tracheae (a. cu-a) are much the larger; the one to the fore wing arises from the transverse connective which lies THE TRA CREATION OF WINGS 45 between the two dorsal lon<2;ittidinal trachea) {d. It) as described by Scott (/. c.) and the one to the hind wing arises from the dorsal longitudinal trachea. It will be noticed that in the side view of Lestes (Fig. 35) the connection of the accessory cubito-anal trachea (a. cu-a) of the mesothorax with the transverse connective cannot be seen for it is hidden by the dorsal longitudinal trachea {d. It). The vestige of the true cubito- anal trachea (ai-a) has doubtless been overlooked by both of the former authors. THE INFLUENCE OF THORACIC STRUCTURE ON THE CONNECTIONS OF THE WING TRACHEAE The main tracheae of insects lie in the spaces of the body allowing for the free passage of air thru their lumina. Their flexibility is a necessity for tracheae entering the appendages of the body must bend readily when these structures are moved and it is all important that the flexible trachea should avoid contact with any structure which might at any time bring pressure to bear upon their walls and interfere with the free passage of air. The course of the tracheae to the wings might be expected to be direct but at the same time the avoidance of rigid structures, such as those concerned with the movement and articulation of the wings, among which the tracheae must pass, would seem to be an important factor in determining their course. The material presented in this paper affords many examples of altered courses of tracheae which seem to be due to peculiarities of thoracic struc- ture. In the wing of the Mayfly, Epeorus, (Fig. 30) a simple example may be found. There is a saucer-shaped disc, which has already been referred to as lying between the medial, cubital, and transverse basal tracheae, and which has been indicated in the drawing by a dotted line, around which the tracheae pass. This disc with its two chitinous walls of such convexity would undoubtedly be a great obstacle to the passage of air if the tracheae crossed it rather than encircled it as they do, altho the latter course of the tracheae is the longer. The case of the stone fly, Pteronarcys (Fig. 19), presents an example of the modification of the medial trachea (w) which is very suggestive. The medial trachea (w) will be noticed to arch caudad toward the cubito-anal group of tracheae and altho there is no transverse basal trachea the medial trachea might, from its position, be considered to be midway between the costo-radial and cubito-anal groups of tracheae. The examination of a n>Tnph which is nearly ready to emerge will show a strong process, evi- dently the wing process, (indicated by a dotted line in the drawing) pro- jecting between the radial (r) and medial (m) tracheae to articulate with an over-lying articular sclerite, evidently the second axillary sclerite. Here again it seems very clear that the course of a trachea is modified from the most direct route by a mechanical obstruction. Moreover this case is very 46 THE TRACHEATION OF WINGS significant for the cause of the migration of the medial trachea along the transverse basal trachea, as stated in the introduction, was one of the problems which prompted this research. Modifications of the thorax which would result in the posterior migra- tion of the wing process would also result in the posterior migration of the base of the medial trachea. Likewise a cephalization of the wing would bring the cubito-anal group of tracheae nearer the medial trachea, if the wing process remained in its original position and did not allow the medial trachea to move cephalad with the rest of the structures of the wing. The consideration of the relationships of thoracic peculiarities and the modifications of the tracheae to the wings may profitably be extended to the more general conditions. The insects considered in this paper have been divided into three groups : those with generalized connections of the wing tracheae, those with reduced connections of the wing tracheag, and those in which the connections of the wing tracheae have been specialized by the addition of new connections. In the group where the conditions of the tracheal connections are generalized the thoracic structure is also general- ized. The examples of the Blattidas among the Orthoptera, the Plecoptera, the Neuroptera, the Lepidoptera, and the others are quite generalized, while the Coleoptera and Hemiptera are mentioned both for their somewhat generalized wing connections and for the reduced connections to the legs, which latter condition is accompanied by the oblique position of the lateral sclerites of the thorax. Among the insects with reduced connections of the wing tracheae there are many varieties of thoracic structure illustrating several lines of speciali- zation. It is not the purpose of this discussion to enter into the details of so complex a region as the thorax or to attempt to ascribe all the modifica- tions of the connections of the wing tracheae to the peculiarities of thoracic structure alone, but rather to call attention to the fact that reduced wing tracheae occur in the insects that have a specialized thorax. In the draw- ings of the Ephemerida (Figs. 30 and 31) and the Hymenoptera (Fig. 32). apodemes are represented by dotted lines just posterior to each wing. These apodemes may be studied in insects which are about to emerge from the pupal state and from their position it seems very probable that they obstruct the course of the cubito-anal tracheae and eventually lead to their reduction and the consequent cephalization of the air supply to the wings. The last group of insects, those with additional tracheal connections to the wings, includes the Odonata (Figs. 34 and 35) and the jumping Orthop- tera (Fig. 33). These two groups of insects are peculiar in that their nymphs have their wing pads flexed dorsally, approximating the mid-dorsal line. Such a position of the wingpads favors the enlargement of any median tracheal connections such as the accessory connections to the wings (a. c-r) and (a. cu-a) because the wingpads are drawn so close to the dorsal longi- THE TRACHEATION OF WINGS 47 tudinal trachcce. On the other hand, the original tracheae {cii-a and c-r) must be greatly elongated and no doubt are obstructed by apodemes along their new path, especially in the case of Odonata. The accessory cubito-anal trachea (a. cu-a) of the Odonata is much larger than the reduced cubito-anal trachea {cu-a) and is also larger than either of the anterior tracheal connections {c-r and a. c-r) , each of which is about half its size. The fact that the two anterior tracheae have had their courses altered by the change in form isprobably the cause of their reduction. Some of the most interesting correlations between thoracic structure and tracheal conditions are to be found in the study of the leg tracheae and the analogy with the conditions of the wings is most instructive in the consideration of their tracheae. In the Coleoptera (Fig. 25) where the thorax is oblique and the legs have moved caudad, the anterior stem {as) to the mesothoracic leg (/o) is absent and the anterior stem to the meta- thoracic leg {h) is small while both of the posterior stems are large. The Hemiptera (Figs. 22, 23, and 24), which have a thorax like that of the Coleoptera, also have modified conditions of the leg tracheas which are most marked in the Belostomidae (Fig. 24) where there is a single stem of the leg trachea, due to the posterior migration of the main leg trachea which results in the obliteration of the posterior stem and the elongation of the anterior stem of the trachea. The Odonata present a thorax which is oblique in the opposite direction from the examples mentioned above, for the legs have moved cephalad. This skewness of the thorax in the Odonata was studied by Needham and Anthony {Journal of the New York Entomological Society, 1903, xi: 117- 125) and the angle formed between the perpendicular and the humeral angle was found to vary from 21 degrees in some of the Aeschninas to 71 degrees in certain Agrioninae. This furnishes a series of modifications in which the tracheae may be studied and it has been found by a study of a series of n^onphs that the proportionate size of the anterior stem of the leg trachea varies in direct proportion to the size of the angle of the thorax, while the size of the posterior stem of the leg trachea varies in inverse proportion to the size of the thoracic angle. The thoracic angle of Anax Junius (Fig. 34) is 27 degrees and the posterior stem of the leg trachea {ps) is slightly larger than the anterior stem {as) . Lestes (Fig. 35) has a thoracic angle of 66 degrees and the posterior stem of the leg tracheae {ps) is very small while the anterior stem {as) is comparatively large. In this connection it is interesting to note that Tillyard (1. c. p. 205) entirely overlooked the posterior stem of the leg trachea in his figure of Ausirolestes. The Odonata, therefore, with their highly modified thorax and greatly altered tracheas to both the wings and legs, present the best example of the correlation between thoracic structure and the tracheal connections. 48 THE TRACHEATION OF WINGS The jumping Orthoptera (Fig. 33) have the tracheation of the hind leg greatly modified in that a part of the tracheae come from the second abdominal spiracle. This is a condition which has been met with nowhere else and is undoubtedly due to the great development of the hind leg and its intrtision into the abdominal region. THE COMPARATIVE VALUES OF THE AIR SUPPLIES TO THE TWO PRINCIPAL TRACHEAE OF THE WINGS The value of the air supply of the wing tracheae has been suspected of being the cause of some of the tracheal modifications of the wings. This theorv^ has been mentioned in the introduction of this paper and it has been given some study by Tillyard (1. c. p. 207-218). This author has made a study of the Odonata where the basal connections of the tracheae are most highly modified. The author has formulated a theory, as the result of his work, which reads as follows: "The peculiarities shown by the wing-venation of the Odonata, as con- trasted with that of other insects, are due primarily to the aquatic habits of the larvae; whereby, thru the development of rectal or caudal breathing, the oxygen-supply of the developing wing is carried from the posterior end of the body, and enters the wing-base at its anal end." The invasion of the area of the radial sector by the medial trachea is accounted for by the fact that the medial trachea is nearer the air supply than the radial trachea. Several other modifications are ascribed to the same cause. The Odonata would seem a very unfavorable order to use in the study of the comparative values of the air supplies to the wings because the conditions are so highly specialized. A better understanding of the subject may be had by returning to the typical condition of the tracheal connec- tions and following the various lines of development. Each wing typically receives air from two spiracles and the meta- thoracic spiracle (Fig. 17, spi) supplies the cubito-anal portion of the front wing and the costo-radial portion of the hind wing. It may be assumed that the three typical sources of air supply are equal. The Lepidoptera (Fig. 21) have the tracheal connections to the wings very generalized but the metathoracic spiracle (r. sp) is rudimentary in the pupa. This would seem to make the air supply to the cubito-anal portion of the front wing and the costo-radial portion of the hind wing inferior to the others and the tracheation of the two wings should be quite different. On the other hand the tracheae seem to have been unmodified and to have retained their typical relationships. It is difficult to apply the air-supply theory to the case of the honey-bee for the hind wing has lost its cubito-anal connection which was with the functional first abdominal spiracle (Fig. 32 spz) and has retained its costo- THE TRA CHEAT ION OF WINGS 49 radial connection (c-r) with the trunk of the vestigial metathoracic spiracle {v. sp'i). The posterior wing has, therefore, forsaken the best source of air supply and retained a connection with a trunk which no longer has an opening to the exterior. In all the cases where the wing tracheation has been reduced it is the cubito-anal trachea which has been lost. If the air supply is the factor causing the one connection to be lost and the other connection to be retained it is not easy to understand how it operates in the case of the metathoracic spiracle, for here the cubito-anal trachea to the front wing is lost and the costo-radial trachea to the hind wing is retained. To return to the case of the Odonata where the air-supply theory seems to find its best support, the posterior connection to the front wing should be considered. If the accessory cubito-anal trachea (Fig. 34, a. cu-a), which is enlarged, connected with the dorsal longitudinal connective {d. It) as Tillyard (/. c.) thought, it would be favorable to the air-supply theory, but it connects with a small transverse connective between the two dorsal longitudinal tracheae. This makes an indirect course for the air to follow in going to the wing and the condition would seem to be more easily explained as having arisen as an enlargement of a simple anastomosis of the accessory cubito-anal tracheae of both sides, a condition found in Melanoplus (Fig. 33) . The union of these tracheae might easily have anastomosed with a transverse connective to form a course which later became enlarged because it was the least obstructed route for the air to follow to the wings. It should be remembered that there are two anterior tracheae leading to the wing pad {c-r and a. c-r), each of which is about half the size of the posterior trachea (a. cu-a) and the two together must supply about as much air to the wing as this posterior trachea does. The condition of the leg tracheae in the Odonata is very instructive in the study of the values of the air supplies in the thorax. The legs, like the wings, each have two tracheal connections which typically derive their air supply from the same sources as do the tracheae to the wings. But in the Odonata the posterior stem of the leg tracheae, which should offer the more direct course for the air to follow in passing to the legs from the gills at the posterior end of the bod}', is greatly reduced, when the thoracic angle is great, while the anterior stem with its less direct cotuse is enlarged. This would indicate that the directness of the air supply has little to do with the modification of the leg tracheae in the Odonata. The belief that the migration of the medial trachea along the transverse basal trachea toward the cubito-anal group of tracheae might be due to the better air supply to that group was referred to in the introduction of this paper. In view of the fact that the basal connection of the posterior group of tracheae is the one to be lost in all cases where there is a reduction of the tracheal connections, it seems more reasonable to ascribe this migration of 50 THE TRACHEATION OF WINGS the medial trachea to pecuharities of thoracic structure, as has been done above. The invasion of the area of the radial sector by the medial trachea in the Odonata was also referred to in the introduction as a condition which might be due to the better air supply of the cubito-anal trachea. In the Odonata the distance between the bases of the radial and medial tracheae is so short that it would hardly seem that the medial trachea should be so modified as to invade the area of the radial trachea in order to save the air from passing so short a distance along the transverse basal trachea. The May-flies illustrate the same condition of the medial trachea invading the area of the radial sector, but here the connection of the cubito-anal trachea has been lost and the air supply of the wing has been completely cephalized so that the medial trachea is less favorably situated, with regard to the directness of the air supply, than is the radial trachea, the area of which it has invaded. From all these considerations it would seem that the directness of the air supply is not an important factor in controlling the relationships of the tracheae. However, the function of the trachege is to supply air and, even tho the surrounding structures may obsti-uct their way and cause them to become modified and adopt new and longer courses, they must still furnish an adequate supply of air to the tissues. The accessory cubito-anal tracheae to the hind wings of the Odonata is a case where the least obstructed course for the trachege to follow is also the shortest course, but the majority of cases, such as the modified base of the medial trachea of the stonefly and the single trachea to the wing of the honey-bee connecting with the trunk of a vestigial spiracle, have seemed to have no regard for the distance which the air must pass in reaching the tissues, but they have all adopted the least obstructed course for the air to follow. The length of the tracheae, therefore, does not seem to be the factor causing the modifications of structure, but rather a course for the tracheae which is free from obstnrction and which will allow for the passage of the air. Whether the air reaching the radial trachea comes from the costo-radial or cubito-anal trachea makes no difference to the tissues so long as the course of the radial trachea is unobstructed. The medial trachea, on the other hand, would not be expected to invade the area of the radial trachea simply because its base might be a millimeter or two nearer to the source of air supply any more than the radius should invade the area of the media in forms like the May- flies where the cubito-anal trachea has been lost. CONCLUSIONS The material upon which this discussion is based does not constitute an exhaustive study of this subject, for each order of insects presents an oppor- tunity for more detailed research. However, there are certain points with THE TRACHEATION OF WINGS 51 regard to the general subject of the basal connections of the wing tracheae which are clearly brought out by the study of the material presented. It has been shown that the basal connections of the tracheas of the wings of insects may be referred to a typical condition which is found in the more generalized forms and which may be traced thru various modifications to the more specialized groups. Very generalized conditions of the connections have been found in the Blattidag, among the Orthoptera, the Plecoptera, Neuroptera, and the Lepidoptera. The Hemiptera, Coleoptera, and Diptera show the begin- nings of specialization but have been grouped with those having generalized conditions. The typical conditions of the connections of the wing tracheae have been specialized by reduction in the Trichoptera, Ephemerida, and the Hymen- optera and this modification accompanies the specialized condition of the thorax. In other groups, the jumping Orthoptera and the Odonata, the tracheal connections of the wings have been modified by the addition of accessory tracheas and this condition is correlated with the position of the wing pads which are dorsally flexed in the nymphs of these two groups. Peculiarities of thoracic structure seem to be the most important factor in governing the course and relationships of the tracheas. Processes and apodomes have been found to lie in the typical paths of the tracheae so obstructing them that the trachea could not do otherwise than go around them even tho their course was lengthened by so doing. The directness of the air supply does not seem to be an important factor in controlling the course and relationships of the tracheae. Both the migration of the medial trachea toward the cubito-anal group, along the transverse basal trachea, and the invasion of the area of the radial sector by the medial trachea in the wing have been found to take place without regard to the directness of the source of the air supply. CHAPTER III THE MORE GENERAL FEATURES OF THE WINGS OF INSECTS The series of investigations briefly reviewed in the preceding chapters has led to the adoption of a uniform terminology of the wing- veins of insects. The application of this terminology to the wings of representatives of the several orders of insects is discussed in later chapters ; but, in order to avoid repetition, the more general features of the wings of insects are discussed here, and the terms applied to the difterent types of wings and to the parts of wings are defined. The presence or absence of wings in insects.^ — Excepting in that divi- sion of the Hexapoda, or insects, that is known as the Apterygogenea, and which includes the Collembola, the Campodeoidea, and the Thysanura, wings are usually present in adult insects. Regarding the absence of wings in the Apterygogenea two very different theories are held. The first was proposed by Friedrich Brauer, in his Systematisch-Zoologische Studien (1885). According to this theory, the Collembola, Campodeoidea, and Thysanura represent a branch of the insect series that separated from the main stem before wings were evolved, and which consequently is phylo- genetically distinct from the series of orders of insects that decended froms the primitive winged insects, and in which the absence of wings, when such is the case, is due to their having been lost. Brauer, therefore, separated the Hexapoda into two primary divisions: the Apterygogenea, or originally wingless insects; and Pterygogenea, or originally winged insects. The former division includes only the Collem- bola, Campodeoidea, and the Thysanura; the latter includes all other insects. The second theory is that maintained by Handlirsch in his Die Fossilen Insekten (i 906-1 908).. According to this theory all living insects have descended from the winged Palseodictyoptera of Paleozoic times ; and the absence of wings in the so-called Apterygogenea is due to specialization by reduction. It does not fall within the scope of this essay to discuss these rival theories. Even in the Pterygogenea many wingless forms are found; there being wingless representatives of nearly all of the orders of winged insects. But here the wingless condition is unquestionably an acquired one. The loss of wings by members of the Pterygogenea is often confined to a single sex of a species ; thus with the canker-worm moths, for example, the females are wingless, while the males have well-developed wings; on the other hand, with the fig-insect, Blastophaga, the female is winged and the male wingless. (52) THE GENERAL FEATURES OF WINGS 53 When the wings are all present, there are two pairs: one pair borne by the mesothorax; and one pair, by the metathorax; prothoracic wings are unknown among living insects. In certain cases one pair of wings may be either wanting or so modified in form as to be not fitted for organs of flight. Thus the flies, or Diptera, have only the first pair of wings fitted for flight, the second pair being repre- sented by a pair of knobbed threads, the halteres, which are probably organs of some special sense ; and with the earwigs and the beetles, the first pair of wings are greatly thickened and serve as wing-covers or elytra, which protect the hind wings when they are not in use. In these cases the hind wings are the chief, if not the only organs of flight. The fundamental structure of the wings of insects. — Studies of the development of wings have shown that each wing is a saclike fold of the body- wall ; but in the fully developed wing, its saclike nature is not obvious ; the upper and lower walls become closely applied throughout the greater part of their extent; and, since they become very thin, they present the appearance of a single delicate membrane. Along certain lines, however, the walls remain separate, and are thickened, forming the firmer frame- work of the wing. These thickened and hollow lines are termed the veins of the wing; and their arrangement is described as the venation of the wing. Many of the older writers termed the wing-veins nerviires and their arrange- ment the netiration of the wing. The thin spaces of the wings which are bounded by veins are called cells. When a cell is completely surrounded by veins it is said to be closed; but when it extends to the margin of the wing it is said to be open. The different types of insect wings. — What may be regarded as the typical foim of insect wing is a nearly flat, delicate, membranous appendage of the body, which is stiffened by the so-called wing-veins. But striking modifications of this form exist; and to certain of them distinctive names have been applied. The fan-like vuings. — While the typical form of insect wing presents a nearly flat surface, in many insects the wings are more or less corrugated, and in some the corrugating of the wings has proceded so far that a fan-like form of wing has resulted. Among the more perfectly fan-like wings two quite distinct types can be recognized; these may be designated as the fixed fan-like type and the folding fan-like type respectively. The most perfect examples of the former are the wings of the Ephemerida; of the latter, the anal area of the hind wings of the Acrididse. In the fixed fan-like type the corrugating of the wing serves to strengthen it, and thus fit it to withstand better the strain brought upon it by beating the air in flight. This can be easily illustrated by first attempting to use a flat sheet of paper as a fan and then after folding the paper in plaits using 54 THE GENERAL FEATURES OF WINGS it for that purpose. In the fixed fan-hke type of wing the so-called concave and convex veins are qtiite evenly developed. The folding fan-like type of wing is better fitted to function as a para- chute when outspread than as an organ for beating the air; in this type the convex veins are much more prominent than the concave veins. The elytra. — In the Coleoptera and in the Dermaptera the front wings are thickened and serve chiefly to protect the dorsal wall of the body and the membranous hind wings, which are folded beneath them when not in use. Front wings of this type are termed wing-covers or elytra. The hemelytra. — The front wings of the Heteroptera, which are thickened at the base like elytra, are often designated the hemelytra. The tegniina. — The thickened fore wings of Orthoptera are termed the tegmina by many writers. The halteres. — The hind wings of the Diptera, which are knobbed thread- like organs, are termed the halteres. The hind wings of the males of the family Coccidae are also thread-like. The pseudo-halteres . — The re- duced front wings of the Strepsip- tera are known as the pseudo-halteres . The margins of wings. — An insect's wing is more or less tri- angular in outline; it, therefore, presents three margins: the costal ,w.i margin or casta (Fig. 36, a-b); the Fig. 36.-Diagram of a wing showing ; • ^p- ^ ^^ ^^ ^^ margins and angles. '^ _ v & o > /> inner margin (Fig. 36, c-d). The angles of wings. — The angle at the base of the costal margin of a wing is the Immeral angle (Fig. 36, a); that between the costal miargin and the outer margin is the apex of the wing (Fig. 36,6); and that between the outer margin and the inner margin is the anal angle (Fig. 36, c). The tegula. — In several orders of insects there is at the base of the costal vein a small, hairy, shghtly chitinized pad; this is the tegida (Fig. 38, Tg). In the more speciaHzed orders, the Lepidoptera, the H\Tnenoptera, and the Diptera, the tegula is largely developed so as to form a scale-like plate overlapping the base of the wing. The tegulse of the front wings of Lepidoptera are specially large and are carried by special tegidar plates of the notum. These in turn, are sup- ported by special internal tegidar arms from the bases of the pleural wing processes (Snodgrass, '09). The axillary cord. — The posterior margin of the membrane at the base of the wing is sually thickened and corrugated; this cord-like structure is termed the axillary cord. The axillary cord normally arises, on each side, THE GENERAL FEATURES OF WINGS oo from the posterior lateral an^^le of the notum, and thus serves as a mark for determining^ the ijosterior limits of the notum. The axillary membrane.— The membrane of the wing base is termed by Snodgrass the axillary membrane; it extends from the tegula at the base of the costal margin of the wing to the axillary cord at the base of the anal area of the wing ; in it are found the axillary sclerites ; it is most prominent at the base of the anal area of the wing, where it is margined by the axillary cord. anr r pnr The alula. — In certain families of the Dip- tera and of the Coleoptera the axillary mem- brane is expanded so as to form a lobe or lobes which fold beneath the base of the wing when the wings are closed; this part of the wing is the alula or alulet. The alulas are termed the squamce by some writers, and the calypteres by others. The articulation of the wings. — The wings are articulated to the lateral margins of the notum of the two wing bearing segments. In the membrane at the base of each wing, the axillary membrane, there are several axillary sclerites, which are relatively constant in posi- tion but which vary in form in different insects. These axillary sclerites have been much studied for nearly one hundred years. They were described in the early part of the last ^rnc centurv bv Jurine (1S20), Chabrier (1820), Fig. 37. — Lateral view of a wing- and Straus-Durckheim (1828); the more * bearing segment" (From ^ ' ' Snodgrass). important of the later papers on this subject are those of Amans (1885), Lowne (1892), Voss (1905), Berlese (1909), and Snodgrass (1909, 1910 a, 1910 b.). The accounts by Snodgrass are the most complete and logical, and as his terminology of the parts of the thorax to which the wings are attached and of the axillary sclerites is the simplest I have adopted it here. The follow- ing statement is compiled from these accounts. The thoracic supports of the wings. — Figure 37 is a diagram, given by Snodgrass, of a lateral view of a generalized, wing bearing segment, left side. The chief supports of the wing are two processes of the notum and one of the pleurum; these are the anterior notal ifing process (ANP). the posterior notal wing process (PNP), and the wing process of the pleurum (WP). Behind the posterior notal wing process is the attachment of the axillary cord (AxC). 56 THE GENERAL FEATURES OF WINGS The above generalization is true of all of the orders of winged insects except the Ephemerida and the Odonata. In these two orders the flexible bases of wing-veins merge into the lateral margins of the notum, and only one distinct axillar}^ sclerite is present at the base of each wing. The axillary sclerites. — There are typically four axillary sclerites; but occasionally some of these are subdivided and sometimes there occur small extra chitinizations in the axillary membrane. The four principal axillary sclerites are designated as the first, second, third, and fourth axillar}^, respectively. Their relative positions are indi- cated in Figure 38. "Two of these, the first {lAx) and \he fourth {4AX), form a hinge with the anterior and the posterior notal wing processes, respectively, while the second {2 Ax) articulates below with the wing process of the pleurum, constituting thus a sort of pivotal element. The third Tg — Fig. 38. — Diagram of a wing showing the axillary sclerites (From Snodgrass). axillary {sAx) inteiTnediates between the bases of the anal veins and the fourth axillary — except when the latter is absent ( as it is in nearly all insects except Orthoptera and Hymenoptera), in which case it articulates directly with the posterior notal process. "The base of the costa is not directly associated with any of the axil- laries, but is specially connected by tough membrane below with the episternal paraptera. The subcosta abuts against the end of the curved neck of the first axillary. The radius is either attached to or touches ui)on the anterior end of the second. The media and the cubitus are usually associated with each other at their bases and also more or less closely with one or two median plates (ni) in the wing base. These plates, however, are not of constant shape and occurrence as are the articulating axillarics. The anals are generally attached to the outer end of the third axillary, which acts as a lever in the folding of the wing. "A few insects have a generalized wing almost identical with the diagram (Fig. 38), but most of them depart from it in varying degrees." (Snodgrass, '10). THE GENERAL FEATURES OF WINGS 57 The costal sclerite. — The costa does not connect with any of the axillary sclerites named above ; but there is very generally at its base a more or less distinct sclerite, which may be termed the costal sclerite. This is regarded by Snodgrass as merely the base of the costa; but as it is frequently a dis- tinct sclerite of considerable width it merits a special name. In the hind wings of many of the Lepidoptera, the costal sclerite forms a strong support for the frenukim. This sclerite is represented, but not lettered, in the figiue of the wings of Prionoxystus (Fig. 62) ; it is also very prominent in the hind wing of Cacoecia (Fig. 353). Figures of the bases of many insect wings showing the form and position of the axillaries are given by Snodgrass. The tuberosities of the base of the wing. — At the base of each wing in most insects, there are two prominent, elevated, shoulder-like areas; these were designated by Amans ('85) as the anterior tiiherosity and the posterior tuberosity respectively. These tuberosities are better defined in the fore wings than they are in the hind wings. The anterior tuberosity is at the base of the radius and the adjacent veins, and the posterior tuberosity is at the base of the anal veins. The posterior tuberosity is sometimes di\ndcd into two or three tuberosities, corresponding to the separate anal veins. The cubito-anal sulcus.^ — The deep channel between the anterior and the posterior tuberosities may be termed the cubito-anal sulcus. In most insects the cubito-anal sulcus is traversed by the basal part of a wing-vein. This vein is either the cubitus, the first anal vein, or the coalesced bases of these two veins, differing in different insects. In a few insects, as in Corydalus, there is no vein at the bottom of the sulcus, the cubitus extending along one side of it and the first anal vein along the other. As a rule, a vein traversing this sulcus is more or less atrophied at its base. The corrugations of the wings. — The wings of comparatively few insects present a fiat surface ; in most cases we find that the membrane of the wing is throwai into a series of folds or corrugations. This corrugating of the wing in some cases adds greatly to its strength; this is well-shown by the wings of dragonflics; and in most orders the costal margin of the wing is strengthened by a plication between the costa and the radius, this is the subcostal fold. In other cases, the conjugations are the result of a folding of the wing when not in use; this is well-shown in the anal area when this part is broadly expanded. The cubito-anal fold. — There is one fold that is almost universally present in the wings of insects. This fold is a continuation of the cubito-anal sulcus, which usualU' extends to the margin of the wing; it may be tenned the cubito-anal fold. It is in this fold that the anal furrow is developed when present. 58 THE GENERAL FEATURES OF WINGS The anal area and the preanal area.-^In descriptions of wings it is frequently necessary to refer to that part of the wing that is supported by the anal veins ; this is designated as the anal area of the wing ; and that part of the wing lying in front of the anal area, including all of the wing except the anal area, is termed the preanal area. The furrows of the wing. — There are found in the wings of many insects one or more suture-like grooves in the membrane of the wnng; these are termed the furrows of the wing. The following furrows have received dis- tinctive names. They occur chiefly in the fore wings. The anal furrow. — The anal furrow is the one of the furrows of the wing that is most often present. It is developed in the cubito-anal fold; but in many insects where there is a well-devel- oped cubito-anal fold there is no definite, suture-like anal furrow. The anal furrow is usually either between the ciibitus and the first anal vein or it is it may supplant in forms in This is the case in the Fig- 39- -Wings of Bombyx mori. coincident with the first anal vein which which the venation of the anal area is reduced Lepidoptera and the Diptera and is well-shown in the wings of Bombyx mori (Fig. 39), in which a vestige of the first anal vein is preserved near the margin of the wing. In the Hymenoptera, where the anal furrow is developed between the cubitus and the first anal vein, it cuts through those veins, the tips of which, coalescing with the anal veins, crosses its course (Fig. 40). In those Heteroptera in which we have been able to determine the veins in the fore wings, the anal fun-ow has been developed in front of the cubitus (Fig. 41, a/). A study of the musical organs of adult Orthoptera throws light on the nature of the anal furrow. In the female this furrow lies in its usual posi- THE GENERAL FEATURES OF WINGS o9 tion between the cubitus and the first anal vein; but in the males of the Locustidas and the Gr\41ida3 the anal furrow crosses vein Cu2. It is evident, Fig. 40.- -Typical hymcnopterous wing. The median furrow and the anal furrow are indicated by dotted lines. therefore, that this furrow is merely a fold in the adult wing, and that its position is variable. The median furrow. — This is a longitudinal furrow which is usually between radius and media (Fig. 41, mf). It is well-marked in many of the Heteroptera, where it separates the embolium from the remainder of the corium (Fig. 42) ; and in the Hymenoptera its course is marked by a series of weak spots, the bullae, in certain veins (Fig. 40). The nodal fiirroiv. — The nodal furrow is a transverse suture beginning at a point in the costal margin of the wing corresponding to the nodus of the Odonata and extending towards the inner margin of the wing. It crosses a varying number of veins in different orders of insects. It is well-shown in %i 'fe S{ » % 'q:9' '■^V^S^f vA<;*«'' Fig. 41. -Fore wing of a bug, Hormostes reflex ul us; ni f, median furrow; a f, anal furrow (After C. & X.). the fore wing of the flower-bug (Fig. 42) ; and in the fore wing of a Cicada (Fig. 43, n f). The nodal furrow is termed the costal hinge by some writers. The axillary furrow. — The axillary furrow is a suture-like line extending from the base of the wing to the inner margin ; it ends at the axillary excision 60 THE GENERAL FEATURES OF WINGS Fig. 42. — Fore wing of a flower-bug (Acanthiidae) ; e embolium. when this is present. This Hne serves as a hinge which faciUtates the fold- ing of the posterior lobe of the wing under that part of the wing in front of it. The axillary excision. — In the wings of most Diptera there is a notch in the inner margin of the wing near its base (Fig. 44, ae) ; this has long been known as the axillary excision; the application of this term may be extended advantageously to the notch in a similar position that exists in the wings of many other insects. The posterior lobe. — That part of the wing lying between the axillary excision and the axillary membrane at the base of the wing in the wings of Diptera (Fig. 44, /) has been termed the posterior lobe. The development of a posterior lobe, however, is not limited to the Diptera, a distinct posterior lobe being found in the representatives of several of the orders. In some cases, as in those Diptera where an alula is developed, the posterior lobe and the axillary membrane are distinctly differentiated ; but in other cases they are not, as in the wings of Corydalus. The posterior lobe of the fore wings of certain insects has been specialized so as to serve as an organ for uniting the fore and hind wings. Two types of such an organ are defined later. The posterior lobe of the wing and the clavus of the Heteroptera are not homologous, the clavus including a much larger part of the wing than does the posterior lobe. The methods of uniting the two wings of each side. — It is obvious that a provision for ensuring the synchronous action of the fore and hind wings adds to their efificiency ; it is as important that the two pairs of wings should act as a unit as it is that the members of a boat's crew should pull together. In many insects the synchronous , "^ action of the wings is ensured by the fore wing of each side overlapping the hind wing. But in other insects special struct- ures have been devel- oped which fasten Fig. 43. — Fore wing of a (Cicada; n f, nodal furrow. together the two wings of each side. Tlic different types of these struc- tures have received special names. The hamuli. — With certain insects the costal margin of the hind wings bears a row of hooks, which fasten into a corresponding fold on the inner THE GENERAL FEATURES OF WINGS 61 margin of the front wings (Fig. 45). These hooks are named the hamuli, and ser^'e to hold the two wings of the same side toegther, thus insuring their action as a unit. The frenulum and the frenulum hook. — In most moths there is a strong spinc-hke organ or a bimch of bristk'S borne by the hind wing at the humeral 2d A+Cu, Fig. 44. — Wing of Conops; a e, axillary excision; /, y)osterior lobe of the wing. angle (Fig. 46,/) ; this is the frenulum or little bridle. As a mle the frenu- lum of the female consists of several bristles; that of the male, of a single, strong, spine-like organ. In the males of certain moths, where the frenu- lum is highly de\-eloped, there is a membranous fold on the fore wing for receiving the end of the frenulum, and thus more securely fastening the two wings together: this is tha frenulum hook (Fig. 46, / h). The jiigttm. — In one family of moths, the Hepialid^e, the posterior lobe of the fore wing is a slender, finger-like organ, which is stiffened by a branch of the third anal vein, and which projects beneath the costal margin of the hind wing. As the greater part of the inner margin of the fore wing over- laps the hind wing, the hind wing is held between the two. This type of posterior lobe of the fore wing is termed the jugum or yoke. Figure 47 represents the wings of a hepialid seen from below and shows the manner in which the two wings of one side are yoked together. The jugum is behind one of the two principal branches of the third anal vein, which may be designated as vein sdAi, and is supported by the other of these two branches, which may be designated as vein sd^lo (Fig. 48). The fibula. — In several groups of insects an organ has been developed that serves to unite the fore and hind wings, but which functions in a way quite different from that of the jugum. Like the jugum it is found at the Fig. 45. — Wings of a honey-bee; /;, hamuli. 62 THE GENERAL FEATURES OF WINGS Fig. 46. — Wings of Thyridopleryx ephemercs- formis; f, frenulum; / h, frenulum hook. base of the fore wing; but unlike the jugum it extends back above the base of the hind wing and is clasped over an elevated part of the hind wing. This organ may be termed the R,^^,^^'^''^ _ fibula or clasp. The fibula differs in different insects in respect to the part of the wing that enters into its composi- tion. In some insects it is strictly homologous with the jugum being composed merely of the posterior lobe of the wing; in other insects it consists of the posterior lobe of the wing and a part of the axillary membrane. These two types are illustrated by the fibulas of Rhya- cophala and of Corydaliis respec- tively. The fibula of Rhyacophila, a generalized member of the order Trichoptera, is a hatchet-shaped lobe at the base of the wing (Fig. 49) . It is joined to that part of the anal area lying in front of it by the axillary furrow, which acts as a hinge. It is evident that the structure of the wing is such that the fibula is depressed with consider- able force, for it is very diffi- cult to mount a wing without the fibula being folded under it. As the longitudinal free margin of the fibula is curved down it seems probable that the fibula clasps the anterior tuberosity of the hind wing and thus the two wings are held together. Although the fibula of Rhyacophila differs greatly in appearance from the jugum of the Hepialidae and functions in a different man- ner, the two are formed of homologous parts of the wing. In Rhyacophila the axillary furrow is immediately behind vein 3dAi (Fig. 49); in the Hepialidae there is a slit Fig. 47. — Wings of a hepirilid, seen from below. THE GENERAL FEATURES OF WINGS 63 in the wing in this position; in both the longitudinal margin is strength- ened by vein 3dA2. The fibula of Corydalus differs in structure from that of Rhyacophila in that it is composed in part of the axillary membrane. It is a triangular, backward projecting lobe of the wing, along the middle of which extends vein 3dAo (Fig. 50). That part of this fibula that lies distad of this vein is homologous with the jugum of the Hepialidge and the fibula of Rhyacophila; the part lying proximad of this vein is a por- evident from the. fact that it is Fig. 48. — Jugum of a hepialid. tion of the axillary membrane; this margined by the axillary cord. The apex of this fibula, which is strengthened by the tip of vein 3dA2, curves down forming a hook. In the living insect, this hook clasps a fold in the hind wing, which extends from the apex of the anterior tuberosity to the margin of the cubito-anal sulcus. In this manner the two wings are clasped together. The relation of the fibula to the hind wing can be seen only rarely in dried insects. In the spreading of specimens the fibula is pulled out of its normal position; and in drying both the fibula and the fold of the hind wing behind which it hooks become 1 • 1 J it, j.j.1. ■ 111.- Fig- 49- — VihwXa. oi Rhxacophila. shriveled so that their normal relations ' are not obvious. The conclusions given above are based on a study of living individuals. The two types of fibular described above will serve as a basis for more extended studies of this organ. Doubtless much material of taxonomic value can be obtained from a study of this part of the wing. But such a study docs not fall \vithin the scope of this essay, the preparation of which was undertaken with special reference to the application of the uniform terminology of the wing-veins. An expanded humeral angle of the hind ivings. — In certain insects the humeral angle of the hind wings is greatly expanded so that it projects Fig. 50. — Fibula of Corydalus. 64 THE GENERAL FEATURES OF WINGS forward beneath the fore wing thus insuring the synchronous action of the two wings. Examples of this exist in several families of moths, in butter- flies, and in the Ephemerida. The clothing of the wings. — With the greater number of insects the wings appear to be naked ; but in many cases they are only apparently so, an examination with a microscope revealing a fine covering of setag or spines. On the other hand, the wings of certain insects are obviously clothed. Thus the wings of the aquatic Trichoptera are densely clothed with long hairs, and those of the Lepidoptera are covered with scales. The more conspicuous clothing of wings is composed of setae more or less modified. In the aquatic Trichoptera the setfe are essentially typical in form, being hair-like stitxctures attached to the membrane of the wing by a C •^^' Sc, Fig. 5I-- -Hypothetical tracheation of a wing of the primitive nymph (After C. & N.). movable joint. In the Lepidoptera and in certain other insects, the setae are modified into scales. In addition to setce the wings of certain insects bear hair-like stmctures which are morphologically different from setae; they lack the joint at the base and are probably merely elongated cuticular nodules. A more extended discussion of the clothing of the wings is given later in the discussion of the wings of the Lepidoptera. The principal wing-veins. — It has been shown in the previous chapters that in the more generalized insects the principal wing-veins arc developed about preexisting tracheae; and a study of these tracheae in the more generalized members of several orders of insects has enabled us to construct a diagram rejjresenting the hypothetical type of the primitive wing-vena- tion. In nymphs and i^upas the tracheae about which the principal veins are later developed extend separate from within the thorax, and their courses are easily traced. But it frecjuently happens that througli the narrowing THE GENERAL FEATURES OF WINGS 65 of the base of the wing, which occurs at the change to the adult form, the bases of these tracheae are crowded together, which results in the coalescing of the bases of two or more veins, so that they appear as a single vein. For this reason the figure of the hypothetical type (Fig. 51) is made to represent the tracheation of a nymph ; and the different tracheae are designated by the same terms as are the veins to which they correspond. By representing the wing of a nymph, we are able to represent the basal connections of the tracheae that precede the wing-veins, and thus show which are principal veins and which are branches of them. A glance at Figure 51 will show that there are eight principal veins; and that the second, third, fourth, and fifth are branched. The names of these veins and the abbreviations by which they are designated are as follows, beginning with the one nearest the costal margin : Names of veins Abbreviations Costa C Subcosta Sc Radius R Media M Cubitus Cu First Anal ist A Second Anal 2d A Third Anal 3d A Enderlein ('02) has revived the old name axillaris and applied it to the second and third anal veins, which he designates as axi, and a.v'2 respectively, and applies, the term an alls to the first anal vein. He states that he does this in order to better characterize the morphologically different veins grouped as anal veins (Z, c. p. 15). But he does not state in what respect the second and third anal veins differ morphologically from the first anal vein. I see no reason for making the change that he suggests. The names of the wing-veins used in the uniform teiTninology arc those adopted by Redtenbacher in his work upon which this terminology was based. The terms costa, subcosta, radius, cubitus, and anal veins were selected by this author from those alread}^ in use; the term media was proposed by him for what appears to be the middle vein, ''in der That als die 'Mittelader' erscheint," that is as the vena media. The changing of this term to the masculine fonn, "medius," as has been done by several recent writers, is unwarranted. The chief branches of the wing-veins of the preanal area. — The chief branches of the principal veins are numbered, beginning with the branch nearest to the costal margin of the wing. The term used to designate a branch of a vein is fonned by compounding the name of the vein with a numeral indicating the number of the branch. The names of the branches 66 THE GENERAL FEATURES OF WINGS that are thus recognized and the abbreviations of these names that are used are the following : Names of branches Abbreviations Subcosta-one Sci Subcosta-two Sc2 Radius-one Ri Radius-two R2 Radius-three R3 Radius-four R4 Radius-five R5 Media-one Mi Media-two M2 Media-three M3 Media-four M4 Cubitus-one Cui Cubitus-two Cu2 In the case of radius and of media, each of which has more than two branches, each division of the vein that bears two or more branches has received a special name. Thus after the separation of radius-one from the main stem of radius there remains a division which is typically four- branched ; this division is termed the radial sector or Rg ; the first division of the radial sector, which later separates into radius-two and radius-three, is designated as radius- two-plus-three or R2+3; and the second division is termed radius-four-plus-five or R4+5- Media is t^^pically separated into two divisions, each of which is two-branched; the first division is media- one-plus-two or Ml +2, the second, media-three-plus-four or M3+4- The veins of the anal area. — The three veins of the anal area, which are designated as the first anal (ist A), second anal {2d A), and third anal {3d A) respectively, exhibit a wide range of variation both as to their persistence and as to their form when present. In wings with a reduced anal area any or all of the anal veins may be lacking; on the other hand, in wings with an expanded anal area some or all of the anal veins may be branched. In those cases where the anal veins are branched there is no indication that the branching has been derived from a uniform primitive type of branching, as is the case with the other branched principal veins. The principal branches of either radius, media, or cubitus can be homologized, when present, throughout the insect series; and, therefore, have been given distinctive names; but there is no reason to believe that this can be done with the branches of the anal veins. For this reason in describing a branched anal vein merely the number of branches is indicated. In many cases where there is a reduction in the number of anal veins a study of the development of the wing will reveal the manner in which the THE GENERAL FEATURES OF WINGS 67 reduction has taken place; it may be due either to the atrophy of one or more veins, or it may be due to the coalescence of veins. In either case if the manner of reduction be determined the homology of the remaining vein or veins can be indicated. But in some cases, as for example, in the Odonata, there has been a specialization b}' reduction, which has resulted in the preservation of a single anal vein, and there is available no data showing the course of this evolution. The three anal veins may have coalesced into one: or one anal vein may ha\'e atrophied and the remaining two coalesced; or two of the veins may have been lost, leaving a single vein. Obviously, in this state of uncertainty, it is best to designate the single vein as the anal vein (A) without any predicate. When a single anal vein is present the identity of which can not be determined and this vein is branched, we are not warranted in designating these branches to make use of the names istA, 2d A, and 3d A, as this would indicate homologies that we do not know to exist. Therefore in such cases the branches of vein A are designated as .4i, A2, A3, etc. The reduction of the number of wing-veins. — In many wings the num- ber of the veins is less than it is in the hypothetical type. In some cases this is due to the fact that one or more veins have faded out in the course of the evolution of the insects showing this deficienc}^; frequently in such insects vestiges of the lacking veins remain, either as faint lines in the positions formerly occupied by the veins or as short fragments of the veins. A much more common way in which the niimber of veins has been reduced is by the coalescence of adjacent veins. In many wings the basal parts of two or more prinicpal veins are united so as to appear as a single vein; and the number of the branches of a vein has been reduced in very many cases by two or more branches becoming united throughout their entire length . The evolutionary process by which the number of the veins has been reduced, whether by the fading out of veins or by the coalescence of adja- cent veins, is termed specialization by reduction. The coalescence of principal veins is usually made evident by a study of their branches, the branches of two or more veins arising from a common stem. The coalescence of the branches of a vein is often shown, by a study of a series of allied forms in which different degrees of it can be observed, the point of separation of two branches being nearer and nearer to the margin of the wing in successive forms until the margin is reached. As the venation of the wings affords some of the most available charac- teristics for use in the classification of insects, it is of the utmost importance .that the same name should be applied to homologous veins in the wings of different insects. For this reason when a vein consists of two or more of the primitive veins united, the name applied to the compound veins should 68 THE GENERAL FEATURES OF WINGS indicate this fact. In the wing of Rhyphiis (Fig. 52), for example, radius is only three-branched; but it would be misleading to designate these branches as Ri, R2, R3, for this would indicate that R4 and R5 are lacking. /?, ^.+3 id A Fig. 52. — Wing of Rhyphus. The first branch is evidently Rr, the second branch is composed of the coalesced Ro and R3, it is, therefore, designated as R2+3; and the third branch, which consists of the coalesced R4 and R5 is designated R4+5- The wing of Rhyphus illustrates also the other method of reduction of wing- veins, as the basal part of media has faded otit. A second method of coalescence of veins is illustrated by the wing of Tahanus (Fig. 53). In this wing the tips of cubitus-two and the second anal vein are united; here the coalescence began at the margin of the wing and is progressing towards the base. The united portions of the two veins are designated as 2d A + Cu2. In those cases where the reduction of the venation has proceeded very far, as in some of the smaller Hymenoptera for example, the logical carrying 2d A^'^Cu, FiS- 53- — Wing of Tahanus. out of the plan of designating a compound vein by a combination of the names of all of the veins that have entered into the composition of it would result in a very cumbersome terminology. Such a designation, however THE GENERAL FEATURES OF WINGS 69 complex, may be desirable in detailed taxonomic work; but for the pur- poses of ordinary descriptions a designation indicating the most obvious element of the compound vein is sufficient. Thus in the hind wing of /st A Fig. 54. — Hind wing of an ichneumon-fly. Exetastes (Fig. 54), where veins M3, M4, Cui, Cuo and 2d A coalesce with the first anal vein the united tips of these veins is designated, by the term indicating its most obvious element, ist A. Serial veins. — In the wings of some insects, where the wing-venation has been greatly modified, as in certain Hymenoptera, there exists what appear to be simple veins that in reality are compound veins composed of sections of two or more veins joined end to end with no indication of the point of union. For compound veins formed in this manner I propose the term serial veins. The following comparatively simple example will sen^e as an illustration of a serial vein. In those Hymenoptera in which the wing-venation is not greatly reduced vein M2 of the fore wings extends in a transverse direction from the point of separation from vein Mi to the point where it is joined by the medial cross- vein ; it then bends more or less abruptly and extends in a longitudinal direction. This condition persists in the fore wing of an ichncmon- fiy (Fig. 55). But in the fore wing of a braconid (Fig. 56) the transverse section of vein M, has been lost; and the p^^ --._wings of an ichneumon-fly. longitudinal section of this vein and the medial cross-vein form together what appears to be a simple vein. The example just cited is perhaps the most easily understood of all of the serial veins. But there are cases where sections of several veins enter 70 THE GENERAL FEATURES OF WINGS Fig. 56. — Wings of a braconid. into the composition of a serial vein, and where, owing to extensive modifica- tions of the wing-venation, it is impossible to determine completely the composition of the resulting vein. The problem of devising a method of designating serial veins now arises. In cases where the vein consists of only two elements it is a simple matter. The serial vein in the braconid wing described above, consisting of the medial cross-vein (m) and vein M2, can be designated as m & Mo. The sign & is used instead of + , as the latter is used to indicate compound veins f onned by the coal- escence of veins side by side, as indicated on an earlier page. In those cases where sections of several veins enter into the composi- tion of a serial vein, it is usually possible to deter- mine the basal element and also that forming the tip of the serial vein, even though it is impossible to determine defin- itely the veins that are represented by the intermediate portion of the vein. In such cases the serial vein can be designated by the abbre\'iation of the name of the basal element connected by a dash with the abbre- viation of the name of the terminal portion. Thus a serial vein, the basal element of which is the cubitus and the terminal element vein Mi is desig- nated as vein Cu-Mi. A serial vein thus formed exists in the hind wings of certain ichneiunon-flies (Fig. 54). This method of designating serial veins is suggested by Professor Bradley as a substitute for that used in his monograph of the Evaniida? (Bradley '08). The increase of the number of wing-vems. — In many insects the wings are furnished with a larger number of veins than that which we believed to have been characteristic of the wings of the primitive winged insects; the evolutionary process by which this change has been brought about is termed specialization by addition. This multiplication of veins is due either to an increase in the number of the branches of the principal veins by the addition of secondary branches, or to the development of secondary longitudinal veins between these branches. In no case is there an increase in the number of principal veins. In some cases the specialization by addition has taken place only in the preanal area; in others it is confined to the anal area; in still others it has occurred in both areas ; and in many insects a reduction in the number of THE GENERAL FEATURES OF WINGS 71 the veins has taken place in one part of the wing while there is an increase in another part. The secondary longitudinal veins. — Omitting certain adventitious veins, which will be discussed later, and branches of the anal veins, the secondary longitudinal veins are of two kinds, which may be designated as the accessory veins and the intercalary veins, respectively. These two kinds of veins differ fundamentally in their mode of origin. Accessory veins arise as secondary branches of the principal veins or of their branches; and their Fig. 57. — Wings of Osmyius. presence is not correlated in any way with a corrugating of the wing. Intercalary veins, on the other hand, are thickened folds; each of which arises more or less nearly midway between two preexisting veins, with which, as a rule, it has no connection except by cross-veins, and can not, therefore, be considered accessory to either. Frequently, however, an in- tercalary vein becomes joined to one of the two veins between which it was developed in such a way as to appear to be a branch of it. The accessory veins. — The secondary longitudinal veins that arise as branches of the principal veins or of their branches are teiTned accessory veins. Accessory veins may be borne by any of the primitive longitudinal ycins; and they may arise from either of the two sides of such a vein. 72 THE GENERAL FEATURES OF WINGS Accessory veins are found only in the Orthoptera, Isoptera, Neuroptera, Plecoptera, Mecoptera, and, rarely in the Trichoptera and Embiidina. Two types of accessory veins can be recognized, although in many cases it is difficult to detennine with which of the two types certain accessory veins should be classed. These two types may be designated as marginal accessory veins and definitive accessory veins respectively. The marginal accessory veins are twig-like branches that are the result of bifurcations of veins that have not extended far back from the margin of the wing. Many such short branches of veins exist in the wings of Osmylns (Fig. 57). Note especially the outer margin of the wing where the forking of the veins is quite regular and where some of the marginal acces- sory veins are themselves forked. The most important characteristic of the marginal accessory veins, for the purposes of this discussion, is the fact that their number and position are not at all constant. Not only do they vary in these respects in different Fig. 58.- — A wing of a pupa of Corydalus (After C. & N.). individuals of the same species but also in the wings of the two sides of the same individual. There is on this account no occasion for providing a terminology for them. In the paleozoic insects the marginal accessory veins were very irregular in length, this is shown by the figures illustrating Chapter IV. Strongly contrasting with this condition is that seen in many recent Neuroptera, where there are one or more ranks of these veins, the members of each rank being of comparatively uniform length. The definitive accessory veins differ from the marginal accessory veins in having attained a position that is comparable in stability to that of the primitive branches of the principal veins; for this reason it is practicable to designate them individually, either by names or by numbers, depending on the manner in which they have been added. Comstock and Needham pointed out that accessory veins arc added to the principal veins in two ways: first, in some insects they are added distally by successive splittings of the tip of a principal vein, thus fomiing a regular series; and second, the number of accessory veins may be increased THE GEXERAL FEATURES OF WINGS 73 in an iiTegular manner by interpolation, i. e. by the splitting of various members of a series of accessory veins. Illustrations of the adding of accessory veins distally are to be seen in Corydalus and allied genera. The presence of fine twigs at the tip of Fig. 59. — Fore wing of a nym])h of a cockroach (After C. & X.). trachea R2 in the pupal wing of Corydalus (Fig. 58) indicates the method of increase, which is doubtless as follows: the branches have been added one after another to the tip of trachea R-j, there being a migration of the base of each accessory trachea towards the base of the wing, thus making room for the addition of new branches. In this case the first accessory vein is the proximal one. In the wing of a nyinph of a cockroach represented by Figure 59 there are manv accessorv trachea? branching from the front side of the radial fsi A Fig. 60.- — -A wing of a pui)a of Chanliodes (After C. & N.). trachea. From the presence of the iinc twigs near the apex of the wing, it is evident that accessory tracheae are being added distally. It is also evident that the number of accessory tracheae is being increased by the splitting of some of these accessory tracheae, i. e., by interpolation. 74 THE GENERAL FEATURES OF WINGS Where the veins are added distaUy so as to form a regular series it is practicable to designate them individually. In this case the accessory veins arising from one side of a primitive vein are considered as a single series, and to each series a distinct set of letters is applied beginning with the Fig. 6i. — Wing of a May-fly, Epeoms humeralis (After Morgan). first developed member of the series of accessory veins. Thus if vein R2 bears three accessory veins they are designated as veins R2a, R2b. and R2C respectively (Fig. 60). By this method homologous veins, when a homology exists, will bear the same designation. But it should be remembered that as accessory veins have arisen independently in many different groups of insects, it often happens that accessory veins similar in position, and bearing the same designation in our system, are merely analogous and not homologous. The discussion of the sequence of development of definitive accessory veins is continued in the Chapter treating of the wings of Neuroptera; as it is in this order that this class of veins reaches its most perfect develop- ment. The intercalary veins. — The intercalary veins are secondarily developed longitudinal veins that did not arise as branches of the primitive veins, but were developed in each case as a thickened line, in a corrugated wing, more or less nearly midway between two preexisting veins, with which primarih^ it was connected only by cross veins. In many cases, however, an intercalary vein has become united to one of the two veins between which it was developed in such a way as to appear to be a branch of it. This resemblance to an accessory vein is greatly increased when the intercalary vein is penetrated by a branch of the trachea of the principal vein; this condition is well-illustrated by the intercalary veins of the Odonata. Excellent illustrations of unmodified intercalary veins are common in the Ephemerida, where most of the intercalary veins remain distinct from the veins between which they were developed, being connected with them only by cross-veins, the proximal end of the inter- calary vein being free (Fig. 61). THE GENERAL FEATURES OF WINGS In the formation of an intercalary vein there is a straightening out of veins that form part of the boundaries of polygonal cells into a longitudinal line. Different stages of this development can be seen in the wings of dragon-flies; see figures in Chapter XL Intercalary veins did not exist in the wings of the Palaodictyoptera. Their development is correlated with the development of corrugated wings. In the preanal area of recent insects, intercalary veins are found only in the Ephemerida and in the Odonata. In the anal area they are found in these two orders and- in the Orthoptera, where the convex veins are accessory and the concave veins intercalar>'. In paleozoic insects they were present in the Protodonata and Protoephemeroidea, and judging from some published figures, they were present in some of the Protoblattoidea. When it is desirable to refer to a particular intercalary vein it may be done by combining the initial I, indicating intercalary, with the designation of the area of the wing in which the intercalary vein occurs. For example, in the wings of most May-flies in which the venation is not reduced, there <^r R' /?. Fig. 63. — The wings of Prionoxyshis. is an intercalary vein between veins Cui and Cu2, i. e. in the area Cui. This intercalary vein is designated as ICui (Fig. 6i). When more than one intercalary vein exists in a single area of the wing, as in area Mi of Epeorus (Fig. 6i), and it is desired to designate them individually, it can be done by the use of appended letters. Thus, for example, the three intercalary veins in area Mi of Epeorus may be desig- 76 THE GENERAL FEATURES OF WINGS nated as IMia, IMib, and IMic, respectively, beginning with the oldest, i. e. the longest one. It will rarely be necessary, however, to resort to this somewhat cumbersome terminology; as it will be siifficient in most cases to indicate the number of distinct intercalary veins in a given area. The Adventitious veins. — -In certain insects there occur secondary veins that are neither accessory veins nor intercalary veins as defined above; these may be tenned adventitious veins. The more important of them are described in later chapters in the course of the descriptions of the wings in which they occur. Examples of them are the supplements of the wings of certain Odonata and the spurious vein of the Syrphidse. The anastomosis of veins. — The typical arrangement of the wing-veins is often modified by an anastomosis of adjacent veins; that is, two veins will come together at some point more or less remote from their extremities and merge into one for a greater or less distance, while their extremities remain separate. This is illustrated by Nemoura (Fig. 12) where tracheae ; 2345 e 7 Fig. 63. — Diagram illustrating the formation of regular cross-veins (After Needham). Sc2 and Ri extend for a short distance in the same vein cavity. In the fore wing of Prionoxystus (Fig. 62) there is an anastomosis of veins R3 and R4+.^- The cross-veins. — It has been shown that so far as the principal longitudinal veins are concerned the wings of all orders of insects are modifications of a single primitive type, and a diagram representing our conclusions regarding this type has been given on an earlier page (Fig. 6). In the discussion of our hypothetical type no reference is made to cross- veins ; for it is evident that a homology, similar to that which exists between the longitudinal veins of the several orders of insects does not exist between the cross-veins of these orders. It is shown in Chapter IV that in the wings of the more generalized members of the Pateodictyoptera, the order of fossil insects from which the orders of recent insects were evolved, there were no definite cross-veins, but instead an irregular networic of veins in the spaces between the longi- tudinal veins (see Fig. 8). A similar network exists to-day in the fore wings of the Acrididac, for example. The transformation of an irregular network of veins into regular transverse cross-veins is discussed by Needham ('03) and is illustrated by the accompanying diagram (Fig. 63). His conclusions were based on his THE GENERAL FEATURES OF WISGS 77 studies of the \vinTTTT77 Fig. 175. — Wings of Hemerobius humuli. character, that it is not probable that it will be abandoned; and no harm will be done by following it if the origin of the various "sectors" be kept in mind. 182 THE WINGS OF NEUROPTERA It has been shown, in the preceding section of this chapter, that in the Sympherobiidag the radius of the fore wings has two sectors ; but this result in that family has been brousjht about in a very different way than that Fig. 176. — Base of hind wing of Hemerobiiis huniuli. just described. In the one case there is a diyision of the stem of the radial sector; in the other, a suppression of this part of the sector, by its coales- cence with vein Ri. The separate origin of one or more branches of the radial sector occurs also in the wings of the Dilaridse described in the next section of this chapter, and in an Australian insect recently described by Tillyard ('16 p. 279), under the name Ithone fulva. In both of these cases there appears to have been a splitting back of the branches that arise separately, and not a suppression of the stem of the radial sector. This, however, is not per- fectly clear in the case of Ithone ftdva. An early stage in the suppression of the stem of the radial sector is shown in the hind wing of Hemerohius himiidi (Fig. 175), where vein R5 has split back nearly to the base of the radial sector, so that it separates from the remainder of the sector, i. e. vein R2+3+4 nearly opposite the anterior end of the first radio-medial cross-vein (Fig. 176). From the point of Fig. 177. — Trachcation of a hcmerobiid wing (After C. & N.). separation of veins R5 and R2+3+4, the latter extends obliquely forward and anastomoses with vein Ri thus forming a small cell opposite the cell closed by the first radio-medial cross-vein. The extending of the union of veins Ri and R2+3+4 from ^he point where they now anastomose towards THE WINGS OF HEMEROBIIDM 183 the base of the wing, so as to obHterate the small cell between them and also towards the apex of the wing for a certain distance would produce the condi- tion that exists in the fore wing. A stage intermediate between that of the radial sector in the hind wing oiHemerobius humiili and that of the fore wing of the same species, where the coalescence of the stem of the radial sector and vein Ri is carried so far that Fig. 178. — Wings of Megalonius via'stiis. both veins R5 and R4 arise separately, is presented by a hemerobiid pupal wing figured by Comstock and Needham (Fig. 177). Here only trachea R5 arises separately. In the more specialized members of the Hemerobiidas a large number of definitive accessory veins have been developed upon the radial sector or upon the coalesced veins Ri and Rsi this development of accessor}'- veins has progressed much farther in the fore wings than in the hind wings. The two pairs of wings differ also in the fact that in the hind wings the stem of the radial sector is not suppressed. 184 THE WINGS OF NEUROPTERA The wings of Hemerohius hmnuli (Fig. 175) are very instructive in that they show the beginning of the speciaHzation of the hemerobiid wings by the development of accessory veins. In the hind wing, the radial sector bears only the four primitive branches. In this enumeration the marginal accessory veins are not counted. In the fore wing, there is a single definitive accessory vein, vein Rqs- The wings of Megalomus mastiis (Fig. 178) will serve as an example of the more specialized hemerobiid wings. In the hind wing, veins Ri and Rg do not coalesce. As the radial sector is seven-branched, it is evident that three accessory veins have been developed upon it. In the fore wing, there is a complete coalescence of vein Ri and the stem of vein Rg, and a greater nimber of radial accessory veins have been developed. The tips of all of the branches of radius are forked in both fore and hind wings. This forking is especially marked in the case of vein R5, and has resulted in the formation of a radial cuneate area of considerable size (Fig. 178, red). In the more generalized Hemerohius hmnuli, the tip of vein R5 is not markedly more forked than are the tips of other branches of the radial sector (Fig. 175). Correlated with the prominent part played by tiie secondary cubital fork in the hemerobiid group of families, there is frequently a tendency for vein Cu2 to SLtvophj. In the hind wing of Megalomus mcestus (Fig. 178), the base of vein Cuo is preserved, but the greater part of this vein is repre- sented by a fold, indicated in the figure by a dotted line. In the hind wing of Sympherobius (Fig. 173), only the base of vein Cu2 is preserved, there is no vestige of the remaining portion. In this case the base of vein Cu2 is no longer curved, as is usually the case, but a part of it is aligned with a cubito- anal cross-vein. In the hind wing of Hemerohius humuli (Fig. 175) only that part of the base of vein Cu2 that is aligned with the cubito-anal cross- vein is preserved, so that there appears to be only a cross-vein in this position. In Figure 176 the two parts of this apparent cross-vein are lettered to indicate their homologies. (/) THE WINGS OF THE DILARID.^ This is a small family and one that is not well represented in collections. There is only one species known from North America, and this is exceedingly rare. I have before me wings of two species from Japan and figures of two other species. The following statement is based on this limited amount of material. Although the radius of the fore wings has from one to three sectors, the increase in the number of sectors appears to be due to a splitting back of one or two branches of vein R^ rather than to a coalescence of veins Ri and Rs, as is the case in the Hemerobiidaj. But as we do not know what was THE WINGS OF DILARIDM 185 the primitive position of the radial fork in this family we can not be sure that this is the case. In the Dilaridse the forking of the branches of the radial sector has progressed to a much greater extent than in the Hemerobiidae. These two features are well shown in Dilar uohirce (Fig. 179), from Yoshino, Japan, wings of which were kindly given to me by Mr. Waro Nakahara, the describer of the species. A remarkable feature of the wings of Dilar nohira: is the fact that the radius of the hind wings has two sectors. Thisisalso the case in the wings Fig. 179. — -Wings of Dilar nohircc. of Rexavius japonicns. In none of the HcmcrobiidcE is there more than one radial sector in the hind wings. The following remarkable series of variations in the structure of the radial sector exists in the Dilaridse. In Dilar americanus, as described by McLachlan, there is a single five- branched radial sector in both fore and hind wings. In Dilar turcius, as figured by Handlirsch ('06), the radius of the fore wings bears two sectors, the second of which is five-branched. The radius of the hind wings bears a single five-branched sector. In a species of Dilar figured by Brongniart ('93), the radius of the fore wings bears three sectors; the third sector is six-branched, making eight branches in all. The radius of the hind wings bears a single seven-branched sector. 186 THE WINGS OF NEUROPTERA In Dilar nohirce and in Rexavins japonicns the radius of both fore and hind wings bears two sectors. In all of these species the branches of the radial sector are so deeply- forked that it seems probable that accessory veins are being interpolated. Marginal dashes are present. {k) THE WINGS OF THE BEROTHID.^ The type species of the Berothidag is Berotha insolita, the wings of which are figured here (Fig. i8o). These wings present some striking features, most of which, however, are foreshadowed in various genera of the hemero- biid group of families. The wings of Berotha differ from those of the hemerobiids in that the first radio-medial cross-vein of the hind wings is transverse, not longitudinal and sigmoid, as it is in those families. trpnTTyri Fig. 1 80. — Wings of Berotha insolita. The most striking features of the wings of Berotha are the following: the absence of a recur\^ed costal vein; the atrophy of the terminal portion of the subcosta; the great reduction in the number of cross-veins, those of the radial area being reduced to a single series of gradate veins; the fan- like branching of the tips of veins; prominent marginal dots alternating with the twiglike branches of the veins; the loss of vein Cuo of the hind wings; the anastomosis of the first anal vein and vein Cui in the hind wings; and, in the hind wings, the extending of vein Cui and the first anal vein quite closely parallel with the inner margin of the wing, leaving THE WINGS OF BEROTHID^ 187 only a narrow area between these veins and the margin of the wing, which is largely occupied by the fanlike branched portion of the accessor}- veins. In Berotha the subcosta enters the pterostigma and then atrophies; the tip of the preserved portion is distinct from vein Ri ; the accessory branches of the atrophied portion are preser\^ed, but are free at the base. A short distance before the end of the preserved portion of the subcosta there is a cross-vein extending from the subcosta to vein Ri. In some allied genera veins Sc and Ri become united at this point so that they appear to coalesce. The loss of vein Cu2 of the hind wings occurs also in the hemerobiids, where all stages of the atrophy of this vein are to be found. There are several genera of Netiroptera that present the same type of wing-venation as that shown by the wings of Berotha, and which, for this reason I believe should be included in the family Berothidae. These are the following : Loniamyia, Spennophorella, Trichoma, and Stenobiella. The most available recognition characters of this family are presented by the hind wings, these are the following: the transverse course of the first radio-medial cross- vein; the absence of vein Cu^; the narrowness of the area of the wing between A-eins istA and Cui and the margin of the wing, this area being largely occupied by the fanlike tips of the accessory veins ; and a bend in the outer part of the first anal vein which results either in an anastomosis of this vein with vein Cui or in these two veins being closely approximate where they are joined by a cross-vein. The shape of the wings varies greatly in this family; in Berotha, Loniamyia, and TricJioma the wings are falcate; in Spermophorella they are oblong, with the terminal portion rounded; and in Stenobiella, they are ver\^ long and narrow. In Berotha the temiinal portion of the subcosta is atrophied and the tip of the preserved portion is free; in Lomamyia and SpermopJiorella, the tip of the subcosta joins vein Ri ; and in Trichoma and Stenobiella the subcosta ends in the margin of the wing. The representatives of this family are widely distributed. Berotha is found in India, Lomamyia is represented by two species in the United States, and the other three genera are Australian. (/) THE WINGS OF THE POLYSTCECHGTID.^ The wings of our covamon Polystoochotes punctatns (Fig. i8i) represent the type of wing-venation characteristic of the family Polystoechotidai. In these wings the humeral cross-vein is recurved and branched; veins Sc and Ri coalesce at the tip; the radial sector is pectinately branched; the number of cross-veins is greatly reduced; but there is in both fore and hind wings a very perfect series of gradate veins; and marginal dots are present, these are not represented in the figure. 188 THE WINGS OF NEUROPTERA In these wings the development of definitive accessory veins on the radial sector and the regularity of the border of marginal accessory veins have reached a very high degree of perfection. Cu- Fig. 1 8 1. —Wings of Polystcechotes ptinctatns. (m) THE WINGS OF THE PSYCHOPSID^ Among the remarkable types of wings that have been developed in the Neuroptera that of the Psychopsidas is one of the most striking. I have selected as an illustration of this type the wings of an undetennined psychopsid from West China (Fig. 182), as in this species the venation of the wings is not obscured by blotches of color as it is in the only species of Psychopsis in our collection. The more striking features of these wings are their rounded form; the unusual width of the costal area throughout its length; a peculiar nexus of the tips of veins Sc, Ri, and Rg; the presence of two series of gradate veins; the irregular distribution of the radial cross-veins in both wings and of the cross-veins in the svibcostal area of the fore wings; and the unusual length of the stalk of media, especially in the hind wings. The branches of the radial sector are very regular except that one in the fore wing, the sixth, is split nearly to its base, indicating that an accessory vein is being interpolated. Nearly all of the veins are forked at the tip; and marginal dots are present. THE WINGS OF CHRYSOPIDM 189 Fig. 182. — Wings of a psychopsid. (n) THE WINGS OF THE CHRYSOPID^E The wings of the Chrysopidte are characterized by a very remarkable and distinctive type of specialization, the details of which can be under- stood only by a study of the tracheation of the wings. A brief account of the wings of Chrysopa plorabunda showing the essential features of this type of specialization was published by McClendon ('06) ; and, quite recently, a much more extended one, based on an Austra- lian species, Chrysopa signata, has been ptxblished by Tillyard ('16). So that now the structure of these wings is well understood. In the preparation of the present account I have made use of the papers just cited, and I have been materially aided also by Mr. R. C. Smith, who during the year previous to the appearance of Tillyard's paper devoted himself to the study of this problem. Mr. Smith has bred many indi\'iduals 190 THE WINGS OF NEUROPTERA of several species of Chrysopa, and has kindly furnished me with the material upon which the accompan^dng figures were based. A superficial examination of a wing of an adult Chrysopa (Fig. 183) reveals the presence of two longitudinal veins between the radial sector and Fig. 183. — Fore wing of Chrysopa nigricornis; M', pseudo-media; Cu', pseudo-cubitus. the inner margin of the wing, one of which appears to be the media and the other vein Cui ; but each of these, as is shown later, is a serial vein composed of sections of several veins. As it would be impracticable to apply to these veins names indicating their composition they have been designated by Tillyard ('16) as the pseudo-media and the pseudo-cubitus respectively; and he makes use of the notation M' for the pseudo-media, and Cu' for the pseudo-cubitus. This Fig. 184. — Trachcation of the wings of a i)upa of Chrysopa nigricornis. THE WINGS OF CHRYSOPID^ 191 designation of these two veins is a happy one and I gladly adopt it with one slight change. As the pseudo-cubitus really appears to be vein Cui, I make use of the notation Cu/ instead of Cu'. ^'*-4Afj +- ^5 /?4 Ri ^" Fig. 185. — Diagram of the wings of Chrysopa nigricornis showing coalesced veins slightly separated. Fig. 186. — Wings of Chrysopa nigricornis. An examination of the tracheation of the wings of a pupa of Chrysopa reveals the nature of the two serial \'cins pseudo-media and pseudo-cubitus. 192 THE WINGS OF NEUROPTERA Figure 184 is a reproduction of camera lucida drawings of the pupal wings of Chrysopa nigricornis made by Mr. R. C. Smith. In order to show more definitely the composition of the two serial veins a diagram of an adult wing is given (Fig. 185) in which the elements of the coalesced veins are represented slightly separated, and the cross-veins connecting the coalesced veins are represented by dotted lines. By com- paring this diagram with Figure 186 the homologies of the different veins can be recognized. {0) THE WINGS OF THE GSMYLID.^ The Osmylidffi is a moderately large family and is composed of very beautiful insects. The wings of Osmylus hyalinahis (Fig. 187) can be taken as an illustration of the type of wings characteristic of this family. There Fig. 187. — Wings of Osmylus hyalindlus. is no difficulty in recognizing the identity of the veins in these wings; but the more important ones are lettered in Figure 188. The subcosta and vein Ri are closely parallel and are somewhat stouter than the other veins; they constitute the supporting axis of the wing. These two veins coalesce at the tip. The humeral vein is not markedly recurved and is not branched. The costal area is broad and the accessory THE WINGS OF MYIODACTYLID.E 193 veins of the subcosta are forked, and sometimes they are connected by cross-veins. Many cross-veins are present in the other parts of the wing, excepting the subcostal area in which they are lacking. In the outer third of the wing the cross-veins are reduced in number, consisting chiefly of a series of gradate veins. Many definitive accessory veins have been developed upon the radius and upon the veins back of this vein ; the tips of most of these are forked. The marginal accessory veins form a quite regular border. The media is two-branched in both wings. In the fore wing the medial fork is a little beyond the first fork of the radial sector in the specimen figured, but it varies slightly in different individuals. In the hind wing the medial fork is quite near the base of the wing; it is not, however, quite so near the base of the wing as it appears to be in this figure; for, owing to a fold in the wing, the base of media is concealed by the base of radius, which overlaps it. A result of this overlapping of the base of media by the base of radius is that the first radio-medial cross- vein appears to join the radius instead of joining media, as it really does. The cubital fork is near the base of the wing in both fore and hind wings ; and the anal veins are much branched. Fig. 1 88. — Base of fore wing shown in Figure 187 enlarged. {p) THE WINGS OF THE MYIODACTVLID.'E The type of this family is Myiodactyhis osmyloides, an Australian species described by Fr. Brauer in 1 866 . This species I have not seen ; but through the kindness of Mr. Nathan Banks I have been able to study a closely allied species, Myiodactyhis pubesceiis, the wings of which I figure here (Fig. 189). This species is from Port Darwin, Australia. 194 THE WINGS OF NEUROPTERA The wings of Myiodactyltis resemble those of Osmyhis (Fig. 187) in their general appearance and in most of their structural details. The most important differences are that in Myiodactyliis there are many cross-veins in Fig. 189. — Wings of Myiodactyliis pubescens. the subcostal area, the media of the fore wings is not forked, and vein Cu2 of the hind wings is lost. The most striking features of the wings of Myiodactylus are a broad border of accessory veins that extends completely about the wing and which is free from cross-veins and a discal area which is abundantly supplied with cross-veins. In this respect the wings resemble those of Osmyhis (Fig. 187); but the border is more complete than in Osniylus, due to the fact that vein Sc-1-Ri is curved back so that it does not interrupt the border, being con- tinued to the margin of the wing only by accessory veins. The humeral vein of the fore wing is forked but it is not of the recurved type. A series of gradate veins extends from the tip of vein Sc+Ri to the tip of media; this is quite similar to the condition in Osmylus, except that in Osmylus vein Sc + Ri extends beyond the end of the series of gradate veins. In the fore wings, media extends free from radius to the base of the wing; it is not forked, and there is no oblique vein representing vein M3+4. The cubital fork is near the base of the wing, and there are three separate anal veins. THE WINGS OF NYMPHIDM 195 The hind winj^^s differ from the fore wings in that media is distinctly two-branched, and it appears to arise from radius, at least the two are so closely oppressed that I can see no space between them; the cubitus is greatly reduced in length and is not forked, probably vein Cu2 has atrophied; and the base of the first anal vein is vestigial. It is quite possible that the forked vein lettered ist A in the figure is vein Cu-j -f ist A. With only a single specimen before me I am unable to decide this point. In the fore wing some of the cross-veins are much more prominent than the others; this is largely due to their being colored brown, but they are also slightly stouter. {q) THE WINGS OF THE NYMPHIDM The Nymphidaj is a small family of insects that is restricted to Australia. I have been able to study only a single representative of it ; this is Nymphes myrmeleonides . The wings of this species are represented in the accom- panying figure (Fig. 190). These wings are long and narrow and are furnished with many cross- veins ; they possess also many definitive accessory veins and many marginal accessory veins; the latter are very numerous and regular in the apical portion of the wings and add greatly to their beauty. In general form and appearance the wings of Nymphes resemble those of the Myrmeleonidas ; but they are sharply distinguished from myrmeleonid Fig. 190. — Wings of Nymphes myrmeleonides. See Figure 191 for the labeling of the veins of the fore wing. wings by the following characteristics : first, the subcostal area is traversed by many cross-veins, while this area is free from cross-veins in the Myrme- Iconidte; second, the media of the fore wings is ob\dously two-branched in Nymphes; in the M\mnelconida3 it is apparently, though not actually, unbranched. 196 THE WINGS OF NEUROPTERA In the fore wings of Nymphes (Fig. 191), the radial fork is quite near the base of the wing; there is a fairly well developed secondary cubital fork; and the primary cubital fork is near the base of the wing ; vein Cuo is easily recognized; it branches off from vein Cu nearly opposite the first medio- Fig. 191. — Base of fore wing of Nymphes myrmeleonides. cubital cross-vein, and the basal part makes a sweeping curve; it touches the first anal vein at the end of this curved portion, and then extends free from it. There are three anal veins; vein ist A is forked. In the hind wings (Fig. 190), media is two-branched, a fact that has not been understood, vein M3+4 having been mistaken for vein Cu. This mistake was due to two facts : first, the medial fork is very near the base of the wing and has been overlooked; second, vein M3+4 bears prominent accessory veins which cause it to closely resemble vein Cui of the fore wing. The cubital fork, as in the fore wing, is very near the base of the wing ; and the basal part of vein Cuo is transverse, appearing like a cross-vein except that it is much stouter ; it resembles in this respect this vein in many of the Myrmeleonidae. The .anal veins are similar to those of the fore wing. Marginal dots are present in the wings of Nymphes; this seems remark- able as they are absent in the other families of the myrmeleonid group ; the marginal dots are not represented in the figures given here. ir) THE WINGS OF THE MYRMELEONID^ In the Myrmeleonidae the wings are long and nan-ow and delicate in structure; they are furnished with many accessory veins, both definitive and marginal, and with very many cross-veins (Fig. 192) ; the radial sector is pectinately branched. Although the wings of myrmeleonids have been much studied no one has published as yet a coirect and complete determination of the homologies THE WINGS OF MYRMELEONID.^ 197 of the principal wing- veins of these insects; but I am convinced that at last this problem is solved; and now that it is solved, it seems a comparatively simple one, so simple that there can be no doubt as to the correctness of the conclusions. The causes of the earlier and incorrect conclusions are indicated in the following discussion. The determination of the identity of the costa, subcosta, radius, and the radial sector presents no difficulties, and no change in the commonly accepted view regarding these veins is necessary. This is true of these veins in both the fore and hind wings. With regard to the other principal veins it is necessar}^ to discuss the fore and hind wings separately Media oj the fore wings. — In the fore wings media appears to be a single unbranched vein; but it has been demonstrated, by a study of the trachea- tion of the wings of pupae, that what appears to be an oblique cross-vein and which is lettered o in the figures, is a branch of media, and consequently that media of the fore wings of ant-lions is two-branched, as it is in most other Neuroptera. The significance of the oblique vein in the fore wings of myrmeleonids was first discovered, twenty years ago, by Dr. Needham and myself, by a study of pupal wings, when we were collecting material for our series of articles on the wings of insects. We did not include our myrmeleonid material in that series of articles, as it was not needed to illustrate the fundamental principles that we were discussing; but I made use of it a little later in a laboratory manual;* and Figure 194 is based on a photo- micrograph that we made at that time. Recently Tillyard has studied the Fi<^. 192. — Wings of Myrmdeon, sp. From Katihar, British India. tracheation of the wings of myrmeleonids and discovered independently that the media of the fore wings is two-branched (Tillyard 1916, p. 739). *The Elements of Insect Anatomy by John Henry Comstock and Vernon L. Kellogg, Third Edition 1901. On page 116 of this work I indicated the significance of this bUque vein. 198 THE WINGS OF NEUROPTERA During the past year Dr. Needham and I have reviewed our study of the wings of ni}Tmeleonid pupae by the examination of fresh material and Figure 193 is based on one of our recent photomicrographs. In no pupal myrmeleonid wing examined by us, and we have examined many, were we able to detect the costal trachea. The subcostal and radial trachege were prominent in all, so too were the accessory tracheae of the subcosta. In some of the forming cross-veins there were distinct tracheae but in most cases the tracheae in the cross-veins were so indistinct that we are not warranted in representing them in the figure. Returning to a study of the medial trachea, it can be seen by an exami- nation of Figure 193 that this trachea is two-branched; that trachea Mi +2 is a direct continuation of the main stem of this trachea and that trachea M3+4 first traverses the oblique vein and then after making a sharp bend extends lengthwise the wing in line with the basal part of trachea Cui. From this it can be seen that the vein which in the adult wing appears to Fig. 193. — Tracheation of a fore wing of a pupa of a myrmeleonid. be vein Cui is really a serial vein consisting in part of vein Cui and in part of vein Ms +4; to this serial vein Tillyard has appHed the designation cubitomedian and makes use of the notation Cui + M2. I see no objection to the application of the term cubitomedian to this vein; but as the nota- tion Cui -f M2 indicates an ordinary coalescence of veins side by side and not the formation of a serial vein, I regard the notation Cui & M3+4 as more appropriate. It will be noted that Tillyard designates the branch of media that unites with vein Cui in this case as Mo. This is an unfortunate failure to apply the uniform terminology; and there are several other authors who are guilty of the same error in similar cases. The two branches of media formed by the first forking of this vein are veins M1+2 and M3+4; it is obvious that these branches are homologous with those thus designated when media preserves its typical four-branched form. If one designates the two branches of media as veins Mi and M2 the question what has become of veins M3 and M4 is suggested. In the wing represented by Figure 193 the trachea M3+4 and Cui were distinct although they came very near together at the end of the oblique vein and the veins formed about them anastomose at this point. In the THE WINGS OF MYRMELEONIDM 199 wing represented by Figure 194 a somewhat different condition exists; in this wing trachea Cui and M3+4 extend closely parallel in the same vein Fig. 194. — Detail of a pupal wing of a myrmelconid showing union of tracheae M3+4 and Cui. cavity. In this case the serial vein consists of three differing sections; the basal part of this vein extending from the cubital fork to the point where it Fig. 195. — Tracheation of the wings of a pupa of a myrmelconid. is joined by the oblique vein is simply vein Cui ; then follows a section in which this vein is compound, consisting of veins Cui and ^U+i united; 200 THE WINGS OF NEUROPTERA and finally, there is a section consisting simply of vein M3+4- Figure 195 represents in the fore wing a condition similar to that shown in Figure 194. Fig. 196.- — Base of a fore wing of Tomatares clavicornis. In the wings represented by Figure 195 the tracheation of the anal area was obscured in both wings. Cubitus oj the fore wings. — In all of the myrmeleonids that I have studied the cubital fork is very near the base of the wing and in many of them vein Cu2 is greatly reduced. In the wing represented by Figure 193, trachea Cu2 is quite prominent and extends nearly one-third of the length Fig. 197. — Base of a fore wing of Balaga micans. of the wing. I reared several pupae of this kind from larva; obtained at Gulf port, Florida; but I do not know the adult of this species. The follow- ing series of figures will serve to illustrate different stages in the reduction of vein Cu2. THE WINGS OF MYRMELEONIDM 201 In the fore wing of Tomatares clavicornis (Fig. 196), vein Cuo is still a well-developed vein and is free from the first anal vein, except that the two are connected by cross-veins. In the fore wing of Balaga micans (Fig. 197), vein Cu^ is reduced to a vestige, which extends closely parallel with the first anal vein for a short distance and then the two coalesce. In the fore wing of Brachynenmrus longipalpus (Fig. 198), only the transverse basal part of vein Cu2 is distinct from the first anal vein ; but this transverse part is still curved. The almost complete coalescence of vein Cu2 with the first anal vein in the greater number of the Myrmeleonidae, as in Brachynenmrus, is doubt- Fig. 198.- -Base of the wings of Brachynemunis longipalpus. less the reason that this vein has been overlooked by writers on this family. There is also another fact that has led to the overlooking of vein Cu2; in most of the M\T-meleonidffi vein Cui of the fore wings bears a prominent accessory vein, vein Cuia (Fig- i97)> which has been mistaken for vein Cu2. Let us now pass to an examination of the hind wings. Media of the hind wings. — Up to the present time the belief that media of the hind wings of mynncleonids is reduced to an tmbranched condition has been undisputed ;' and an explanation of the way in which this condi- tion has come about has been eagerly sought; hundreds of m^mneleonid wings have been examined in the hope of finding an oblique vein in the hind wing like that of the fore wing, but without success; and an examina- tion of pupal wings seemed to throw no light on the subject. This was the 202 THE WINGS OF NEUROPTERA state of our knowledge of the subject at a time when this volume was about to be sent to the printer, when I made a reexamination of a series of myrmeleonid wings which revealed the explanation of the problem. This explanation is as astonishing as it was unexpected. Media of the hind wings in these insects is two-branched and there is no anastomosis of vein Ms +4 with cubitus as there is in the fore wing. The medial fork is very near to the base of the wing; and what has been believed to be the cubitus is vein M3+4. I had observed the medial fork but had regarded it as the result of the coalescence of veins M and Cu; and had explained in the same way the forking of the medial trachea which I had observ^ed in the hind wings of pup« (Fig. 195). Fig. 199. — Hind wing of Symmatheles contrarins . In wings mounted for study, as they are commonly mounted, the radial fork is concealed in many cases by an overlapping of it by the radius, due to a fold in the wing; but with a little care it can be easily seen. In the hind wing of Brachyueniurus represented by Figure 198 the medial fork is overlapped by the radius; the figure is a reproduction of a photograph and it represents the veins as they appear when seen from above. Figure 199 represents the base of a hind wing of Symmathetes contrarins; in this wing the forked condition of the media is obvious. Cubitus of the hind wings. — A study of the cubitus of the hind wings confirms the conclusion regarding media given above, as in certain fonns one finds a typical cubitus which is distinct from the vein that has been believed to be the cubitus and which is really vein M3+4. In the hind wings of most myrmeleonids, as in the fore wings, vein Cu2 is greatly reduced, but a series illustrating different degrees of this reduction can be easily found. THE WINGS OF MYRMELEONID.E 203 In the hind wing of Symmatheies conirariiis (Fig. 199), vein Cuo is well- preserved, being about one-half as long as vein Ctti. There is also in this wing a prominent vein Cui^. In the hind wing of Brachynemiirus longipalpus (Fig. 198), veins Cuo and I St A anastomose and the basal part of vein Cuo appear like an un- usually stout cross-vein. In concluding this discussion of the veins M and Cu of the myrmel- eonids, I wish to call attention to a ver\^ interesting feattire of the wings of these insects, which has doubtless been an important factor in delaying the determination of the homologies of the wing-veins. Examine the figure of the wings of Brachynemurus longipalpus (Fig. 198) and note the similarity in structure of the fore and hind wings. If we omit in each case a study of the base of the wijig, the venation of the two wings appear to be almost identical except that in the fore wing there is an oblique vein. Behind the radius and the radial sector there is in each case an apparently unbranched vein, formerly regarded as an unbranched media; behind this simple vein there is a forked vein, formerly regarded in each case as the cubitus. It is not strange that the similarity in form of these two veins should have led to the belief that they are homologous. But it has been shown above that the forked vein in the fore wing is vein Cui and the forked vein in the hind wing is vein M3+4. There has been developed in these two wings a similar bracing of each by the use of very different material in the two wings. Generalizations and definitions of special terms. — In the descriptions of wings in which peculiar methods of specialization have arisen, it becomes necessary to make use of special terms in referring to the resulting struc- tures. Several such terms have been proposed by Dr. Needham, who has prepared an extensive monograph on the wing-venation of the Myrmeleon- idre. As the publication of this monograph has been delayed it seems desirable to introduce these terms here, so that they may become available. I also include some generalizations kindly handed to me by Dr. Needham. MvRMEi.EOMu Venation- "The more striking cliaraeters of the venation of the Myrmeleonidae are: "i. The apical fusion of veins Sc and Ri and the clearing out of all cross-veins from the inclosed space. "2. The development behind the ])oinl of fusion of an elongate truss cell of variable form but constant position. "3. The extensiv^e development of branches upon the radial sector, pectinately arranged, and diminishing in length from the base outward. The basal branch is much more extensively forked and more deeply forked at its tip than are any of the others. In the subfamily Palparina: and in a few scattering genera {Myrmecalurus, etc.) that portion of the radial sector distal to the base of the truss cell, becomes switched upon a cross-vein, and attached to Vein R\, so as to stimulate a second sector [see Figure 200I. 204 THE WINGS OF NEUROPTERA "4. The reduction of media in the fore wing to an apparently simple vein, the base of M3+4, appearing as an oblique cross-vein joining Cui. When this ceases to be oblique, the forking of media is entirely obscured. "5. The development of a triangular brace across the basal third of both wings, similar in form but unlike in composition. This is a striking example of parallelism in vein development. This brace is formed about a strong secondary fork in a principal vein. The vein is Cui in the fore wing, M3-I-4 in the hind wing. Beyond the fork this vein in the fore wing fuses with M3-|-5i, . Fig. 244. — Wings of a nymph of Nemoiira (From C. & N.). at least, these trachese do not arise from the same trunks as do those tracheae that precede the wing veins.* (c) THE CLASSIFICATION OF THE PLECOPTERA As it is necessary in the following discussion of the wings of the Ple- coptera to refer to the divisions of the order, and as the classification adopted here has not been included in any of the text books, I give a brief outline of it. It is the classification proposed by Dr. Gtinther Enderlein ('09). Enderlein divides the order Plecoptera into two suborders, the Holog- natha and the Systellognatha. In the former he includes three families; in the latter, two, as shown by the following table : r Fam. Gripopterygidas Suborder Holognatha \ Fam. Capniidee I Fam. Nemouridas / Fam. Pteronarcidae I Fam. Perlidas Suborder Systellognatha *I have been greatly aided in my study of the trachcation of the wings of the Plecop- tera by the opportunity of examining an extended .series of jihotographs and sketches of the wings of nymphs made by Miss Lucy W. Smith, which she kindly jjlaced at my disposal and which supplement in several important particulars those made by Dr. Needham and myself. THE WINGS OF PLECOPTERA 245 The family Gripopterygidae is believed to include the most generalized of living Plecoptera, on account of their strongly developed mandibles ; and the Capniidas and Nemouridas are associated with this family because in these families also the mandibles are well preser\'ed. On the other hand, in the Pteronarcidae and Perlidae only vestiges of mandibles exist. It may be that farther studies will show that this is not a natural division of the order; but with the data at hand it seems to be the best grouping of the families yet suggested. The family Gripopterygidae is restricted to the tropics and to the Southern Hemisphere; the other four families are well represented in the United States and Canada. In m}' studies of the wings I have been unable to find any characters indicating a dichotomous division of the order; each of the methods of specialization observed exist in both the Holognatha and the Systellognatha. id) THE SPECI.\L FEATURES OF THE WINGS OF THE PLECOPTERA One of the more striking features of the venation of the wings of the Plecoptera is a lack of uniformity in the number and courses of the sub- 2d A Fig. 245. — Wings of Isogenus sp. ordinate veins. Not only are striking differences in wing-venation to be observed between different members of the same species, but frequently the wings of the two sides of an individual will vary greatly in venation. 246 THE WINGS OF PLECOPTERA This is especially true as to the number of cross-veins and the branching of the veins in the distal parts of the wings. On the other hand, the characters presented by the trunks of the principal veins are quite constant. There is one characteristic of the wings of the Plecoptera that is so constant that it may be considered an ordinal character. This is the fact that in the wings of the adult the radial sector of the hind wings is attached to media instead of to radius (Fig. 245). This switching of the radial sector of the hind wings is true only of the venation of the adult. In the wings of nymphs the trachea Rg is a branch of trachea R (Fig. 244).* There are certain features of the wings of Plecoptera, which although not always constant, occur in so large a proportion of the members of the order that they may be considered characteristic; these are the following, all of which are represented in Figure 245 : The presence of the radial cross-vein (Fig. 245, r). The absence of cross-veins in cell R and in the basal part of area Ri. Cross-veins are found in cell R in Pteronarcys. The strengthening in the fore wing of the area between media and vein Cui and of that between veins Cui and Cu2 by the development of many cross- veins. The reduction of media to a two-branched condition. The reduction of the radial sector to a two-branched condition. This fact is apparent only after an extended study of the wings of stone-flies. In many cases, of which the form represented by Figure 245 is one, acces- sory veins have been developed on vein R2+3, which appear to be the primitive branches of the radial sector. But these accessory veins are very inconstant in number and position. I am convinced that only the first forking of the radial sector, the division of it into veins R2+3 and R4+5 has been retained from the stem form of the order. The farther branching of either of these veins is too erratic to be considered primitive, frequently differing in the wings of opposite sides of an individual insect. The unbranched condition of the first anal vein in both fore and hind wings; rarely, as in some species of Pteronarcys, the first anal vein is branched at the tip. In the unbranched condition of the first anal vein the Plecoptera resemble the Orthoptera. But the Plecoptera differ from the Orthoptera in that in the expansion of the venation of the anal area only accessory veins are developed; intercalary veins are not found in the Plecoptera; although in the Australian genus Eusthenia (Fig. 246) there are what appear to be *In spite of this fact, Klapalek ('12) states: "Im Hinterfliigcl finden wir in dem reifen Fliigel kcinen Sektor radii, dafiir ist die Media zweimal gegabelt." In carrying out this idea he incorrectly labels the branches of the radial sector and of media in his figure of the wing- venation. This is a remarkable example of a lack of appreciation of the value of ontogenetic facts. THE WINGS OF PLECOPTERA 247 the first stages in the development of intercalary veins, in the anal area of the hind wings. In concluding this brief summary of the special features of the wing of the Plecoptera it seems desirable to define some terms freqtiently used by writers on this order. The transverse cord. — In many genera of this order there is a nearly continuous series of cross-veins extending across eacli wing just beyond the Sci SCi . Fig. 246. — Wings of Eusthenia specatbilis. middle of its length ; this series of cross- veins is termed the anastomosis by writers on the Plecoptera. As it is not formed by an anastomosing of veins the use of the term, transverse cord, defined in chapter III, is preferable. The pterostigma. — In most members of this order a specialized ptero- stigma has not been developed ; but the term pterostigma is commonly applied to the cell beyond the end of the subcosta and between the costa and vein Ri, even though it is of the same color and texture as the remainder of the wing. The basal anal cell. — A very constant feature of the anal area of the wings of Plecoptera is the presence of a cross-vein near the base of the wing, which extends from the first anal vein to the second. The cell that is closed by this cross-vein is termed the basal anal cell (Fig. 245, ha). 248 THE WINGS OF PLECOPTERA Enderlein has introduced a new terminology of the veins of the anal area, which I thinlc is unfortunate, as it tends to confusion. The veins of this area of all orders of insects have long been known as the "anal veins." When Comstock and Needham deterinined that these veins were typically three in number, we designated them as the ^^^5^ anal (ist A), the second anal {2d A), and the third anal {jd A) respectively. This preserved the term anal, b}^ which they were generally known, and at the same time afforded a specific designation for each. Enderlein restricts the term anal (an) to the first anal vein; and desig- nates the second anal vein the axillary (ax), and the third anal vein the accessory (acce) . He numbers the branches of the axillary vein from before backward, and those of the accessory vein in the opposite direction. (Zool. Anz. Vol. 28 p. 810). (e) THE PRIMITIVE PLECOPTEROUS TYPE OF WINGS The logical method of determining the various ways in which the wings of the different members of an order of insects have been specialized is to discover what is the most generalized type of wing-venation to be found within the order and then to trace the various ways in which this type has been modified in the different divisions of the order. This method has been followed in the preparation of several of the chapters of this book. In carrying out this method the t3rpe of wing-venation that most closely resembles the hypothetical primitive type (Fig. 6, p. 16) is considered the most generalized and is used as a starting point from which to trace the various methods of specialization found within the order. In my efforts to apply this method to the study of the wings of the Plecoptera I have found what, at first, was a most perplexing aiTay of facts. Forms were found that appeared to possess a generalized type of wing- venation. But when these forms were compared with closely allied forms, other species of the same genus for example, it was found that certain of the supposed generalized features were of little taxonomic value.* The lack of stability of the wing-venation is most apparent in the number and distribution of the cross-veins and in the number and courses of the branches of the longitudinal veins near the margin of the wings. This is especially true of the secondary branches of the radial sector and *The lack of stability of the wing-venation of stone-flies is apparent to any one that studies the order carefully. Ris ('96) in writing of the genus Dictyopteryx, after referring to the fact that it is practically impossible to find two specimens of the same species with identical wing-venation, and that seldom individuals are found that show the same network of veins in the wings of the two sides, remarks: "Es sicht so aus, als ob sich diese Thiere fur die wcnigen Tage, die sic als Imago zu leben haben, und fur den geringen Gebrauch, den sic von ihren Flugeln machen, gar nicht den Luxus eines streng gesetzmassig ausgebildeten Aderwerks gestatten konnten." THE WINGS OF PLECOPTERA 249 the secondary branches of media. In this region of the wings there are frequently striking differences between the wings of the two sides of individual insects. I am convinced, as already stated in the discussion of the special features of the wings of the Plecoptera, that only the first forking of the radial sector, the division of this vein into veins R2+3 and R4+5, is primi- tive; and that in those cases where the radial sector is more than two- branched, the additional branches have been developed secondarily. It is also evident that only the first forking of media, the division of this vein into veins M1+2 and M3+4, is primitive, for the farther branching of these veins is too inconstant and erratic to be considered primitive. The result of these conclusions is that forms that have only a two- branched radial sector and a two-branched media, like A^eniozira for example, are believed to resemble more closely what must be regarded as the primi- tive type of the order than do those in which these veins are more than two- branched, even though among them individuals may be found that much more closely resemble the hypothetical primitive type. In other words the known facts indicate that the Plecoptera have been evolved from a form in which the radial sector and the media where each only two-branched. In some of the descendants of this primitive stone-fly the reduction of these veins has been carried still farther, in others, additional branches have been developed upon these veins, but in a very erratic manner. As yet paleontology affords little data as to the form of the primitive members of this order, but what data we have tends to confirm the con- clusions stated above. I do not, however, place much weight on the paleontological evidence; for so few ancient Plecoptera are known that we can not claim to have an adequate conception of the ancient plecopterous fauna. Remains of two nymphs and one adult have been found in Jurassic deposits; and this is all the data that we have from times preceding the Te^tiar3^ The adult from the Jurassic is Mesonemottra Maaki Brauer. The fore wing of this species is quite well -presented and is quite similar to a wing of the recent genus Nemoura. Less than a score of species are known from the Tertiary; all of these belong to existing genera, and the greater nttmber of them belong to the Nemouridae ; the other species are placed in the genus Perla. Returning to an examination of recent forms, from a study of which we must draw our conclusions, we find that the type of wing that I have concluded represents best the primitive plecopterous wing exists practically unmodified in one or both pairs of wings of representatives of four of the five families of the order; and in the fifth family, the Pteronarcidae, where extended specialization by addition has taken place, the modification of it is comparatively slight. 250 THE WINGS OF PLECOPTERA The important features in which the assvuned plecopterous type of wing differ from the hypothetical primitive type are the reduction of the radial ^ Rz^i Fig. 247. — Wings of Gripopteryx tessellata (From Enderlein '05, but with some changes in the labehng of veins). sector to a two-branched condition and a similar reduction of media to a two-branched condition. The following examples illustrate these features in each of the five families. The Gripopterygidae. — Figure 247 represents the venation of the wings of Gripopteryx tessellata as shown by Enderlein ('05). In the fore wing of Cu2 Cu, Fig. 248. — Wings of Nemoura. this Species the radial sector and media are each clearly two-branched; but the branches of the radial sector coalesce nearly to the apex of the wing. THE WINGS OF PLECOPTERA 251 In certain closely allied forms, the coalescence is complete with the result that the radial sector is not branched. In the hind wing of this species the homology of the veins is somewhat obscured by an anastomosis of veins Ms +4 and Cui. In Eusihenia speciahilis (Fig. 246), the most "richly veined" member of the family that I have seen, the two-branched condition of both the radial sector and of media of both fore and hind wings is perfectly preser\^ed. The Nemouridae. — In Nemoiira (Fig. 248) the plecopterous type is per- fectly prescr^•cd in both fore and hind wings. A detailed discussion of the wings of Nemotira follows on a later page. The Perlidae. — In the fore wing of Chlorcperla sp. (Fig. 249), the radial sector and the media are clearly only two-branched; the same is true of the hind wing except that the homology of the veins is somewhat obscured, as in Gripopteryx, by an anastomosis of veins M3+4 and Cui. /?4+5 2d A Fig. 249. — -Wings of Chloroperla sp. The Capniidae. — In the Capniidae there appears to be a degeneration of the wing-venation, which has resulted in a remarkable lack of constancy in the courses of the veins, even within the limits of a single species. But it is easy to find examples, at least of fore wings, in which the plecopterous type of venation is fairly well preserved (Fig. 250). 252 THE WINGS OF PLECOPTERA Fig. 250. — Wing of Capnia. The Pteronarcidae. — Of the three known genera representing this family, it is evident that so far as the venation of the wings is concerned the genus Pteronarcella is the most generahzed.* In Pteronarcella hadia (Fig. 251) the plecopterous type of wing- venation is very shghtly modified. Media is clearly two-branched in both fore and hind wings. The first branch of the radial sector, vein R2+3, bears a few branches; but here, as elsewhere in the order, a comparison of allied species, or even of different individuals of the same species shows that these branches can not be considered pri- mitive. The data given above show that in tracing the methods of specialization of the wing-veins in the Plecoptera we should start with a form in which the radial sector and the media are each only two- branched. I have selected Nenionra,\he better known representative of the more generalized suborder, as the form with which to make comparisons. The tracheation of the wings of a nymph of Nemoura was figured by Comstock and Needham (Fig. 252). This represents very closely the hypothetical primitive type of tracheation except that the trachea of the radial sector and of media in both fore and hind wings are only two- Wings of Pteronarcella hadia. *I have seen neither specimens nor figures of Diamphipnoa, which occurs in Chile. But as this is said to resemble Pleronarcys in having cross-veins in cell R of the fore wings, I feel warranted in making the above statement. THE WINGS OF PLECOPTERA 253 branched. But from the facts given above, this condition of the radial sector and of media may be considered as typical of the primitive plecop- terous wings. In Nemoiira Comstock and Needham failed to find a costal trachea; but as we found a well-developed costal trachea in the closely allied genus Fig. 252. — Wings of a nymph of Nemoura (From C. &■ N.). Tceniopteryx, we may assume that this trachea was present in the primitive type of the order. The more important features of the primitive type of tracheation of the wings of the Plecoptera may be described as follows: The absence of tracheee in the cross-veins. The absence of a basal transverse trachea, and correlated with this, the origin of the medial trachea from the costo-radial group of tracheae. A typical two-branched subcostal trachea without accessory tracheae. The trachea Sci extends to the margin of the wing; the trachea Sc2 extends towards trachea Ri until it nearly reaches it, and then curving away from this trachea extends to the margin of the wing. A typical radial trachea except that its sector is only two-branched. It should be noted that the radial sector trachea of the hind wing, like that of the fore wing, springs from the main stem of the radial trachea. But the switching of the radial sector of the hind wing to media is foreshadowed by the course of the forming vein (Fig. 252). A typical medial trachea except that it is only two-1) ranched. 254 . THE WINGS OF PLECOPTERA A typical two-branched cubital trachea. Three unbranched anal trachece in the fore wing. Three anal tracheas in the hind wing; the first anal trachea is un- branched; the second and third anal tracheae are each two-branched. The characteristic features of the venation of the preanal area of the wings of the adult Nemoura (Fig. 248) need not be described in detail, as the veins follow quite closely the courses of the tracheas that precede them, which have been described above. There is an anastomosis of veins Sco and Ri in the region where the trachea Sco was closely parallel with trachea Ri; and the radial sector of the hind wing has been switched to media. There are certain features of the venation of the anal area of the fore wing that merit special attention. In Nemotira, and in fact in the greater number of the genera of this order, the anal area of the fore wings contains three, and only three, unbranched anal veins. The first and second anal veins are connected by a cross-vein, near the base of the wing, which closes the basal anal cell. The third anal vein either coalesces with the second at the base of the wing; or, when the two veins are separate at the base, they anastomose opposite the basal anal cell.* (/) THE METHODS OF SPECIALIZATION OF THE WHSTOS OF THE PLECOPTERA If we accept the conclusion that the type of wings represented by Nemoura can be taken as illustrating the primitive plecopterous wings, there follows the conclusion that within this order the specialization of the wings has proceeded in opposite directions in different members of the order. In some there has been a farther reduction of the wing-venation; in others, a specialization by addition. An equally remarkable fact is that differences in the direction of the specialization of the wing-venation do not indicate important divisions of this order. Within a single family forms exist in which there has been a reduction of the wing-venation and also others in which the specialization has been by addition. A few illustrations of the more striking modifications of the primitive plecopterous type follow. Specialization by reduction. —There are many cases of specialization by reduction in this order; a common example is the reduction of the radial sector of one or of both pairs of wings to an unbranched condition; this occurs in each of the five families. *Much use is made by systematists of the characters presented by this part of the anal area. When the coalescence or anastomosis of the second and third anal vein docs not extend beyond the end of the cell, as in Iso^eniis (Fig. 245), it is said that "two simple veins extend from the cell below," when the union of these two veins extends beyond the limits of the cell, as in Nemoura (Pig. 248), it is said that "a single forked vein extends from the cell below." The inclusion of the word below in these expressions is to indicate that the first anal vein is not included in the enumeration. THE WINGS OF PLECOPTERA 255 A more striking example of specialization by reduction is afforded by certain forms in which the hind wings are so reduced in size that they are smaller than the fore wings and have fewer veins. An excellent illustration of this is the common Chloroperla cydippe (Fig. 253). In this species the radial sector of the fore wing is nearly reduced to an unbranched condition and in the hind wing this reduction is complete. The most remarkable feature, however, is the extreme reduction of the anal area of the hind wing, which results in the hind wing being much smaller than the fore wing. Sci Fig. 253. — Wings of Chloroperla cydippe. It is unnecessary to cite other examples of specialization by reduction, as they will be easily recognized when met. Specialization by addition. — A few examples of specialization by addi- tion are illustrated by the accompanying figures. The most convenient method of discussing these examples is to treat each of the principal veins of the wing separately. The subcosta. — In many genera there are in addition to the very con- stant humeral cross-vein other veins that extend from the subcosta to the costa. These are commonly termed cross- veins; but as they are preceded by tracheae and the true cross-veins are not preceded by tracheae in this order, they are morphologically accessory veins. These accessory sub- costal veins are present in the wings of Isogenus (Fig. 245); and the presence of a trachea in each was clearly evident in a pupal wing of Acro- neuria, which I studied. The radius-one. — Accessory veins, closely resembling those described in the preceding paragraph, extend from radius-one to the costal margin of the wing in some genera. These are also shown in Figure 245. 256 THE WINGS OF PLECOPTERA The radial-sector. Accessory veins are frequently developed upon the radial sector. In most cases when such veins are present they are borne by vein R2+3; vein R4+5 is, as a rule, unbranched. Occasionally the arrange- ment of the accessory veins is such as to give the radial sector the appear- ance of preserving its primitive four-branched condition. But as already stated this condition is merely fortuitous. In Isogenus (Fig. 245), vein R2+3 bears two accessory veins in the fore wing and one in the hind wing. In certain species of Acroneuria and in Pteronarcys several accessory veins is borne by this vein. As a rule, these accessory veins extend in approx- imately parallel courses except near the apex of the wing. But in certain Fig. 254.- — Wings of Pteronarcys dorsata, female (Drawn by Miss L. W. Smith). forms, their courses are so irregular that in the adult wing the accessory veins and the cross-veins extending between them form a chaotic network. This is well-illustrated by the fore wing of Pteronarcys dorsata (Fig. 254); in the hind wing the confusion is not quite so great. The media. — Accessory veins are comparatively rarely developed on the branches of media, this vein remaining two-branched even when both the radial and the cubital areas are expanded. This is wcll-sliown by Acro- neuria. The cubitus. — Accessory veins are frequently developed on the distal part of vein Cui. In Acroneuria for example, there are three such veins that are preceded by tracheae beyond the ordinary cross veins extending between veins Cui and Cu2, which are not preceded by tracheae. THE WINGS OF PLECOPTERA 257 Correlated with the development of one or more accessory veins on vein Cui there is a tendency of this vein to curve forward towards media, which frequently results in an anastomosis of these two veins. In the fore wing of Isogenus (Fig. 245) this extension forward of vein Cui is marked; but the cross-vein m-cu still persists. In the hind wing of this species, the cross- vein m-cu has been obliterated by the anastomosis of veins M3+4 and Cui. The result of this anastomosis is to produce the appearance of media being three-branched; but what appears to be the third branch of media is vein Cui. The first anal vein. — Occasionally, as in some species of Pteronarcys, the first anal vein is forked at the tip ; but, as a rule this vein is unbranched in both fore and hind wings. The second and third anal veins. — The anal area of the fore wings is rarely expanded, the second and third anal veins remaining unbranched in most genera; but in a few cases, as in Pteronarcys (Fig. 254), these veins are branched. In the hind wings, the anal area is expanded in many genera; and correlated with this expansion, there is a branching of both the second and third anal veins; this is shown in Isogenus (Fig. 245) and in Pteronarcys (Fig. 254). The most remarkable expansion of the anal area of the hind wings that I have observed in this order is that of Eusthenia spectabilis (Fig. 246). There is here a certain resemblance to the hind wing of an orthopterous insect ; for in several places there are what appears to be the beginnings of the development of intercalary veins. CHAPTER XIII THE WINGS OF THE CORRODENTIA The winged members of this order have four membranous wings; the fore wings are larger than the hind wings ; and both pairs when not in use are placed rooflike over the body, being almost vertical, and not folded in plaits. The wing-veins are prominent, but the venation of the wings is reduced. The determination of the homologies of the wing- veins in this order was a problem that sorely puzzled all who worked upon it until it was approached by the ontogenetic method by Comstock and Needham. As soon as we 4+5 Fig. 255. — The wings of Psocus (After C. & N.). Studied the tracheation of the wings of nymphs the difficulties vanished. These difficulties are due to the fact that what appear to be cross-veins are sections of longitudinal veins. The type of venation of the wings characteristic of the Corrodentia is well-illustrated by the wings of Psocus (Fig. 255). In the fore wings the branching of the longitudinal veins corresponds quite closely with the hypothetical type except that there is a reduction in the number of the branches, subcosta being unbranched, and radius and media being each only three-branched. In the hind wings, the venation is much more reduced. A study of Figures 256 and 257, which represent two stages in the development of the fore wing of Psocus will convince one of the correctness of the homologies of the veins indicated in Figure 255. These figures were (258) THE WINGS OF CORRODENTIA 259 made from photographs of wings, which were so mounted that the develop- ing wing-veins appeared as pale bands, and the wing-tracheae as dark lines. A remarkable feature of the fore wing of Psocus (Fig. 255) is that, although it is braced in every direction, there is not a single cross- vein in it, Fig. 256. — Fore wing of a nymph of Psocus (After C. & N.). the bracing being accomplished by the zigzag courses of the principal veins. There is an arculus (ar) near the base of the wing, which is formed of the base of media ; and what appear as cross-veins in the central portion of the wing are sections of media and of cubitus. It was these apparent cross- veins that made the determination of the homologies of the wing- veins difficult before the tracheation of the wings was observed. Although there are no cross- veins in the wing represented by Figure 255, cross-veins exist in the wings of certain members of this order. In some genera the radial cross-vein is present and in some, instead of an anasto- Fig. 257. — Fore wing of a full-grown nj'mph of Psocus (After C. & N.). mosis of veins M and Cui, these veins are connected by a medio-cubital cross-vein. Both of these cross-veins exist in the fore wing of Stenopsocus (Fig. 258). 260 THE WINGS OF CORRODENTIA In the remarkable genus Neurostigma described by Enderlein ('oo) from Peru, the pterostigma of the fore wing is crossed by from eight to ten cross- veins. Enderlein ('03) in a paper on the Corrodentia of the I ndo- Australian fauna adopts the uniform terminology of the wing-veins and gives an extended discussion of the venation of the wings of this order; this has been of much aid in the making of the following generalizations: The costa. — In Psocus, according to the observations of Comstock and Needham, the margin of the adult wing is tubular throughout, there being present an ambient vein R ^^2 + 3 Fig. 258.^Vings of Stenopsocus (After Enderlein '03). A diagram in which the coalesced veins are represented separate. The costal and anal portions of this vein doubtless represent the costa and the third anal vein respectively, al- though the con^espond- ing tracheae are appar- ently lost. The distal portion of this ambient vein was preceded by the anastomosing tips of all of the wing tracheae, as is shown in the figures of the wings of nymphs. The subcosta. — The subcosta is not forked. In the fore wings of nymphs of Psocus (Fig. 256 and 257), the subcostal trachea extends unbranched to the apex of the wing, where its tip enters the forming ambient vein. In the wings of an adult Psocus (Fig. 255), the subcosta is short, and its tip coalesces with the radius in the fore wing and with the costa in the hind wing. In many genera the subcosta consists of two parts : a basal part which lies in the basal part of the costal cell and a distal part which extends from vein Ri to the costa at the base of the pterostigma, the inter- mediate part of the vein being lost. This condition is well shown in the fore •wrng oi Stenopsocus (Fig. 258). The radius. — The first forking of the radius, the separation into veins Ri and the radial sector, is quite typical. In the fore wing vein Ri curves back from the costa, bounding the pterostigma; in the hind wing the pterostigma is wanting. The radial sector is usually reduced to a two- branched condition, as in Psocus (Fig. 255) ; in some fomis it is reduced to an unbranched condition, and in others it is not reduced, the four branches remaining separate. THE WINGS OF CORRODENTIA 261 The media. — The media coalesces with radius for a short distance at the base of the wing; it then bends backward and joins cubitus with which it coalesces for a considerable distance; it is usually three-branched in the fore wings, but in Ptiloneura hidorsalis, described by Enderlein ('oo) from Peru, it is eight-branched. This is a remarkable departure from the almost universal method of specialization in this order; and even in this genus, where accessory veins have been developed upon media, the radial sector is reduced to a two-branched condition. In the hind wings, media is usually reduced to an unbranched condition ; but in some forms it is forked. The cubitus. — The cubitus is quite widely separated from media at the base of the wing; this is well shown in the wing of a nymph of Psocus (Fig. 257); but it soon joins media, with which it coalesces for a considerable distance. In the fore wing, vein Cui anastomoses with media in some forms (Fig. 255), in others the two veins are connected by the medio- cubital cross-vein (Fig. 258), and in other forms there is no connection between veins M and Cui. Vein Cu2 is short when present; it is wanting in some forms. In the hind wings cubitus is reduced to an unbranched condition. The anal veins. — As a rule the first and second anal veins are present in both fore and hind wings ; in a few genera a short third anal vein is present in the fore wings. All of the anal veins are unbranched. In the fore wings, the first and second anal veins usually end at the same point in the margin of the wing. CHAPTER XIV THE WINGS OF THE EMBHDINA The winged members of this order have two pairs of wings, which are quite similar in form and structure. The wings are elongate, membranous, extremely delicate, and folded on the back when at rest. The venation of the wings is considerably reduced; this reduction has been brought about ^'^'^^^^^^ '^\M<'<'mf^ Fig. 259. — Fore wing of Oligotoma saundersi: A, the wing; B, outline of the wing showing the existing venation ; C, outline of wing showing the venation restored (After Wood-Mason). both by the coalescence of veins and by the atrophy of veins. In a single species, Clothoda nohilis from South America, the cubitus bears accessory veins; but even in this species the radial sector is only three-branched and the media is reduced to an unbranched condition. The cross-veins are comparatively few in number. The females are always wingless; and in two of the eleven known genera both sexes lack wings. Each of the veins of the wings extends along the middle of a brown band; between these bands the membrane of the wing is pale in color. (262) THE WINGS OF EMBIIDINA 263 The alternating brown and pale bands give the wing a very characteristic appearance (Fig. 259, A). In those forms where the venation of the wings has been reduced by the atrophy of veins, the brown bands persist after the veins have faded out ; hence it is easy to determine by these bands the former position of veins that have beefi lost. In Figure 259, A represents the appearance of a wing of Oligotoma sann-dersi; B, the existing venation; and C, a restoration of the venation based on a study of the brown bands. Here and in other figures that follow, the restored veins are indicated by dotted lines. On each side of vein Ri there is a narrow line of deeper color than the brown bands; these lines have been termed, by Enderlein, the fore and hind radial border lines (Radiussaumlinie) , respectively; these are repre- sented in the following figures by series of dots, parallel with vein Ri. The membrane of the wing is clothed throughout with fine setae, and along each vein and along each margin of each brown band there is a series of larger setae. The tracheation of wings of nymphs has been studied by Melander ('02), who has figured that of Emhia texana (Fig. 260). As the order is very poorly represented in the collections to which I have access, I have been forced, in my studies of the wing venation of mem- bers of it, to depend almost entirely on the published figures of wings of these insects. Fortunately there has recently appeared a monograph of the order by Enderlein ('12) in which figures of the wings of representatives of eight of the nine winged genera are given, and the wings of the remaining winged genus are described in detail. The following account is based on the data given by Enderlein. I have not seen the monograph of this order published by Dr. H. A. Krauss ('11); but have made use of one figure which is copied from it by Enderlein.* The wings of Embia sahulosa (Fig. 261) will serv^e to illustrate the fea- tures of the wing-venation most commonly found in the wings of insects of this order. The subcosta is well-preserved, but is unbranched; the radial sector arises near the base of the wing and is only three-branched, veins M,+, Fig. 260. — Wing of a pupa of Emhia texana (After Melander). *In copying the figures of Enderlein I have made some sHght changes in the lettering of the wing-veins. The vein that he designates the "Cubitalstamm {cast)" is vein Cuo; the accessory veins borne by vein Cui I have designated as Cuia and Cuib instead of Cu2 and Cu3, his vein Cu2 in this case is not homologous with the vein Cu2 of the uniform terminology; and the vein that he designates as the "Axillaris (a.v)" is the second anal vein of the uniform terminology. The modification of the terminology of the veins of the anal area by this author is discussed in the chapter on the wings of the Plecoptera (p. 248). When media is two-branched, the two branches are doubtless veins Mi-1-2 and Ms-f 4, and not veins M 1 and Mo as labeled by Enderlein. 264 THE WINGS OF EMBIIDINA R2 and R3 coalescing to the margin of the wing ; the media is reduced to an unbranched condition ; there is what appears to be a typical arculus at the base of the wing; the cubitus is two-branched; the first anal vein is Ciii Fig. 261. — Wing of Embia sabulosa (After Enderlein, with some changes in the lettering of the veins). replaced by an anal furrow; the second anal vein is well-developed; and there are comparatively few cross veins. Illustrations of the more important modifications of the type of wing- venation possessed by Embia sabulosa are shown in the following figures. ^3U ^' + ^ Fig. 262. — Wings of Donaconethis abyssinica (After Enderlein). In several genera veins Ri and R2+3 coalesce at the tip; this condition is shown in Figures 259 and 262. THE WINGS OF EMBIIDINA 265 Enderlein states that in the subfamily Embiinas, vein R4 or R5 is forked in some very rare cases; in Figure 262 a vestige of an accessory vein is shown on vein R5 of the hind wing. While the media is usually reduced to an unbranched condition, in the genus Donaconethis (Fig. 262) it is two-branched. Fig. 263. — Wing of Teratembia geniculata (From Enderlein after Kraus). Vein Cui is reduced to a vestigial condition in many forms; this is the case in the hind wing of Donaconethis (Fig. 262). The radial sector is reduced to a two-branched condition in Oligotoma (Fig. 2 59) ; but is usually three-branched. In those forms where it is three- branched, it is almost invariably veins R2 and R.3 that coalesce completely; but in the genus Teratembia (Fig. 263) veins R2 and R3 are separate, while veins R4 and R5 coalesce throughout their length. Fig. 264. — Wings of Clothoda nobolis (After Enderlein). In the genus Clothoda (Fig. 264), of which only a single species is known, vein Cui bears two accessory veins in the fore wing and one in the hind wing. In Teratembia (Fig. 263) and in some species of Embia there appears to be a vestige of an accessorv vein borne by vein Cuo. 266 THE WINGS OF EMBIIDINA The presence of accessory veins in the preanal area of the wings of Clothoda and occasional occurrence of vestiges of accessory veins in other genera raises a question as to the position of this order in the table of the methods of specialization of the wings characteristic of the orders of insects given at the close of Chapter VI, and upon which the linear arrangement of the orders adopted in this essay is based. In spite of the presence of accessory veins in the preanal area of the wings in the cases mentioned, it is evident that the characteristic method of specialization of the wings in this order is by reduction; even in Clothoda the number of the wing-veins is greatly reduced, the radial sector being only three-branched and the media reduced to an unbranched condition. CHAPTER XV THE WINGS OF THE THYSANOPTERA The winged members of this order have four wings; these are similar in form, long, narrow, membranous, not plaited, with but few or with no veins, and only rarely with cross- veins ; they are fringed with long hairs, and in some species are armed with spines along the veins or along the lines from which veins have disappeared. The two wings of each side are united by spines, the arrangement of which are described by Hinds ('03) as follows : "Upon the costa of the hind wing, near its base, stand about five short spines in the Terebrantia and two or three in Tubulifera, which are hooked at their tips. When the wings are spread in flight these tiny hooks engage a membranous fold on the underside of the scale [anal area] of the fore wing. Beyond these small Fig. 265. — Fore wing of JElothrips nasttirtii (After Jones). The lettering is original. hooks stand a single stouter spine which also forms a hook. From the hind angle of the scale of the fore wing proceed two long, stout spines, standing so closely together as to often appear like one, and these engage the solitary stouter hook on the hind wing. Thus united the wings move together, but as the connection is so near the bases of the wings it can not be very strong. ' ' When at rest the wings are folded back flat upon the abdomen. Although the wings are usually present in adults, certain species are apterous. Even in the most generalized forms the venation of the wings is greatly reduced. A fore wing of ^olothrips nasturtii, one of the more generalized members of the order, will serve as an illustration of this fact (Fig. 265). This figure is a copy of one given by Jones ('12), to which I have added letters indicating my conclusions regarding the homologies of the wing- veins. The costal vein is present and is continued by an ambient vein, which margins the entire preanal area of the wing (Fig. 265, avi). The ambient (207) 268 THE WINGS OF THYSANOPTERA vein is termed the "ring vein" by writers on this order, although the term ambient vein has been long in use for veins in this position. There is a short longitudinal vein separating the anal and preanal areas (Fig. 265, A) ; this is doubtless the anal vein. Between the costa and the anal vein, there are only two longitudinal veins (Fig. 265 i? and Cu); of the four longitudinal veins that typically traverse this area, the subcosta, radius, media, and cubitus, the radius and cubitus are almost invariably the more persistent when there is a reduction of the wing venation. I therefore conclude that these two veins are the radius and the cvibitus respectively. These two veins have been designated as the anterior longitudinal vein and Fig. 266. — Fore wing of Ervthrolkrips arizonce (After Moulton). the posterior longitudinal vein respectively; but it is obvious that terms that pertain to the uniform terminology are preferable so long as there is no reasonable doubt as to the homologies of the veins. In some members of the order there are a few veins extending trans- versely to the length of the wing. It is possible that these are branches of the longitudinal veins that have become transverse by a coalescence of their distal portions with adjacent veins; but in the absence of any data to confirm this suggestion, it is better to designate them simply as cross-veins. A good illustration of this type of wing is that of Erythrothrips arizoncB, figured by Moulton ('11) (Fig. 266). CHAPTER XVI THE WINGS OF THE HOMOPTERA (a) THE MORE GENERAL FEATURES OF THE WINGS OF THE HOMOPTERA In the order Homoptera the two pairs of wings are usually similar in texture, and each wing is practically of the same thickness throughout. This is in marked contrast to the conditions in the Heteroptera, where the two pairs of wings differ in structure, the basal half of the front wings being thickened. It was this difference in the structure of the wings of the two groups that suggested the contrasting names Homoptera and Heteroptera for what were regarded until recently as two suborders of a single order, the Hemiptera. The wings are usually membranous, but in some the front wings are subcoriaceous. In these cases, however, they are of quite uniform texture throughout, and not thickened at the base as in the Heteroptera. A striking dift'erence be- tween the wings of the Hom- SeR+M+ Cu,+ lstA optera and those of the Heteroptera is that in the Homoptera when an anal furrow is developed it occu- pies the usual position, that is behind the cubitus, along the first anal vein instead of in front of the cubitus, as in the Heteroptera. Many wingless forms exist in this order; in the family Coccidae the females are always wingless; and in the family Aphididas the males may be either winged or wingless, while the females and certain generations of the agamic forms are wingless. In the Coccidae the males have only a single pair of wings, the hind wings being represented by a pair of club-like hal teres. Each of these is furnished with a bristle, which in all of the species that I have studied is hooked, and fits in a pocket on the wing of the same side. If only the wings of adult Homoptera be studied, the venation of these wings appear to depart widely from the hypothetical primitive type. There seems to be little in common with this type in the wings of an aphid (Fig. 267), of a membracid (Fig. 268), or of a cicada (Fig. 269). But in each case when the trachea that precede the wing-veins are studied it is easy to determine the homologies of the wing-veins. (269) Fig. 267. — The wings of an aphid (After Patch). 270 THE WINGS OF HO MO PT ERA Fig. 26'8. — The wings of a membracid, Thelia bimaculata (After Funkhouser). Fig. 269. — The wings of a cicada (After C. & N .). THE WINGS OF A CICADA 271 This has now been done in the case of representatives of each of the families of this order, with the result that the homopterous type of wing- venation is now well-known. The first work of this kind in this field was by Comstock and Needham ('gS-'gg) who traced the development of the wing-veins of a cicada. Ten years later Miss Patch ('09) studied the development of the wing-veins of Aphididse, Psyllidae, Aleurodidse, and Coccidas; and still later, Funkhouser ('13) determined the homologies of the wing-veins of the Membracidse. In the same year there appeared two papers by Professor Z. P. Metcalf, one on the wings of the Jassidse (Metcalf '13a), and one on the wings of the Fulgoridas (Metcalf '13b); and more recently this author has completed the series by a paper on the wing-veins of the Cercopidas (Metcalf '16). Abstracts of each of these papers except those on Fulgoridas and Cercopidas are given in the following pages. In addition to these papers, each of which treats of the wings of a family of the Homoptera, Dr. Karel Sulc ('11) has made a careful study of the tracheal development in the successive stages in the development of the wings of a cercopid, PhilcBnus lineatus. (b) THE WINGS OF A CICADA The development of the wing-veins of a cicada was traced by Comstock and Needham by a study of a series of nymphs of different ages and recently emerged adults. The results obtained were exceedingly gratifying. We had antici- pated encountering much difficulty in determining the homologies of the wing-veins of the Homoptera. We were filled with delight, therefore, when we found within this order, preserved almost un- changed, what we had come to regard, from a study of other orders, as the primi- tive type of wing- venation. The following account of the development of the Fig. 270. — The fore wing of a voung nvmpli of a cicada (After C. & N.). wing-veins of a cicada is abstracted from our joint paper. The close correspondence of the wing-venation of a cicada with the primitive type is not obvious if one studies only the wings of the adult (Fig. 269); for in this stage there is a massing of several veins along the costal margin of the wing, and the cross-veins have the same appearance as the branches of the primary veins. 272 THE WINGS OF A CICADA In the wings of a young nymph, on the other hand, the tracheae that precede the veins are not massed as they are later; and in the older nymph, where the forming veins appear as pale bands the cross-veins contain no tracheae, and can be thus easily distinguished from the longitudinal veins. Figure 270 represents the tracheation of the fore wing of young nymph ; and Figure 271, that of the hind wing. In each of these figures the dotted line a-b indicates approximately the line along which the hinge of the wing of the adult is formed. In the fore wing of this young nymph the only depar- tures from the typical branch- ing of the tracheae are the following ; trachea Ri coalesces with the radical sector to a point beyond the separation of trachea R4+5; the first anal Fig. 271. — The hind wing of a young nymph of a cicada (After C. & N.). trachea coalesces with trachea Cu for a short distance; and the second and third anal tracheae are united at the base. These differences are remark- ably slight compared with the great changes that have taken place in the specialization of the mouth-parts and other organs of the adult cicada. In the hind wing the tracheation is much more reduced. An especially striking feature is the complete loss of trachea Ri, which is considerably reduced in the fore wing. In a wing of a mature nymph (Fig. 272) trachea Ri is completely aborted. In fact one of the most characteristic features in the venation of the Homoptera, and of the Heteroptera also, is the absence or very great reduc- tion of vein Ri in the adult wings of most members of the order. The Fig. 272. — The fore wing of a mature nymph of a cicada (After C. & N. THE WINGS OF A CICADA 3/a- monanthe pitlchella and in Smilia camelus, veins R2 and R3 end separately. The media. — The medial tra- chea is easily recognized by its position in the costo-radial group of trachea (Fig. 276) ; thisposition being that of the radial trachea in the Plecoptera and in those cockroaches in which a transverse basal trachea has not been developed, and which was adopted in the construction of the hypothetical type (Fig. 278). In the Membracidae, media is usually reduced to a two-branched condi- tion; and in most members of the family vein M1+2 unites with vein R4+5 for a distance (Fig. 268). In a few genera media is three-branched ; as in Xantholobus trilineatiis, where vein M1+2 is forked, and in Archasia belfragci, where vein M3+4 is forked. The cubitus. — An examination of the tracheation of the wings of n^^mphs shows that the first member of the cubito-anal group of trachese is a large branched trachea which at first sight appears to be the cubital trachea (Fig. 276 and 277); bvit the branching of this trachea occurs proximad of what becomes the hinge line of the adult wing; so that the two branches represent two distinct veins of the adult wing, i. e. veins Cu and ist A. This condition is almost identical with what exists in the cicada (Fig. 270, p. 271). The recognition of cubitus in the Membracidae is more difficult than in the cicada, because in the Membracidae this vein is reduced to an un- branched condition. But Funkhouser found that the cubital trachea is Fig. 283. — Highly magnified portion of fore wing of a nymph of Vanduzea arquata showing the region of trachea Ri. 280 THE WINGS OF J A SSI DM branched in the wings of many nymphs. This is weh shown in his figure of the wing of a nymph of Ceresa diceros. This interpretation of the identity of cubitus makes the position of the anal furrow, namely along the first anal vein, in the membracid wing agree with that found in those other families of the Homoptera in which an anal furrow is developed. The anal veins. — In most members of this family the three anal veins are presented more or less distinct in both fore and hind wings. In the wings of many nymphs the third anal trachea is branched (Fig. 277) ; but in the wings of the adults the third anal vein is simple. The tip of the first anal vein usually coalesces with cubitus, thus forming a part of the "marginal veins." (Fig. 279). In the fore wings of the larger number of the genera the second and third anal veins are separate at the base of the wing but coalesce throughout the greater part of their length (Fig. 279); but in the hind wings they frequently coalesce at the base and end separately (Fig. 281). The cross-veins. — Funkhouser states that "Of the cross- veins which appear in the fore wing, three only are constant and characteristic of the family, the others being peculiar to certain genera and species and of little comparative importance." These three cross- veins are shown in Figure 279; they are the sectoral (5), the medio-cubital (m-cu), and the posterior arculus (pa). This last cross-vein is usually in the basal third of the wing but is surprisingly variable in position sometimes migrating beyond the middle of the wing. Funkhouser makes no suggestion as to its identity; to me it seems probable that it is the posterior arculus moved out of its typical position. I am led to this conclusion by a comparison of the membracid wings with that of the cicada in which the arculus is typical. The marginal vein. — A very characteristic feature of most membracid wings, and of the wings of cicadas also, is the presence of an undulating vein parallel with the outer margin of the wing. This vein is formed by the united tips of the longitudinal veins, and is termed the marginal vein, as it forms the edge of the veined part of the wing. (d) THE WINGS OF THE JASSID.« An extended study of the wing-venation of the Jassidaj has been made by Metcalf ('13), who figures in his published account the tracheation of the wings of nymphs of representatives of twenty genera and also the wings of adults of a larger niimber of genera. These represent all of the sub- families and tribes of the Jassidce commonly found in Eastern North America. I reproduce here one of his plates illustrating the tracheation of the wings of nymphs (Fig. 284). The results of this investigation are brieflv as follows: THE WINGS OF J A SSI DM 281 Fig. 284. — Wings of nymphs of jassids: 1, iore wing oi A gallia 4- ptoiclata; 2, hind wing oi A pallia 4- punctata; 3, fore wing of Oncometopia undata; 4, hind wing of Oncometopia undata; 5, fore wing of Diedrocephala cocchiea; 6, fore wing of Draecidacephala mollipes; 7, hind wing of Draeculacephala mollipes; 8, fore wing of Gypona 8-lineata; hind wing of Gypona S-lineata (After Metcalf). I 282 THE WINGS OF JASSIDM The wings of the Jassidse show marked specialization by reduction. This reduction is usually accompanied by the atrophy of one of the branches of one of the main tracheae and the shifting of a branch of a neighboring trachea until it occupies the region of the atrophied trachea. The atrophy of these tracheae with the subsequent shifting of other trachea which take their places gives to the wings of the Jassidse their characteristic aspect. The casta of the fore wing. — The costal trachea was absent in all of the jassid wings examined with the exception of Gypona (Fig. 284, 8). This indicates that the costal trachea has practically disappeared from the Jassidae. The suhcosta of the fore wing. — A subcostal trachea was found in only six of the genera examined; among these are Agallia (Fig. 284, i) and Gypona (Fig. 284, 8). A subcostal vein, however, is well developed in all of the adult wings studied, and it shows very clearly as a distinctly lighter area in all of the older nymphs examined. The radius of the fore wing. — The radial trachea of the fore wing of Jassidse is typically two-branched although in some forms three or even four branches do occur. The two branches of the typical radius represent R2+3 and R4+5. Ri has almost completely disappeared from the fore wings in this family. It does occur, as a delicate branch in a few genera (Fig. 284, 3) but gives rise to a very characteristic cross- vein between sub- costa and radius, which is known currently as the "nodal vein." The media of the fore wing. — The medial trachea is typically two branched in the Jassidae. These branches are Mi +2 and M3+4. Trachea Mi+2 is well developed in only a few genera; in Gypona (Fig. 284, 8) it is vestigial; in the other fore wings represented in this figure the medial trachea is reduced to an unbranched condition ; and this is the case in most of the forms examined by Metcalf . The cubitus and the first anal of the fore wings. — In all of the genera of the Jassidae examined the cubital and the first anal tracheae were the most constant and formed one of the best landmarks in the study of the relations of the tracheae. They coalesce for some little distance from the base of the wing. The cubital trachea is frequently two-branched (Fig. 284, 8) ; but in the greater ntimber of genera studied by Metcalf it is reduced to an unbranched condition. The presence of forms in which trachea Cu2 is reduced to a mere spur indicates that it is this branch that is lost in those foniis where the cubital trachea is unbranched. The first anal vein of the fore wings.- — The first anal vein lies along the anal border of the claval suture. It has not been usually recognized as a distinct vein owing to the fact that as a vein it is rather inconspicuous while the claval suture or fold is very distinct. It is, however, preceded by a conspicuous trachea in all of the genera studied. THE WINGS OF PSYLLID.E 283 The second and third aiials of the fore iving. — The second and third anal trachese in the fore wing are well developed and the third anal is frequently two-branched (Fig. 284, 3). The costa and the subcosta of the hind wings. — In all of the Jassida^ proper the hind wing is very uniform. No costal or stibcostal tracheae have been disco\'ered although the subcostal vein was well defined in all of the older nymphs studied (Fig. 2 84, q). The radius of the hind wing.— The radial trachea of the hind wings is typically two-branched in this family. In a single genus Spanghergiella, the radial trachea is unbranched. No indication of trachea Ri was observed in the hind wing. The media of the hind wing. — The medial trachea of the hind wing is two branched in all of the genera examined. The cubitus of the hind wing. — The cubital trachea of the hind wing is reduced to an unbranched condition in all of the genera examined. The anal tracJiece of the hind wings. — The first anal trachea and the cubital trachea coalesce for a distance as in the fore wing. The second and third anal trachccC are also present in nearly all cases; and the third anal trachea is frequently two-branched. In the adult wing the second anal vein and the anterior branch of the third anal frequently coalesce at the base or anastomose near the middle of their course. The basal connections of the trachece of the wings. — The above abstract, which consists largel}- of quotations from the paper by Metcalf, indicates the more important conclusions reached by this writer. A study of his illustrations indicates another conclusion not mentioned by him; this is that in the Jassidae, as in the Membracidge, a transverse basal trachea has not been developed. With one exception, in all of the forms where the basal connections of the tracheae were traced, the costo-radial and the cubito-anal groups of tracheae are not connected by a transverse basal trachea, and the medial trachea is a member of the costo-radial group. Several illustrations of this are on the single plate reproduced here. (Fig. 284, 2, 5), and there are many of them on the other plates in Metcalf's paper. The one exception to this arrangement of the tracheae indicated by Metcalf's figures is that of the hind wing of Gypona (Fig. 284, 9). As the basal connections of the tracheae are only partly shown in this figtu^e the evidence presented by it is not conclusive. {e) THE WINGS OF THE PSYLLID^ We are indebted to Miss Patch ('09) for the working out of the homol- ogies of the wing veins of the Psyllidas. The following account is based on her paper. See note on page 291. 284 THE WINGS OF PSYLLIDM A general view of the wings. — Figure 285 represents the tracheation of a fore wing of a nymph of Psyllafloccosa. All of the principal tracheae except the third anal are present and each arises separately from a transverse basal Fig. 285. — The tracheation of a fore wing of a nymph of Psylla floccosa (After Patch). trachea. The subcostal trachea is not branched; the trachea that precedes radius-one is well developed; the radial sector trachea is reduced to an unbranched condition; the medial trachea is reduced to a two-branched condition; the cubital trachea is typical; and the two anal tracheae are unbranched. Fig. 286. — The tracheation of the wings of a freshly emerged Psyllafloccosa (After Patch) The wings of a newly emerged adult show remarkable changes in the tracheae, due to the reduction of some and to the coalescence of others (Fig. 286). The tracheation in this stage corresponds very closely with the definitive venation; the venation of the fore wing is represented by Figure 287. THE WINGS OF APHIDID.^ 285 In the adult wing the first anal vein is not developed as a vein but is represented by the anal furrow, or the calval suture as it is termed by writers on the Homoptera. The other changes are indicated by the letter- ing of Figure 287. The stigma. — In some of the Psyllidae the costal margin of the fore wing is strengthened by a stigma. Fig. 287. — The venation of a fore wing of Psylla floccosa (After Patch). (/) THE WINGS OF THE APHIDID^ The tracing of the homologies of the wing veins of the Aphididaj was the subject of a painstaking investigation by Miss Patch ('09) ; in the course of which approximately 100 species representing 17 genera were studied, Fig. 288.- — The tracheation of a wing of a nymph of Schizmxeura rileyi (After Patch). and wings of many nymphs and of more than 2000 newly emerged aphids were examined. The following brief simimar}^ is based on this paper. A general view of the wing. — The most generalized aphid wdng figured by Miss Patch is that of a nymph of Schizoneiira rileyi (Fig. 288). In this wing there are four main tracheal tnmks, which arise from a transverse basal trachea. The costal and subcostal trachese are lacking; and only 286 THE WINGS OF APHID ID^ one anal trachea is present. The trachea of the radial sector is unbranched ; that of media is only two-branched; and that of cubitus is not forked. In the wing of a newly emerged adult (Fig. 289) all the tracheae of the cubito-anal group coalesce at the base of the wing. Fig. 289. — The tracheation of a freshly emerged adult Schizoneura rileyi (After Patch). It should be observed that here, as is usually the case where there is a transverse basal trachea, the medial trachea becomes a member of the cubito-anal group. In two other cases the basal connections of the trachese are figured by Miss Patch. Figure 290 represents the tracheation of the fore wing of a newly emerged Aphis; and Figure 291 the tracheation of the hind wing of a newly emerged Chaitophorus populicola. The data given above furnishes the key to an understanding of the tracheation of the wings of Aphids. The two principal tracheae, which extend from the base of the wing outward parallel to each other and to the costal margin of the wing, represent respectively the costo-radial and the Fig. 290. — The tracheation of a fore wing of a newly emerged Aphis (After Patch). cubito-anal groups of trachea. Of the former only the radial trachea is preserved in this family; to the latter belong the medial as well as the cubi- tal and anal tracheas. A comparison of Figure 292, which represents the wings of an adult Schizoneura americana, will serve to show the relation of the tracheation of I THE WINGS OF APHID ID^ 287 the wing to the definitive venation. This relation will be made more clear by the following discussion of the separate veins. The costa. — A costal trachea was not found in any of the wings examined. But Miss Patch observed that not only is the costal margin of the wing stiffened by a vein-like structure, but this part of the wing contains what appears to be a vein-cavity; as was shown by the fact that in severing the Fig. 291. — The tracheation of a hind wing of a newly emerged Chaitophorus populicola (After Patch). wing from a freshly killed aphid, the yellow body fluids frequently flow into this vein and extend along to about the region of the stigma (Fig. 290). The subcosia. — No subcostal trachea was found in any member of this family. It may be that the subcostal vein is completely lost, or, as Miss Patch concluded, this "is present in the large main vein channel of the wing, and extends from the base of the wing to the stigma wh.re it approaches the margin of the wing." The lettering Sc*^R+M+Cu,+ lstA of her figures is based on this conclusion. The radius. — In the more generalized members of the family, as shown by the figures given above, radius is two-branched, in the fore wings radius-one being pre- served, and the radial sector being unbranched. In the hind wings, radius-one is lost. In the Chermesinae, vein Ri is wanting in the fore wings as well as in the hind wings. Figure 293 represents the tracheation of a fore wing of Chermes abietis; Figure 294, the tracheation of a hind wing of Chermes piniJolicB; and Figure 295, the definitive venation of both wings of a Chermes. See note on page 291. A striking feature of the tracheation of the wings of Chermes is the presence of many secondary branches of the tracheae. This feature was observed in other genera also. Fig. 292 — The wings of Schizoneura americana (After Patch) . ' 288 THE WINGS OF APHIDIDM The media. — The data upon which the identification of media is based is given above. As to the nature of this vein, it may be three branched, as Fig. 293. — The tracheation of the fore wing of Chermes abietis (After Patch). in Aphis (Fig. 290), or two-branched, as in Shizoneura (Fig. 292), or it may be reduced to an unbranched condition, as in Chermes (Fig. 295). The cubitus. — "Cubits is present in all genera of Aphididae and in all of them unbranched." The anal veins. — A single anal vein is developed in this family. That this is the first anal vein was demonstrated by the finding of the second and third anal tracheae in wings of newly emerged aphids (Fig. 290); but veins are not developed about these tracheae. Fig. 294. — The tracheation of a hind wing of a nymph of Chermes pini- folicB (After Patch). .Sc+jR+^f+Ct/ ^'«'^^ gc Fig. 295. — The venation of the wings of Chermes (After Patch). I THE WINGS OF ALEURODIDJE 289 Comstock-Needham COMPARISON OF TERMINOLOGIES OF THE WING-VEmS OF THE APHIDID.^ The following table is given by Miss Patch: — After Buckton Current terminology FORE WING Costal Nenoire Costal Cubitus or post-costal Subcostal vein nervnire Stigmatic First furcal Second furcal Cubital nervnire Second oblique First oblique Stigmal vein First branch Second branch Third discoidal or cubital Second discoidal First discoidal HIND WING Cubitus or post costal Subcostal vein nervure Second oblique Second discoidal First oblique First discoidal (g) THE WINGS OF THE ALEURODID^ In the fore wings of freshly emerged Aleurodes, Miss Patch found four fine but distinct tra- chese, the costal, subcostal, radial, and cubital (Fig. 296, a). All of these extend separate to the base of the wing. The medial trachea is suggested merely by a very faint and delicate but constantly appearing tracing in the wing. In the definitive venation of the fore wing (Fig. 296, 6), costa, subcosta, vein Rj, and media are lacking, only the main stem of radius, the radial sector, and cubitus being developed. Costa (C.) Subcosta (Sc) Radiusi (Ri) (together with basal portions ofthe remaining wing veins). Radial sector (Rs) Media (Ml) Media (M2) Media (M) Media 3+4 (M3 +4) Cubitus (Cu) First Anal (ist A.) Radial sector (Rs). Media (M.) Cubitus (Cu.) Fig. 296. — Wings of Aleurodes (After Patch). 290 THE WINGS OF COCCIDM The hind wing of Alenrodes (Fig. 296, c) has but one trachea and one vein; this is doubtless the main stem of radius and tlie radial sector. {h) THE WINGS OF THE COCCID^ The tracheation of the wings of a species of Dactylopius found on cactus is figured by Miss Patch (Fig. 297) ; and the definitive venation of the same species is also figured (Fig. 298). The following is from her account. "The trachese in the wing of this coccid remain distinct until after the veins begin to form so that the relation of the two is at once discerned. One vein follows the general trail of the subcostal and radial tra- cheae. This vein very evi- dently represents the radius." (In her figure it is labeled Rg. Fig. 297. — The tracheation of a wing of Dactylopius (After Patch). I have taken the liberty of changing the designation to R) . "The second vein follows the base of the first tracheal group to about the point where the medial trachea separates from the subcostal and radial. The vein here takes a direct line for the middle of the caudal margin of the wing. For slightly less than one-third the length of this vein it frequently joins the path of the cubital trachea. This corresponds most closely with media. "Besides these two main veins a short spur representing the subcosta is present. "In a wing so highly special- ized as the coccid wing it is not improbable that the trachea- tion has lost its value as a basis for the venation. Certainly in the species studied there seems no necessary connection between the tracheas and the veins which are found later." The fact that the medial trachea is a member of the costo-radial group of trachese suggests, in the light of the condition found by Funkhouser in the Membracidae, that perhaps in the Coccidce a transverse basal trachea has not been developed. Fig. 298. — The venation of a wing of Dactylopius (After Patch). \ THE WINGS OF HOMOPTERA 291 {i) SUPPLEMENTARY NOTE It has been shown in the preceding pages that in a Cicada, the Alem- bracida?, and Jassida^, vein Ri is either absent or greatly reduced and that, as a result of this, the two principal branches of radius are veins Ro+sand R4+5. This fact suggests the possibility that the two branches of the radius in the Psyllidas and the Aphididas are veins R2+3 and 4+5 and not veins Ri and Rg as indicated in the accounts of the wings of these families given above. But as I know of no observation demonstrating the atrophy of vein Ri in these families, I have not felt free to change the determinations of homologies of the branches of radius published by Miss Patch; and I merely suggest that the method of the reduction of the radius in these and allied families be investigated farther. CHAPTER XVII THE WINGS OF THE HETEROPTERA Fig. 299. — Diagram of a front wing of a bug; cl, clavus; co, corium; ni, membrane. (a) THE MORE GENERAL FEATURES OF THE WINGS OF THE HETEROPTERA In the Heteroptera there is a remarkable difference in the texture of the two pairs of wings, which suggested the name of the order. The basal half of the front wings is thickened so as to resemble the elytra of beetles, only the terminal half being wing- like. The hind wings are mem- branous, and are folded beneath the front wings. On this account the front wings are often termed wing- covers; they are also termed /z^w^/j- tra, a word suggested by their structure. The front wings present characters much used in the classification of these insects; and consequently special names have been applied to the different parts of them. The thickened basal portion is composed of two pieces joined together at their sides ; one of these is narrow and is the part next to the scutellum when the wings are closed (Fig. 299, d); this is distinguished as the clavus. The other broader part is the corium (Fig. 299, co). The terminal portion of the front wing is designated as the membrane (Fig. 299, w). In cer- tain families, the Acanthiidae for example, a narrow piece along the costal margin of the wing is separated by a suture; this is the emhoUum (Fig. 300, e). In certain other cases as the Capsidae, for example, a tri- angular portion of the terminal part of the corium is separated as a dis- tinct piece; this* is the cuneus (Fig. 301, cu). Fig. 300. — Diagram of a front wing of an acanthiid : e, embolium ; co, corium ; cl, clavius; m, membrane. Fig. 301. — Diagram of a front wing of a capsid; cu, cuneus; e, embolium; co, corium; cl, clavus; m, membrane. (6) THE TRACHEATION OF THE WINGS OF THE HETEROPTERA The wings of the Heteroptera exhibit remarkable departures from the primitive type of wing-venation. So great are these that, at first, one sees very little in common between the wings of a bug and those of insects of (292) \ THE WINGS OF HETEROPTERA 293 any other order. But an examination of the tracheation of the wings of nymphs of bugs shows that these wings are merely modifications of the primitive type. Fig. 302. — Tracheation of a fore wing of a pentatomid nymph (After C. & N.). A quite extended study of the development of wings of the Heteroptera was made by Comstock and Needham; and as I have no additional data that tend to modify our conclusions I will abstract our account of the results of this investigation.* In our studies of Heter- optera we examined n^Tnphs of the following families: Notonectidse, Nepidae, Belo- stomidae, Reduviidae, Nabi- d^e, Capsidse, and Pentato- midae. Of these there is no doubt that the most generalized condition of wing-venation is found in the family last named, but further studies in other fami- lies may reveal a still more primitive type. Fig. 303. — Tracheation of a hind wing of a pentatomid nymph (After C. & N.). The tracheation of the fore wing of a pentatomid nymph is represented by Figtire 302, and that of the hind wing by Figure 303. In the fore wing trachea C is well-preserved. Tracheae Sc and R are closel}^ approximate in the basal half of the wing, foreshowing the coalescence of subcosta and radius. In the distal half of the wing trachea Sc traverses that part of the *Handlirsch ('o6-'o8) figures the tracheation of the wings of Lygceiis (Lygaeidae), Syromastes (Coreidae), and Nepa (Nepidaj). 294 THE WINGS OF HETEROPTERA wing which would be traversed by Trachea Ri were it well developed. Trachea Ri is present, but is reduced to a vestigial condition. It is evident that a supplanting of vein Ri by the subcosta takes place here as in Cicada. Trachea Rg has its characteristic bend at the base, and is two-branched. Trachea M is typical except that the branch M3 coalesces with Mi +2 for a short distance. Trachea Cu is six-branched ; it is evident that a specializa- tion by addition has taken place here. Only a single anal trachea has been preserved. The hind wing of the same nymph (Fig. 303) presents a very similar arrangement of tracheae, except that their branches are reduced. (c) THE VEINS AND FURROWS OF AN ADULT WING In those pentatomids in which we have been able to trare the courses of the trachese of the wings, the wing-veins of the adult are comparatively inconspicuous. It is better on this account to take as an illustration of an adult wing that of a coreid, Harmostes reflexulus, in which the tracheae are distinctly visible within well-developed veins ; the courses of the veins are indicated in Fig. 304. Fig. 304. — Fore wing of an adult coreid, Homostes reflexus (After C. & N.). The median furrow of the wing is in its typical position between radius and media. In the pentatomids that we have studied it is more closely parallel with the radius and extends across the radial sector, showing that its position is not determined by the courses of the veins. The anal furrow is in front of cubitus instead of in its more usual position, behind this vein. In fact, in all of the Heteroptera that we have examined, when an anal furrow is distinctly developed it is in front of the cubitus. The anal furrow is the line of division between the corium and the clavus; the median furrow when well-developed separates the embolium from the corium ; and the development of a nodal furrow limits the corium still more by cutting off the cuncus. CHAPTER XVIII THE WINGS OF THE DERMAPTERA The winged members of this order have four wings. The first pair of wings are leathery, very short, without veins, and when at rest meet in a straight Hne on the back, resembling the elytra of staphylinid beetles; the second pair are large, with radiating veins, and when at rest are folded both lengthwise and crosswise. The radiating veins extend from a point near the middle of the length of the wing (Fig. 305). When the wing is not in use that part over which these veins extend is folded in plaits like a fan, after which the wing is folded twice crosswise. The part of the wing traversed by the radiating veins is the greatly expanded anal area. The preanal area is much reduced and contains only two longitudinal veins; this area is quite densely chitinized. Fig. 305. — Hind wing of an earwig. The tracheation of a hind wing of a nymph is represented by Figure 306. There is a transverse basal trachea (Fig. 306, tb) connecting the costo-radial and the cubito-anal tracheae. The tracheae of the cubito-anal group all coalesce at the base forming a single large trachea, from which, however, the cubital trachea soon separates (Fig. 306, Ctt). The anal trachea is twelve-branched. In the wing of the adult there are three distinct ar.al veins (Fig. 305). The first anal vein is ten-branched ; eight branches fonn a group of radia- ting veins, and the other two branches arise separately between this group of radiating veins and the base of the wing. The trachea: in the wing of the nymph (Fig. 306) corresponding to the ten branches of the first anal vein (295) 296 THE WINGS OF DERMA PTERA are easily recognized, as they occupy exactly similar positions. The two remaining branches, those that arise near the base of the wing, are doubt- less the tracheae that precede the second and third anal veins respectively. The costo-radial group of trachese is represented by a single trachea ; this I believe to be the radial trachea. It is surely not the costal trachea, as the vein that it precedes is far from the margin of the wing of the adult; and of the other three veins of the costo-radial group of veins, the radius is the more persistent one in other orders of insects. In the adult wing the preanal area is divided into two parts by the nodal furrow (Fig. 305, nf), a basal part and an apical part. Fig. 306. — Tracheation of the hind wing of an earwig. The radius extends about one-half the length of the basal part of the preanal area ; it is connected with the cubitus by a cross-vein or a branch near the tips of the two veins. The cubitus arises from the first anal vein and extends transversely until it nearly reaches the radius and then extends longitudinally a little more than half the. length of the basal part of the preanal area. It is forked near its end, and the posterior fork is connected with the anal vein by a vestige of a vein, which may be the remnant of vein Cu-i. A striking feature of the adult wing is a series of intercalary veins alternating with the branches of the first anal vein. There is no indication of the presence of these veins in the wing of the nymph. Neither is there in the wing of the nymph any indication of the series of cross-veins that extends parallel with the margin of the wing and connects the alternating branches of the first anal vein and the intercalary veins. CHAPTER XIX THE WINGS OF THE COLEOPTERA The members of this order have four wings; the first pair of wings, which are called elytra, are greatly thickened and meet in a straight line down the back (Fig. 307) ; the second pair of wings are folded beneath the elytra when not in use. The great modifications in form and structure of the wings of the Coleoptera make the study of the wing-venation of these insects a difficult one. Correlated with the loss of the flight function of the elytra in the course of their change to thickened protective organs, there has been a reduction of their venation ; and in the case of the hind wings, the fact that in most cases they are transversely folded when at rest has resulted in a great modification of the courses of the veins and in the formation of secon- dary vein-like thickenings of the wing. In spite of the difficulties just referred to, the problem of determining the homologies of the wing-veins of the Coleoptera is not an insuperable one if use be made of the ontogenetic method of study. In the pupal wings, so far as they have 1 -1.1 • • 1 . 1 Fig. "^OT. — Desmoceriis been exammed, the pnncipal tracheae are com- "^ palUatus paratively well-preserv^ed ; and a study of these tracheae furnishes data for detennining the homologies of the wing-veins. The beginning of the application of this method of study to the wings of the Coleoptera was made by Comstock and Needham ('98). Our dis- cussion of the matter, however, was brief and devoted largely to a demon- stration of the fact that the elytra are modified wings and not the greatly enlarged paraptera of the mesothorax, as was believed by Meinert ('80). The most conclusive evidence that we found regarding the homology of the elytra was presented by their development. We traced the develop- ment of the two pairs of wings in a coccinellid beetle, Hippodamia ij- punctata; and found that previous to their emergence from the larv^al wing- pockets, there is no appreciable difference between the fore and the hind wings; after this, however, the elytra show a distinctly thicker layer of h^'podermis on their dorsal side, and the thinness of the hind wings steadily increases with their expansion in area. The hind wings are greatly expanded at their final transformation, while the elytra are almost as large in the pupa as in the imago (C. & N. '99, p. 850). Ftirther evidence is the fact that a very close correspondence exists between the tracheation of the elytra and that of the hind wings; and what (297) 298 THE WINGS OF COLEOPTERA is especially striking is that similar modifications occur in the two pairs of organs. We figured the tracheation of the wings of two cerambycid pupse ; these figures are repeated here (Fig. 308 and 309). "In the species represented by Figure 308, the radial trachea is the most prominent one in both elytra and hind wings. On the other hand, in Fig. 308.- — Tracheation of the wings of a cerambycid pupa (From C. & N.). the species represented by Figure 309, the radial trachea is reduced in both elytra and hind wings to a mere vestige. If the elytra and hind wings were not homodynamous organs, it is not probable that the modifications of the two would be so closely correlated. "In comparing the tracheation of the elytra with that of the hind wings, the most striking difference obsen^ed is the greater reduction of the anal ■^^ Fig. 309. — Tracheation of the wings of a cerambycid pupa (From C. & N.). THE WINGS OF COLEOPTERA 299 area of the former. This is doubtless due to the fact that the meeting of the elytra, when at rest in a straight line along the middle of the back does not admit of an expanded anal area. "The extent of the correspondence between the venation and the tracheation of the hind wing of a full-grown pupa is shown by Figure 310. The principal tracheae are within the veins, but the branches of these tracheae extend irregularly through the wing. In the region where the wing is to be folded the secondary vein-like thickenings are only partly supplied with tracheae." (C. & N.) Contemporaneous with the publication of the series of articles by Comstock and Needham, Kruger ('98) presented to the University of Gottingen a thesis on the development of the wings of insects with especial reference to the elytra of beetles. This author studied the development of the wings of Tenehrio molitor and of two species of Lema; and, notwith- standing the fact that he found that the hind wings and elytra arise simul- taneously and develop in an exactly similar manner for a major part of the Fig. 310. — Hind wing of a pupa of a beetle (From C. & X.). lar\-al life, reached the unwarranted conclusion that the elytra are divergent structures and not specialized wings. This conclusion was based on differ- ences in structure between the fore wings and hind wings that appear late in their development and which are correlated with the specialization of the elytra as wing covers. Since the publication of the papers referred to above, several papers have appeared in which the development of the wings of Coleoptera are discussed; these are by Needham ('00), Tower ('03), and Powell ('04)- As these all confirm our conclusion that the elytra are modified wings, and as none of them discuss the homologies of the wing veins, it is not necessar}^ to review them here. The elytral tracheation of the tiger beetles (Cicindelidae) has been described in two papers by Shelf ord ('13 and '15). These studies are based on examinations of adult dried elytra. In only two genera {Amblychila and Mantichora) were all of the six principal tracheae found. Of course the basal connections of the tracheae could not be seen. 300 THE WINGS OF CO LEO PT ERA An extended series of papers on the venation of the hind wings of Coleop- tera has been pubHshed by Kempers ('gg-'og). This contains excellent figures of the venation of the hind wings of representatives of many families. But as these studies are based entirely on studies of the wings of adult insects, it is desirable that the conclusions regarding the homologies of the wing-veins should be confirmed by ontogenetic studies. There has recently appeared, however, a contribution to our knowledge of the tracheation of the wings of Coleoptera based on a study of the wings of pup£e ; this is by Kiihne ('15). This author gives figures of the trachea- tion of the fore wings of pupse of four genera of beetles and of the hind wings of pupae of seven genera; but in only one genus (Cantharis) are the basal connections of the principal tracheae shown. In this genus the radial trachea has not followed the medial trachea in its migration along the transverse basal trachea towards the cubito-anal group of trachege to so great an extent as Comstock and Needham found it had in the cerambycid pupal wings represented by Figure 308. This is markedly the case in the hind wings of Cantharis as figured by Kiihne; but in one of the fore wings of Cantharis figured by this author, his Figure 7, the radial trachea has evidently begun its migration towards the cubito-anal group of tracheae. With this limited amount of data before him Kuhne concludes that the labeling of the tracheae of cerambycid pupae given by Comstock and Needham is an incorrect one, and that the radial trachea always remains a member of the costo-radial group of tracheae. When we consider the great differences in the extent of the migration of trachea R within the order Orthoptera, as is shown in Chapter VII, one is not warranted in making generalizations regarding the conditions in the Coleoptera after studying the basal connections of the tracheae in the wings of a single genus of beetles. As yet I see no reason for changing the labeling of the tracheae in Figures 308 and 309. It is evident from this review of the literature of the subject that much remains to be done before our knowledge of the homologies of the wing- veins of the Coleoptera can be regarded as fiiTnly established. The solution of the problem is rendered difficult by the fact that in no coleopterous pupa yet examined have the principal wing-tracheae retained the mode of branching characteristic of the hypothetical primitive type; in most cases the tracheae extend in nearly direct lines with only small, irregularly arranged branches. It seems also to be made more difficult by another factor, namely, that the observations that have been made indicate that in the Coleoptera, as in the Hymenoptera, the venation of the wings precedes their tracheation. The courses of the tracheae, therefore, are determined by the nature of the highly modified venation. I CHAPTER XX THE WINGS OF THE STREPSIPTERA rst A ^ 2d A Fig. 311. — Wing of Paraxenos eheri (From Pierce after Saunders). In this order the fore wings, which are termed elytra by some writers and pseudo-halteres by others, are reduced to slender club-shaped append- ages. The hind wings are large, compared with the size of the tiny body, fan-shaped, furnished with radiating wing-veins, and folded longitudinally. The venation of the wings is degen- erate. There is a variable number of radi- ating veins, which in the most generalized wings are eight in number. These are supposed, by Pierce ('09), to be the eight principal veins of the typical wing, the costa, subcosta, radius, media, cubitus, and the three anal veins, respectively, and are so designated by him in his most excellent monograph of the order. With our lack of knowledge of the tracheation of the wings, this conclusion can hardly be questioned. There are no cross- veins. The veins are rarely forked ; but there are usually detached veins in the outer half of the wing. These extend longi- tudinall}' in the area between the radius and the media (Fig. 311). The detached veins are usually one or two in number. Pierce makes no suggestion as to the homology of these veins; but the arrangement of the veins in the wing of Acroschismus huhhardi (Fig. 312) leads me to believe that the first detached vein is the radial sector; and the second one, a branch of media. As this con- clusion, however, is not incontrovertible, it will probably be best, in descriptions of species, merely to state the nimiber of detached veins between the radius and the media, without attempting to determine their homologies. Fig. 312. — Wing of Acros- chismus hubbardi (From Pierce) . (301) CHAPTER XXI THE WINGS OF THE MECOPTERA (a) THE MORE GENERAL FEATURES OF THE WINGS OF THE MECOPTERA. In nearly all of the winged members of the Mecoptera the wings are long, narrow, membranous, and furnished with a considerable number of cross-veins. The two pairs of wings are similar in form and nearly equal in size ; the hind wings are usually a little shorter than the fore wings. In many species the wings are conspicu- ously spotted or banded (Fig. 313). In a few forms the wings are either absent or vestigial. In two genera Merope and NoHothauma, the represent- atives of which are very rare insects, the wings are com- paratively broad. Fig. 313 — Panorpa. (b) THE VENATION OF THE WINGS OF THE TYPICAL MECOPTERA Our conclusions regarding the homologies of the wing-veins of the Mecoptera are based entirely on a study of the wings of adults; for, as yet, the tracheation of the wings of pupae has not been observ^ed. It is not probable, however, that a study of pupal wings would modify these con- ^j__^. J^W ?d^ ' ist A Cu Cm Mi, Cui Mi, Mi Fig. 314. — The wings of Panorpa. (302) THE WINGS OF MECOPTERA 303 elusions, as the venation of the wings is so nearly typical that the identifica- tion of the veins presents no difficulties. In some cases, as in the fore wings of certain species of Panorpa, the number and the arrangement of the wing-veins is that of the hypothetical primitive type, with the addition of a considerable number of cross-veins; and in all cases, except- ing the aberrant genera Merope and Notiothamna, the modifications of this type are comparatively slight. The most obvious modification of the prim- ^. ^ c u- a ■ ^t -d^.,^,^^ ^ Fig, 315. — Base of a hind wing of i^OMorpa. itive type is the presence of one or two accessor}^ veins on one or more of the branches of the radius. These are borne most frequently by vein Ro. Another modification of the primitive type that exists commonly is the anastomosing of one or of both branches of the cubitus with the adjacent vein ; that is vein Cui anastomoses with vein M, and vein Cuo with the first anal vein. In the fore wing of Panorpa (Fig. 314) each of the principal veins extends separately from the base of the wing; but in the hind wing each of the branches of the cubitus anastomoses with the adjacent vein. This is shown more clearly in the enlarged figure of the base of the hind wing (Fig. 315). The relation of the branches of the cubitus to the adjacent veins in the three common winged genera was pointed out by Miyake ('13). In the fore wings of Panorpa and of Panorpodes the principal veins extend separ- ately from the base of the wing. In the fore wing of Bittacus, veins Cui and M anastomose, but vein Cu2 is separate from the first anal vein. In the hind wings of all of these genera, each of the branches of the cubitus anastomoses with the adjacent vein. (c) THE ABERRANT MECOPTERA There are two genera of insects which, although they differ greatly in appearance from the typical members of the Mecoptera, are doubtless members of this order. These are Merope and Notiothaurna. The genus Merope is represented by a single species, Merope tuber which was described by Newman in 1828 (Ent. Mag. V. p. 180). Newnnan was unable to decide as to the zoological position of the genus. Later West- wood ('41), from a study of the mouth parts decided that it should be placed in the Panorpida;. 304 THE WINGS OF MECOPTERA Merope tuber is an American insect. Although it is quite widely dis- tributed in the Atlantic States, less than a score of individuals are known to exist in collections. Professor Needham had the good fortune to collect one at Ithaca; and I am indebted to him for an opportunity to study its wings and to give a figure of the insect (Fig. 316). A striking feature presented by the wings is the presence of a minute semicircular tuberculous appendage near the base of the inner margin of the fore wing. It was this that suggested the specific name of the species. The wings of Merope tuber (Fig. 317), at first sight, appear to be quite different from those of the typical Mecoptera. But when they are exam- ined in detail it is found that, aside from the fact that they are less elongate and that they bear an unusually large niimber of cross-veins, the only important difference is that the costal area of the wings is broad and resembles this area of neuropterous insects more than it does that of Pan or pa. This resemblance is increased by the presence of many cross-veins extending from the subcosta to the costa and a smaller nimiber of cross-veins between the subcosta and vein R]. The presence of one or two accessory veins on one or more of the branches of the radius and the anastomosing of cubitus with the adjacent veins, which were pointed out above as characteristic of the Mecoptera, are also to be found in this genus. In the Ithaca specimen, vein R2 of all of the wings bears a single accessory vein. In the specimen figured by Westwood, vein R2 bears two accessory veins in the fore wing and only one in the hind wing. But Dr. Asa Fitch, who had both sexes of this species, states in his "Fourteenth Report," that both pairs of wings are liable to vary in the number of these branches. In fact in his female specimen the fore wing of the left side and the hind wing of the right side had each one more branch than the corresponding wing of the other side. In the fore wing of the specimen taken at Ithaca cubitus and the first anal vein coalesce for a short distance. Immediately after vein Cu separ- ates from I St A, Cui extends transversely to the long axis of the wing and anastomoses with vein M for a considerable distance. Vein Cui bears one accessory vein on the right wing and two on the left wing. The anal furrow is along the first anal vein. There are three anal veins, all unbranched. Fig. 316. — Merope tuber, slight- ly enlarged (Photographed by J. G. Needham). THE WINGS OF MECOPTERA 305 In the hind wing of this specimen vein Cui anastomoses with vein M4 for a short distance. Vein Cui bears one accessor}- vein on the right side, but is unbranched on the left side. Dr. Fitch gave the popular name Earwig-fly to this insect on account of the stout pair of forceps at the caudal end of the abdomen of the male. Fig. 317. — Wings of Merope tuber. The genus Notiothamna was established by McLachlan ('77) to receive a remarkable insect from Chili , which he named Notiothamna Reedi. Only a single mutilated individual was known, the head and pronotum of which were nearly entirely destroyed. But the form of the wings and the presence of a well-developed, nearly semi-circular lobe at the base of the inner margin of the fore wings indicated that the species was allied to Merope tuber. McLachlan gave a figure of his specimen, which was carefully copied by Brongniart ('93). I reproduce Brongniart's copy, as it is better suited for reproduction by photoengraving than the original (Fig. 318). 306 THE WINGS OF MECOPTERA The wings of Notiothauma resemble those of Merope in their more general features; but they present a much more intricate network of cross- veins; and in some parts of the wing it is difficult to distinguish the longitu- dinal veins from the cross-veins. A similar condition exists in the wings of certain Plecoptera of which Pteronarcys dorsata (Fig. 254) is a good example. Fig. 318. — Notiothauma Reedi (After McLachlan by Brongniart). CHAPTER XXII THE WINGS OF THE TRICHOPTERA The order Trichoptera as heretofore recognized includes those insects the larvffi of which are commonly known as caddice-worms and the adults as caddice-fiies. This is a well-defined, homogeneous group of insects. There is, however, a group of moth-like insects that have been included in the order Lepidoptera which so far as the structure of their wings is con- cerned and in some other respects are more closely allied to the Trichoptera than they are to the Lepidoptera, this is the Micropterygina. To continue to include the Micropterygina in the Lepidoptera raises a question of phylogeny to which I can find no answer; while the transference of this family to the Trichoptera removes this difficulty. This phase of the subject is discussed later, in the concluding part of this chapter. Although the Micropterygina are closely allied to the caddice-flies, there are differences between the two groups that warrant the regarding of each as a distinct suborder. I therefore propose the division of the Trichoptera into two suborders, the Phryganeina or aquatic Trichoptera andtheMicrop- ten^gina or terrestrial Trichoptera. The distinction in habits between the Phryganeina and the Microp- terygina is not an absolute one. Among the Phryganeina the lan'S of the genus Enoicyla live under moss at the foot of trees, chiefly in woods and often at great distance from water; and, on the other hand, the known lar\'as of the genus Micropteryx, as now restricted, might be considered as semi-aquatic since they live in wet moss. SUBORDER PHRYGANEINA The Aquatic Trichoptera The suborder Phryganeina includes those Trichoptera in which the wings are clothed with long silky hairs, the tracheation of the wings of the pupa is reduced, and the larvae are aquatic. (a) THE MORE GENERAL FEATURES OF THE WINGS OF THE PHRYGANEINA The two pairs of wings are membranous and usually more or less densely clothed with long silky hairs. The fore wings are denser than the hind wings and are often slightly coriaceous ; in a few forms the wings are naked. The hind wings are shorter than the fore wings; but they are usually broader; this is due to an expansion of the anal area of the hind wings. In a few- species the hind wings are reduced so that they are smaller than the fore (307) 308 THE WINGS OF TRICHOPTERA wings; in one species the female is apterous and in another the wings of the female are vestigial. When not in use the wings are folded roof-like over the abdomen. The posterior lobe of the fore wings is specialized as a fibula, which is well-developed in the more generalized forms, as Rhyacophila, but more or less reduced in the more specialized genera. The costal border of the hind wings is furnished with hamuli in some forms, as in the Leptoceridae and in the Macronematinse. {b) THE TRACHEATION OF THE WINGS OF THE PHRYGANEINA It was shown by Comstock and Needham that the tracheation of the wings bears but little relation to the wing-venation; this being one of the orders in which the trachea- tion of the wings is greatly reduced. If a wing of a pupa of a caddice-fiy be examined after the cavities of the developing wing-veins have been formed it will be seen that usually only two or three main trachea are T.. „r. r r , -,■ r, present; and that although Fig. 319. — Wmg of a pupa of a caddice-flv. . • ■ -, • , these may comcide with forming veins, their branches bear no relation whatever to the veins (Fig. 319). (c) THE PHRYGANEID TYPE OF WING-VENATION As the tracheation of the wings of pupae of the aquatic Trichoptera affords no help in the determination of the homologies of the wing-veins, we are forced to base our conclusions regarding these homologies on studies of the wings of adults. Fortunately in the more generalized members of this suborder it is easy to identify the veins; and the conclusions that have been reached regarding the homologies of the wing-veins of the Phry- ganeina are confirmed by a study of wings of ten-estrial Trichoptera in which the wing-trachea are presei^ved. The wings of Rhyacophila fuscida (Fig. 320) can be taken as illustrating the phryganeid type of wing-venation. Beginning at the costal margin of the jore wing and proceeding backward the following features can be observed. The subcosta of the fore wing is forked ; the forking takes place near the tip of the vein; and vein Scs is connected with vein Ri by a cross vein. Near the middle of the length of the subcosta, an accessory vein (Fig. \ THE WINGS OF TRICIIOPTERA 309 320, a) is given off, which extends to the margin of the wing. The humeral vein is present in its usual position. The radius of the fore wing is typical; the radial fork is quite near the base of the wing; and the radial sector is dichotomously four-branched. The media is also typical, being dichotomously four-branched. The cubitus and the first anal vein coalesce at the base, where they traverse the cubito-anal sulcus. From the point where these two veins separate vein Cu extends in an oblique direction towards media, which it nearly reaches, and then making a bend extends in a longitudinal direction to the outer margin of the wing. It is of the typical two-branched form. /?. /?. 2d A islA Oi2 ^"' Fig. 320. — Wings of Rhyacophila fiisctda. The basal part of the free portion of vein Cu appears very much like a cross-vein. This appearance is accentuated by the fact that a serial vein has been developed, which consists of three parts, the basal part of media, a short cross-vein connecting media and vein Cu (Fig. 320, pa), and the longitudinal part of vein Cu. This results in the longitudinal part of vein Cu appearing to be a nearly direct continuation of the stem of media. The short cross-vein connecting media and cubitus corresponds to the posterior arculus of those orders in which an arculus is developed. The separate portion of the first anal vein is a direct continuation of vein Cu -|- I St A and extends in the bottom of the cubito-anal fold to the margin of the wing. The outer portions of the second anal vein and of two branches of the third anal vein coalesce, forming a single vein, which ends 310 THE WINGS OF TRICHOPTERA in the margin of the wing near the tip of the first anal vein. A third branch of the third anal vein supports the margin of the posterior lobe of the wing, the fibula. In some species of Rhyacophila the terminal portion of this third branch of the third anal vein extends across the fibula to the axillary- furrow. The disposition of the anal veins of the fore wings is one of the most distinctive characteristics of the order Trichoptera. In the hind wing the subcosta resembles the subcosta of the fore wing except that it bears no accessory vein. The radius also resembles the corresponding vein of the fore wing. The media is only three-branched, veins M3 and M4 coalescing to the margin of the wing. The cross-vein connecting media and cubitus near the base of the wing, the posterior arculus, is transverse. Veins Cu and ist A coalesce at the base of the wing for a much shorter distance than in the fore wing. The anal A-eins differ from those of the fore wing in that they all end separately in the margin of the wing. There is one feature in the disposition of the anal veins of the hind wing which, like the coalescence of the tips of anal veins in the fore wing, is a distinctively ordinal characteristic; this is the course of the second anal vein. This vein, near the base of the wing, extends forward until it reaches the first anal vein, with which it anastomoses for a considerable distance; it then bends away from the first anal vein abruptly and extends obliquely until it is joined by a cross-vein, it then bends again and extends longitudin- ally to the margin of the wing. The cross-vein connecting the second and third anal veins is nearly longitudinal. The interpretation of the homologies of the wing-veins given above differs in one respect from that which has been commonly accepted. This is the recognition of the fact that in the fore wings veins Cu and ist A coalesce at the base of the wing and what appears to be an oblique cross- vein between veins Cu and ist A is really the base of the free part of vein Cu. This conclusion does not necessitate any change of view regarding the homologies of the terminal portions of cubitus and the first anal vein ; con- sequently the lettering of figures of wings that is placed at the tips of veins will remain as before. The necessity for reopening the question of the homologies of the wing- veins of the Trichoptera was suggested by Dr. Cornelius Betten, who reached the conclusion that the vein designated here as vein Cu -|- i st A is the cubitus and that designated as the first anal vein is vein Cu2. This necessitated the conclusion that what heretofore has been regarded as vein Cu2, and which I still believe to be this vein, is a definitive accessory- vein, which Dr. Betten terms vein Cuia- Although Dr. Betten's discussion of this question is not yet published, he has ptiblished a figure in which the veins of the fore wing of a species of \ THE WINGS OF TRICHOPTERA 311 Rhyacophila are lettered in accordance with this interpretation (Betten '13.) During my studies of this subject Dr. Betten has aided me in every possible way, placing in my hands portions of his manuscript and figures and many mounted wings. I am under great obligation to him for this assistance. {d) THE MORE GENERAL FEATURES IN' THE SPECIALIZATION OF THE WINGS OF THE PHRYGANEINA The wings of Rhyacophila Juscnla (Fig. 320) probably resemble very closely the wings of the primitive Trichoptera, as they resemble in their more general features the hypothetical primitive type of insect wings. The more important modifications of this type are the following. In the fore wing the tips of the second anal vein and two of the branches of the third anal vein coalesce. This is a distinctively characteristic feature of the wing-venation of the Trichoptera. The subcosta bears an accessor}' vein; this, however, is unimportant; accessory veins borne by SC^ Si2 Cu2 Cu\ Fig. 321. — Fore wing of Rhyacophila sp. the subcosta exist in only a few genera of this order; in some there are several of these veins, as in the Japanese genus Perissonetira, figured by Ulmer ('07). The coalescence of veins Cu and ist A at the base of the wing and the formation of the serial vein consisting of the base of media, the posterior arculus, and the distal part of vein Cu is an ordinal characteristic. In the hind wings media has been reduced to a three-branched condition by the coalescence of veins M3 and AI4. By comparing more specialized forms with Rhyacophila fnscnla it will be seen that in the preanal area the specialization of the venation of the wings is always by reduction. In the anal area of the hind wings the speciahzation is in some cases by addition, resulting in a broadly expanded anal area, in others it is by reduction. It is unnecessary to indicate, in this place, the methods of specialization of the wings in the different families of this order, as this has been done by Ulmer, Betten, and others. 312 THE WINGS OF TRICHOPTERA There is, however, one method of speciahzation, which has been eluci- dated by Dr. Betten and to which he has called my attention, that should be referred to here. It is the modifications of the basal part of the free Sc Fig. 322. — Fore wing of Hydromaniciis dilatus. portion of vein Cu of the fore wings, that part which appears to be an oblique cross-vein. In the hind wing of Rhyacophila Juscula veins M and Cu are connected near the base of the wing by a transverse cross- vein (Fig. 320). In the fore wing of this species this cross-vein has become longitudinal and forms a part of the serial vein M — Cu. In a species of Rhyacophila from India, in Dr. Betten's collection, this cross-vein has been obliterated by the anastomosis of veins M and Cu (Fig. 321), and the basal part of vein Cu simulates a cross-vein so completely that but for the evidence presented by forms in which veins M and Cu do not anastomose its identity would not be sus- pected. This is the usual condition of the base of vein Cu in the fore wing. More remarkable still is the fact, also discovered by Dr. Betten, that after the basal part of vein Cu is transfoimed into a cross-vein, it may migrate from its first position. Thus m Hydromaniciis (Fig. 322) it has moved a considerable distance toward the outer margin of the wing, following along the longitudinal part of Cu. This results in the posterior arculus becoming great- ly lengthened. {e) THE METHODS OF UNITING THE TWO WINGS OF EACH SIDE Fig. 323-- -Fibula of Rhyacophila fuscula. In RhyacopJiila the posterior lobe of the fore wing fonns a well-developed fibula (Fig. 323). This is hatchet- shaped and supported by the third branch of the third anal vein; in some individuals the third branch of the third anal vein extends along the free margin of the fibula (Fig. 320), in others it is curved so as to extend across i THE WINGS OF TRICHOPTERA 313 the fibula (Fig. 323). The axillary furrow is immediately behind the second branch of the third anal vein. It is evident that this fibula is fitted to clasp the anterior tuberosity of the hind wing. There are two facts that indicate that this is the most primitive form of fibula that exists among the living Phryganeina: first, the generalized condition of the wing-venation of Rhyacophila; and second, the fact that this fibula is identical in structure with that of Mnemonica, one of the most generalized of the terrestrial Trichoptera. It is evident that Rhyacophila and Mnemonica resemble in the venation of their wings and in the form of their fibulae the stem form from which the Phn-ganeina and the Microp- terj'gina have been evolved. When more specialized Phrs'ganeina are studied it is found that there is a marked reduction in the size of the fibula and a modification of its form. In Hydromanictis, for example, the posterior lobe of the fore wing (Fig. 322) is greatly reduced in size and has little appearance of being a fibula. It is evident that the uniting of the two wings of each side is being attained in other ways than by a fibula alone. In some cases this is by the overlapping of the wings; in some cases as in Phanostoma and Leptocerus there is a series of prominent hamuli on the costal margin of the hind wing; and in others, as in Gcera, there are strong spines borne at the humeral angle of the hind wing, which probably function as a frenulmn. SUBORDER MICROPTERYGINA The Terrestrial Trichoptera The suborder Micropteiygina includes those Trichoptera in which the wings are clothed with scales, the tracheation of the wings is preserved, and the lar\'se are not aquatic; it is represented b>' a single family, the Micro- pterygidte.* The Micropterygidas is a small family of minute moth-like insects which are generally believed to belong to the order Lepidoptera, although several writers have suggested the close affinity of these insects to the Trichoptera. f *The genus Microptcryx and its allies were long considered as constituting a single family the Micropterygida;. In 1894 Chapman proposed the separation of this family into two families, the Alicropterygidfe and the Eriocephalida; (Trans. Ent. Soc. Lond. 1894, p. 336). In 1912 Meyrick in his monograph of this group (Genera Insectorum, Fascicule 132) regards it as constituting a single family, the Micropterygidae, which he divides into three subfamilies as follows: the Mnesarchajinac, which includes a single genus, MnesarchiEa, represented by three species in New Zealand; the Eriocraniinae, which corresponds to the Micropterygidae of Chapman; and the Micropteryginae, which corresponds to the EriocephaHdae of Chapman. This shifting of family or subfamily names is due to the conclusion that the generic name Eriocephala is a synonym of Micropteryx. For the sake of simplicity I follow Meyrick in regarding the Microptery- gina as including a single family, although it is highly probable that ultimately the mandibulate and liaustilate members of this group will be placed in separate families. fN'ote especially the remarks of Dr. Sharp and Mr. McLachlan (Proc. Ent. Soc. London, 1896, p. XVII), and Dr. T. A. Chapman (Trans. Ent. Soc. London, 1896, p. 569). 314 THE WINGS OF TRICHOPTERA (a) THE VENATION OF THE WINGS OF THE MICROPTERYGID.^ The family Micropterygidse has been monographed recently by Mey- rick ('12), who recognizes eight genera represented by fifty-five species. This author calls attention to the resemblance of some of the more general- ized forms, as Sabatiiica, to Rhyacophila; but does not suggest the trans- ference of this family from the Lepidoptera to the Trichoptera. The wings of a species of Mnemonica (Fig. 324) can be taken as illus- trating the more generalized type of wing-venation found in the Microp- terygidse. The fore and hind wings are quite similar in form and venation. The subcosta is distinctly forked in the fore wing and slightly so in the hind wing. Vein Ri in both wings bears an accessor}^ A^ein, vein Ri^; the pres- Sc, Sr. ^i 2d A istA Cu2 <^«' Fig. 324. — Wings of Mnemonica sp. ence of this accessory vein is limited to the more generalized^members of the family. The radial sector is four-branched in both fore and hind wings; in the fore wings of the individual figured here vein R3 arises from vein R4+5, but this is an exceptional feature; usually the forking of the radial sector is dichotomous. The media is reduced to a three-branched condition in both fore and hind wings, by the coalescence of veins M3 and M4. In the fore wings veins Cu and ist A coalesce at the base of the wing; the base of the free part of vein Cu appears like a cross- vein; and a serial vein is formed by the base of media, the posterior arculus, and the longitudinal part of the cubitus. In the fore wings the tips of the second and of the first two branches of the third anal vein coalesce, forming a single vein, which reaches the margin of the wing near the tip of the first anal vein. i THE WINGS OF TRICHOPTERA 315 There are two features of the venation of these wings that are of especial interest: first, the course of cubitus in the fore wing, described above; and second, the course of the second anal vein of the hind wing, which anastomoses with the first anal vein near the base of the wing, in the same manner as in Rhyacophila, described on an eariier page. The fibula of Mnemonica (Fig. 325) is identical in structure with that of Rhyaco- phila (Fig. 320). It is hatchet-shaped, the axillary furrow is immediateh' behind the second branch of the third anal vein, and the longitudinal margin is supported by the third branch of the third anal vein. Fig. 325. — Fibula of Mnemonica sp. (b) THE TRACHEATIOX OF THE WINGS OF THE MICROPTERYGINA A study of the tracheation of the wings of Mnemonica confirms the conclusions reached regarding the homologies of the wing-veins of this insect and of those of Rhyacophila* In Figures 326 and 327 which represent the basal portion of the wings of Mnemonica, the tracheae are shown in the wing-veins in so far as they can be clearly distinguished in the specimen figured. In the fore wing, no costal trachea was found. There is a single trachea in the subcosta, which is not clearly visible except in the basal fourth of the vein, where it is distinct. At the base of the radius there is a single trachea, but this forks just before the point where subcosta and radius diverge. At about one-third of the distance from this point to the radial fork, one of these two tracheae again divides. A little farther distad the other of the tw^o branches divides. All four of these branches pass into the radial sector. 1 am unable to find any trace of trachea Ri.f In the basal part of the stem of media there is a single trachea; this divides into two a con- siderable distance before the cross-vein that extends to vein Cu {p a) is reached; one of the two again di\'ides a short distance beyond this cross- vein; and the other divides into two a little farther on; all of these tracheae *I am indebted to Dr. W. T. M. Forbes for an opportunity to study a preparation of the wings of a species of Mnemonica, in which the tracheae can be clearly seen in the basal portions of most of the wing- veins. Dr. Forbes placed this preparation in my hands with the suggestion that the tracheation of these wings indicates the necessity of a modification of the commonly accepted view regarding the homologies of some of the wing-veins of the Microptcrygidse. A careful study of this specimen has fully confirmed Dr. Forbes' suggestion. fA similar forking of trachea; near the base of the wing which results in the stem of a forked vein containing several parallel trachea;, one for each of its branches, occurs in the Sesiidas, and in some hepialids, at least. 316 THE WINGS OF TRICHOPTERA become invisible before the medial fork is reached. The cubital and first anal tracheae coalesce for a short distance at the base of the wing and then separate; the cubital trachea divides into two a short distance before the Fig. 326. — Base of the fore wing of Mneinonka. point where vein Cu separates from the first anal vein and both branches follow the Z-shaped course of this vein. There is no trachea in the pos- r«+ Fig. 327. — Base of hind wing of Mnemonica. terior arculus. The first anal trachea I am unable to trace beyond the point where vein Cu branches off, owing to the atrophied condition of the first anal vein. The second anal trachea is clearly visible. THE WINGS OF TRICIIOPTERA 317 In the hind wings the cubital and first anal trachea? coalesce for a short distance, but the two soon separate; they extend parallel in the coalesced vein until the point is reached where the two veins separate and then each follows its vein. The base of the free part of vein Cu makes a sharp cuxxq towards the base of media but does not quite reach it, the two veins are connected by a short posterior arculus in which there is no trachea. Directly opposite the point of separation of the cubitus and the first anal veins, the first anal vein is joined by the second anal vein, the trachea of which is perfectly distinct. The tracheae of the two veins continue parallel until the two veins separate and then the second anal trachea follows its vein. (c) THE ZOOLOGICAL POSITION OF THE MICROPTERYGID.5 If the wings of Mnemonica, one of the more generalized of the Alicro- pterygidae (Fig. 324), be compared with those Rhyacophila, one of the more generalized of the Phryganeina (Fig. 320), it will be seen that they agree in the more essential features of their venation ; the more striking character- istics of the one are presented by the other. These characteristic features are the following. In the fore wings, the coalescence of veins Cu and ist A at the base of the wing ; the Z-shaped course of vein Cu ; the formation of a serial vein consisting of the base of media, the posterior arculus, and the longitudinal part of ^•ein Cu ; the coalescence of the tips of the second anal vein arid of two of the branches of the third anal vein ; and the cross-vein between the first and second anal veins. In the hind wings, the coalescence of veins Cu and ist A at the base of the wing; the Z-shaped course of the cubitus; the anastomosis of the first and second anal veins; the longitu- dinal direction of the cross-vein connecting the second anal vein and the first branch of the third anal vein; and the form of the branching of the third anal vein. In addition to these common venational features the fibulse of the two insects are identical in structure. The possession of this remarkable series of common features of their wings by these representatives of the Phryganeina and the Micropterygina, and which is found in no insect not belonging to one of these two groups, can be explained only by assuming that it indicates a community of descent of the two groups. This conclusion is confirmed by the results of Dr. T. A. Chapman's studies of pupae.* For these reasons, the Micropterygina must be regarded as more closely allied to the Phryganeina than they are to any other group of insects; that is, they are obviously trichopterous insects. It is also obvious that the Phryganeina and the Micropterygina represent two quite distinct lines of descent from the common stem of the order and should be regarded as distinct suborders. *Trajis. Ent. Soc. London, 1896, p. 569. 318 THE WINGS OF TRICHOPTERA If the Micropterygidas be retained in the order Lepidoptera they must be considered the most generahzed members of the order, being near the stem form from which the Trichoptera and the Lepidoptera have been evolved. This view necessitates the explanation of the manner in which the Hepialidas, with their peculiar jugum, and the Frenatse were evolved from a form ha\nng a well developed fibula, like that of Mnemonica and Rhya- cophila. This must be done if the Lepidoptera, including the Micro- pterygidcC, is to be shown to be a monophylitic group. PLATE IX Ma + CUi 2d A '^^^ 3dA Ma + CUy -TdA The wings of Prionoxystus rohinlce. CHAPTER XXIII THE WINGS OF THE LEPIDOPTERA (a) THE MORE GENERAL FEATURES OF THE WIXGS OF THE LEPIDOPTERA In the Lepidoptera the wings are membranous and are covered with overlapping scales. They are usually broadly expanded; but in some of smaller forms they are long and narrow. When at rest, the wings are variously disposed in the different members of the order. In a few species the males are wingless. The wing-venation is comparatively simple ; the only puzzling features are due either to the coalescence of veins or to the atrophy of veins. If we except the humeral veins, which are described later, in no family is there normally a specialization of the venation by the development of either accessory or intercalary veins; and only the principal cross-veins are present. The most striking departure from the hypothetical primiti\'e type is the fact that media is only three-branched. In this respect this order agrees with the Diptera; but the Lepidoptera differs from the Diptera in its characteristic method of coalescence of veins. In this order the coalescence of veins almost invariably procedes outward ; while in the Diptera it often procedes from the margin of the wing towards the base of the wing. (b) THE CLOTHING OF THE WINGS OF THE LEPIDOPTERA The clothing of scales. — The most distinctive feature of the wings of the Lepidoptera is the coating of scales with which they are covered. This coating of scales is the dust that comes off upon one's fingers when a moth or butterfly is handled. When this dust is examined with a micro- scope it is found to be composed of very minute scales of various forms but regular in outline ; and when a wing is looked at in the same way, the scales are seen to be arranged with more or less regularity upon it (Fig. 328). Thebody, the legs, and other append- ages are also covered with scales. ' It is well-known that these scales are Fig. 328 —Part of a wing of a butterfly greatlv magnihed. merely modified setcC. That is, they are setae which, instead of growing long and slender as setae usually do, grow very wide as compared with their length. Every gradation in form can be found from that of the ordinary seta, which occurs most abundantly (319) 320 THE WINGS OF LEPIDOPTERA Fig. 329. — Scales of Eudea cippus (After Kellogg, '94). upon the body, to the short and broad scale, which is best seen upon the wings (Fig. 329). This fact was pointed out by Reaumur nearly two hun- dred years ago (Memoires V. I. 1734); and in more recent times the morphological identity of setae and scales has been established by studies of their development. This identity was inferred by Semper ('57) and Landois ('71). Schaffer ('89) point- ed out that both scales and hairs are evaginations of greatly enlarged hypoder- mal cells and figured one stage in the development of the scales. Mayer ('96) gave a complete account of the development of scales and illustrated his paper by most excellent figures of all stages of this development. The structure of scales is what would be expected from the fact that they are modified setae, the scales, like setae, being hollow; and the manner of their attachment to the cuticula of the body and its appendages is the same as that of the setae, each scale being provided with a pedicel which fits into a cup-like socket in the cuticula. A striking feature of the scales of Lepidoptera is the mark- ings that exist on their exposed surface. These may consist merely of many, very fine, longitudinal ridges (Fig. 329); or there may be series of transverse ridges between the longi- tudinal ones (Fig. 330). A cross section of certain scales indicates that the ridges are produced by foldings of the outer wall {i. e. the wall of the scale that is exposed when the scale is in place on the wing). Figure 331 represents cross sections of a scale illustrating this condition. In some scales, however, the lumen of the scale has been filled to a considerable extent by chitin, and the origin of the ridges is not so obvious. Fi^'. 3^0. -Scale The scales of the Lepidoptera were probably developed of Lycomorpha from that type of setae known as clothing hairs, and were Kellogg) primarily merely protective in function. This is doubtless their chief, if not only, function on most parts of the body, where they form a very perfect armor. THE WINGS OF LEPIDOPTERA 321 The development of ridges on the surface of scales adds greatly to their stiffness, and thus increases their efficiency as a protective covering, as the corrugations in the sheets of iron used for covering the sides of buildings adds to the stiffness of the metal. Upon the wings a covering of rigid scales would serve not merely to protect the wings but would tend to stiffen them, and thus arose a secondary function of scales which has resulted in the perfecting of their arrangement upon the wings in the more specialized members of the order as already indicated. There are great differences among the insects of this order regarding the regularity of the arrangement of the scales upon the wings. With some of the more generalized moths the scales are scattered irregularly over the surface of the wings. But if a wing of one of the more specialized butter- flies be examined with a microscope the scales will be found arranged in regular overlapping rows ; the arrangement being as regular as that of the scales on a fish or of the shingles on a roof. Figure 328 represents a small portion of a wing of one of the more specialized butterflies, where the ^--^-jj;;-;^.,;-;;?^;;;:?^^^ arrangement of the scales is most ,;K3;:^^'^^^"'--^v--''^^^^^:^;;^^^ii^ perfect. In the upper part of the figure the membrane is represented ^^g- 33i. -Cross-section of soiles^ of Par- ^ ^ nasstus smtntheus (After Kellogg). with the scales removed. Even in those insects in which a very perfect arrangement of the scales upon the wings has been attained, great differences in the degree of perfec- tion of this arrangement exists in the tv/o wings of the same side and in the different parts of the same wing. The arrangement is most perfect in those wings and in those parts of each wing that are subjected to the greater strain during flight. And is more perfect in swift flying species that it is in those of slow flight. The taxonomic value of these differences in the arrangement of the scales of the wings of the Lepidoptera and also of the different types of scales found in different divisions of the order was investigated by Professor Kellogg ('94), to whose extended account the reader is referred for a dis- cussion of this phase of the subject. A secondary use of the scales of the Lepidoptera is that of ornamenta- tion ; for the beautiful colors and markings of these insects arc due entirely to the scales, and are destroyed when the scales are removed. The various colors of insects, and of other animals are produced in quite different ways ; and classifications of these colors have been proposed based on the methods of their production. The literature of this subject is too extensive to be referred to in detail here. A most enjoyable, popular account is given by Professor Kellogg in his Aniericaii Insects (Kellogg '08, pp. 583-614) and a detailed analysis of the methods of the production of 322 THE WINGS OF LEPIDOPTERA color is given by Professor Tower in his Colors and Color-patterns of Coleop- tera (Tower '03). Following the classification of Tower the colors of the scales of the Lepidoptera may be either chemical, physical, or chemico-physical. The chemical colors are produced by pigments in the scales; the physical colors are produced either by reflection, refraction, or diffraction of light ; and the chemico-physical colors are produced by either a reflecting, refracting, or diffracting structure overlying a layer of pigment. There are also what Tower calls combination colors due to a combination of the causes just mentioned. As the production of colors by pigments is the most obvious method in nature, it is the one to which the colors of the Lepidoptera are commonly attributed. But it is now well-known that a large proportion of the most beautiful colors of these insects are either physical or chemico-physical; this is true of the various metallic and iridescent colors so commonly found in butterflies and many moths. Explanations of the methods of production of physical colors are given in text books on physics; it is, therefore, only necessary here to point out a feature in the structure of the scales of Lepidoptera that results in the pro- duction of these colors. This feature is the presence of the fine longitudinal striae described above. When the striae are very fine and close together they act in the same way as does a diffraction grating, producing the beauti- ful iridiscent colors. Kellogg ('94) found that on certain scales from a species of Morpho the striae were from .0007 mm. to .00072 mm. apart or at the rate of about 35.000 to the inch. The fact that certain colors are due to the way in which light is reflected from the scales can be shown by the following experiment. Place on the stage of a microscope the wing of a bright blue butterfly, and shade the specimen so that it is viewed only by transmitted light from the mirror of the microscope; when examined in this way the blue color will be absent. This is due to the fact that the light passing directly through the scales is not broken up and only the colors produced by pigment are visible. There is still another function of the scales of Lepidoptera; they may serve as the outlets of scent glands. As the scales that serve this purj^ose are found chiefly on the wings of males, they have received the special name of androconia, signifying male dust. Androconia are most readily found in the "brand" of the wings of males of the subfamily Pamphilinas, the skippers with a discal patch ; in the costal fold of males of the Hesperiinae, the skippers with a costal fold; and in the discal patch of the wings of certain Lycenidas, the blues. They occur, however, in many other situations. The androconia are of various shapes (Fig. 332); they are frequently fringed at the distal end, with each tip of the fringe finely divided. This is THE WINGS OF LEPIDOPTERA 323 probably a provision for insuring the rapid evaporation of the product of the scent gland. Androconia have been figured and described by many authors. Associated with the patches of androconia there are frequently covering scales of various forms. In many Lepidoptera the scales are lacking on portions of the wings; familiar examples of this are most of the Sesiidas and certain members of the SphingidoD. The clothing of fixed hairs. — In addition to the clothing of scales, Kellogg ('94) discovered upon the wings of Micropteryx, which at that time was believed to belong to the suborder jugate of the Lepidoptera, and Hepialns "a covering of very fine hairs differ- ing radically from the scales in size, arrange- ment, and mode of attachment to the membrane, and agreeing essentially with the fixed hairs of the Trichoptera." Kellogg applied the term fixed hairs to this type of clothing, as the hairs are directly continuous with the cuticula of the wing instead of being jointed at the base, as are setas. The fixed hairs are much smaller than are the scales. In Micropteryx unimaculella they average .005 mm. in length, and are dis- tant from each other at their bases a length approximately equal to the length of the hairs. The scales of M. unimaculella average from .1 mm. to .15 mm. in length. In Hepialus sylvinus, he found that the fixed hairs were about one-tenth as long as the scales. As Kellogg was unable to find the fixed hairs on the wings of any of the Frenatae examined by him he concluded that their presence was a distinctive character of the Jugataj. In a paper published a year later, Spuler ('95), who evidently had not seen Kellogg's paper, describes the fixed hairs under the name Stacheln, and indicates their occurrence not only in the Micropter\^gidffi and the Hepialidae, but also distributed over the entire wings in certain Tineids {Incurvaria, Adela, Nematois, Nemophora, and in the Nepticulidas, and in limited areas on the wings of certain other Lepidoptera, in what he terms the "Haftfeld" or holding area, on the lower side of the inner margin of the front wing. The "Haftfeld" of Galleria mellonella is especially mentioned. The fixed hairs are referred to by Busck ('14), in his paper On the Classification of the Microlepidoptera, as the aculei; and the Lepidoptera that are characterized by their presence on the wings are designated as the Fig. 332. — Androconia from the wings of male butterflies (After Kellogg). 324 THE WINGS OF LEPIDOPTERA Aculeatce. The introduction of these two terms in this connection is unfor- tunate on account of their long and very general use in another sense in works on the H^Tnenoptera. The term, fixed hairs, used by Kellogg is perfectly satisfactory. The mode of origin and development of the fixed hairs has not been studied. They may be merely elongated cuticular nodules; but Spuler states that they are hollow, which indicates a different mode of origin from that of the ordinary cuticular nodules. The fixed hairs are designated by Marshall ('15) as the small surface hairs. (c) THE METHODS OF SPECIALIZATION OF THE WINGS OF THE LEPIDOPTERA The Lepidoptera belong to the series of orders in which the number of wing-veins does not exceed that of the hypothetical primitive type, the divergences from this type being the result of specialization by reduction. Fig. 333. — Fore wing of a pupa of Pieris rapce. Neither accessory nor intercalary veins are normally developed in this order; and only the principal cross- veins are present. In some cases, how- ever, there appear to be four anal veins, which indicates that one of the anal veins is two-branched. That this was more generally the case in the primitive Lepidoptera than it is in recent forms is shown by the fact that, although there is only one anal vein in the fore wing of the adult Pieris, in the pupal wing there are three anal tracheae and the third is two-branched. (Fig. 333). Although accessory veins are never normally present in members of this order, abnormal specimens have been found, especially of hepialids, in which there are extra branches on the branched veins. Reduction in the number of veins is of frequent occurrence, and is the result cither of the coalescence of adjacent veins or of the fading out of veins. Reduction by the coalescence of veins is the more usual method of THE WINGS OF LEPIDOPTERA 325 reduction; and, as already stated, this coalescence almost invariably pro- ceeds outward. This method of the coalescence of the branches of a vein is often shown, byastudy of aseries of allied forms in which different degrees of it can be observed, the point of separation of two branches being nearer to the margin of the wing in successive forms until the margin is reached and a single vein remains where there were two before. There are also many instances where the reduction is due to the fading out of veins. Frequently where a vein has atrophied a vestige of it remains, either as a faint line in the position formerly occupied by the vein, or as a short fragment of the vein; but in other cases no trace of the lost vein exists. {d) THE PRIMARY DIVISIONS OF THE LEPIDOPTERA INDICATED BY THE STRUCTURE OF THE WINGS In papers published in 1892 and 1893 I called attention to the fact that in certain moths the two wings of each side are united by an organ borne by the fore wing which I termed the jugum and in all other Lepidoptcra the wings are united either by an organ borne by the hind wing, which had long been known as the frenulum, or by a substitute for the frenuliun, a greatly expanded humeral angle of the hind wing. The discovery of the fact that there are two distinct modes of uniting the wings during flight suggested the inference that in the primitive Lepi- doptcra the wings were united in neither way ; for it is not easy to see how one mode could have been developed from the other. Correlated with each of the two methods of uniting the two wings of each side there is a distinctive feature of the wing-venation. In those moths in which the wdngs are united by a jugum the venation of the two pairs of wings is similar ; in other Lepidoptera there is a striking difference in the venation of the two pairs of wings, due to the fact that in the hind wings the radius is greatly reduced. If in the primitive Lepidoptera there were neither a jugum nor a frenu- limi, and in some of the descendants of these primitive Lepidoptera there was developed a jugum and in others a frenukun, there arose in this manner two distinct lines of descent. These supposed two lines of descent I con- sidered of subordinal value and proposed the temis JugatcB and FrenatcB for them respectively. Suborder Jugatse. — The suborder Jugatae includes the descendants of those ancient Lepidoptera in which a jugum was developed and in which the venation of the two pairs of wings is similar (Fig. 334)- To these, the most obvious characteristics of the Jugatae, can be added the following: In the Jugatas the cubitus and the first anal vein of the fore wings coalesce at the base of the wing; the basal portion of the free part of cubitus has the appearance of being an oblique cross vein; and a 326 THE WINGS OF LEPIDOPTERA serial vein is fonned consisting of three parts, the basal part of media, what appears to be the posterior arculus, and the longitudinal part of cubitus. The suborder Jugat^e includes only a single family of moths, the Hepialidce. I formerly included the family Micropterygidse in the Jugatas, because I accepted the general belief that they are lepidopterous insects, and because they agree with the Hepialidas in the similarity of the venation of the fore and hind wings, and also because the posterior lobe of the fore wings serv^es as an organ for uniting the fore and hind wings. But a more Fig. 334. — Wings of a hepialid, Pieliis labyrinlhecus. careful study of this family has convinced me that it belongs to the order Trichoptera ; the reasons for this conclusion are set forth in the preceding chapter. Suborder Frenatae. — The suborder FrenatcC includes the descendants of those ancient Lepidoptera in which a frenulum was developed, and in which the venation of the two pairs of wings is different, owing to the reduction of the radius of the hind wings. The Frenatffi differ from the Jugatas in the fact that the cubitus and radius of the fore wings do not coalesce at the base and also in the fact that a serial vein M — Cu is not formed. THE WINGS OF LEPIDOPTERA 327 Fig. 335. — Wings of a hepialid, Pielus labyrin- theciis, seen from below. a, abnormal accessory vein. The suborder Frenats includes the greater part of the Hving Lepidop- tera. Within this suborder are found great differences in the form of the wings, in the details of wing-venation, and in the manner of uniting the fore and hind wings. But, in spite of this, it is a very homogeneous group; all of the variations in the struct- ure of the wings are modi- fications of a common type, which is represented by living members of the sub- order. {e) THE WINGS OF THE JUGATE In the wings of the Jugatae the posterior lobe of the fore wing is a slender, finger-like organ, the jugum, which aids in uniting the fore and hind wings; and the venation of the fore and hind wdngs is quite similar, the radial sector of the hind wing being four- branched, like that of the fore wing. The jugum. — The jugum is a peculiar type of the posterior lobe of the fore wing that is found only in the Hepialidse. It is slender and projects beneath the costal margin of the hind wing. As the greater part of the inner margin of the fore wing overlaps the hind wing, the hind wing is held between the two. This is shown in Figure 335, which represents the wings of one side seen from below. The jugum is behind one of the two principal branches of the third anal vein, which may be designated as vein 3dAi,and is supported by the second of these two branches, which may be designated as vein 3dA2 (Fig. 336). The tracheation of the wings of pupae. — The tracheation of the wings of pupa) of Hepialus thulc was studied by Dr. A. D. MacGillivray ('12), who gives figures showing the courses of the tracheae in the forming wing-veins of both the fore and hind wings. This is the only paper that has been published on the tracheation of the pupal wings of a member of the Jugatae. Fig. 336. — Jugum of a hepialid. 328 THE WINGS OF LEPIDOPTERA In several adult hepialids that I have studied the tracheae are visible in the wing- veins. The results of my studies of these specimens differ in some details from those reached by Dr. MacGillivray. It is quite possible that these differences are due to the fact that the material studied by Dr. MacGillivray had been preserved in formal and may not have been in good condition. It is obviously desirable that fresh pupal wings be studied. The most important of the differences referred to is the fact that in all cases where I have seen trachea in the wing-veins of hepialids what appears to be the posterior arculus is traversed by a branch of the medial trachea, which extends towards the margin of the wing in the longitudinal part of the cubitus, and parallel with the cubital trachea. I am unable to explain satisfactorily the presence of this trachea in this position. It may have originated as an adventitious trachea ; but if so, it has become a definitive part of the tracheation of the wing for it is as large as or larger than the cubital trachea. It may be the trachea of the missing vein M4, which has been separated from the other branches of the medial trachea. An analogous splitting back of a trachea is that of trachea R4+.5 in the butter- flies. I leave the problem until more data bearing on it is obtained. In some hepialids there is a forking of trachese near the base of the wing which results in the stem of a forked vein containing several parallel branches. In a wing of Sthenopis that I have examined there are two tracheae in radius at the base of the wing, and these divide into four a short distance beyond the humeral vein. The venation of the wings of hepiahds. — In the hepialids the subcosta is often forked; the radius of both fore and hind wings is of the hypotheti- cal primitive type ; media is reduced to a three-branched condition in both fore and hind wings ; and in the fore wings veins Cu and i st A coalesce at the base, as in the Trichoptera. The way in which media has been reduced in the hepialids is still an open question. Dr. MacGillivray, as a result of his studies of tracheation of Hepialus thule, reached the conclusion that the reduction is the result of the coalescence of veins M^ and M4 ; but had he found the branch of the medial trachea that traverses the posterior arculus he probably would have reached a different conclusion. Comstock and Needham concluded that the reduction of media had been attained by the coalescence of veins M4 and Cui, and figured the wings of an abnormal Sthenopis in which the tips of veins M4 and Cui are separate in the hind wing (Fig. 337), believing that this condition is atavistic. The value of this evidence is seriously questioned by Dr. MacGillivray ; but the presence of the trachea that traverses what appears to be the posterior arculus, described above, adds weight to it. Until this question is more definitely settled, it will be best to omit any reference to vein M4 as is done in Figure 334. THE WINGS OF LEPIDOPTERA 329 In the fore wings, veins Cu and ist A coalesce at the base, where they He at the bottom of the cubito-anal sulcus (Fig. 334)- From the point of separation of the two veins, the cubitus extends at first in an oblique direc- tion, appearing like a cross- vein; when it has nearly reached the media, it bends abruptly and extends longitudinally to the outer margin of the wing. The cubitus is joined to the media by a short vein, which appears to be homologous with the posterior arculus of the Trichoptera; but which, as it contains a branch of the median trachea, may be the base of vein M4, the more distal part of vein M4 coalescing with vein Cu. I have been unable to find anything corresponding to the posterior arculus in the Frenats; for Fig. 337. — Venation of an abnonnal individual of Sthenopis. this reason it seems probable to me that it has not been developed in the Lepdioptera; and that what appears to be the posterior arculus in the hepialids is a section of vein M4. Lest the similarity of the course of vein Cu in the fore wing of the hepialids to that of this vein in the Trichoptera be given undue weight, I call attention to the fact that in the fore wing of the Cicada (Fig. 269) where veins Cu and ist A are crowded together in the cubito-anal sulcus and coalesce there, the free part of vein Cu follows a similar course. There are three anal veins in the fore wings of the hepialids. The first anal vein is aligned with vein Cu + ist A and extends a short distance in the cubito-anal fold, but the distal part of it has atrophied. No distinct anal furrow exists in any of the hepialids that I have studied; there is 330 THE WINGS OF LEPIDOPTERA merely a cubito-anal fold. The second anal vein is apparently simple; but in an adult wing before me it contains two tracheae, one of which extends into the cross-vein between this vein and vein Cu + ist A. This may indicate that the second anal vein was formerly forked and that the cross- vein is a portion of one of the forks. The third anal vein is forked; the second branch of it supports the jugum, as already described. The posterior tuberosity of the fore wing is divided by the second anal vein; the elevated portions of this tuberosity are between veins; this is an unusual condition. The venation of the hind wings (Fig. 334) resembles very closely that of the fore wings except for some differences at the base of the wing. The vein which appears to be homologous with the posterior arculus but which may be the base of vein M., extends transversely, instead of longitudinally as it does in the fore wing; veins Cu and ist A do not coalesce at the base; the third anal vein is forked as in the fore wing, but as there is no axillary excision the second branch of this vein is not in a detached portion of the wing. (/) THE WINGS OF THE FRENAT^ In the Frenatae the two wings of each side are united by an organ, which is termed the frenulum., or by a substitute for this organ, the greatly expanded humeral angle of the hind wing ; the fore and hind wings differ greatly in venation, due to a reduction of the radius of the hind wings ; the media of both fore and hind wings is reduced to a three- branched condition; and the base of the first anal vein does not coal- esce with the cubitus. The frenulum and the frenulum hook. — The frciiulitni (Fig. ^,^8, f) is a strong spine-like organ or a bunch of bristles borne by the hind wing at the humeral angle, and which projects beneath the fore wing. In the males of certain moths, where the frenulum is highly developed, there is a membranous fold on the fore wings for receiving the end of the frenulum, and thus more securely fastening the two wings together; this is the frenulum-hook (Fig. 338, / h). Except in the Microlepidoptera the frenultun of the male consists of a single strong spine-like organ ; and that of the female, of two or more bristles. Fig. 338.— Wings of Thyridopteryx; f, frenulum; f h, frenulum hook. THE WIXCS OF LEPIDOPTERA 331 The bristles of which the frenulum of the female is composed are spine- like setcB. This is shown by their basal articulation, and by the fact that they are hollow. Frequently when a wing is mounted in Canada balsam, the cavity in each bristle can be easily seen. That the spine-like frenulum of the male consists of united bristles is also shown in the same way, the spine-like organ containing two or more longitudinal cavities. Evidently the frenulum of the male is the more highly specialized form of the organ. This is doubtless correlated with the more active flight of the males, in seeking their mates. This also explains the development of a frenulum hook in the male ; while as a rule this organ has not been attained by the female. In the family Sesiidas, where both sexes fly swiftly, the bristles composing the fren- ulum are consolidated in the female as well as in the male. The female also possess a frenulum hook; but this is not so highly specialized as that of the male. The loss of the frenulum in certain of the Frenatae. — The loss of the frenulum in certain members of the Frenatae has followed the development of a substitute for it. It is obvious that if the two wnngs of each side over- lap to a great extent, their acting together will be assured by this fact. And this is what has taken place with the butterflies, the skippers, and certain moths. With these insects the humeral angle of the hind wing has been greatly enlarged, so that it projects far beneath the fore wing (Fig. 339). When this has taken place there is no longer any need of a frenulum, and consequently this organ is no longer preserved by natural selection. We find, therefore, that several families of the Lepidoptera that belong to the suborder Frenata?, being descendants of ancient frenulum-bearing moths, no longer possess a frenulum. These are classed in the S>mopsis of the Lepidoptera, given in the writer's text book of entomology as the " frenulum -losers. ' ' It is a very interesting fact, and one that bears out the theory just stated, that in the more generalized of the frenulum-losing moths, as the Bombycidse, the frenulum has not yet entirely disappeared but is preser^'-ed in a vestigial condition (Fig. 340). Objection has been urged to the establishment of the suborder Frenatae because many of the forms included in it do not possess a frenuliim. Those 2d /I Fig. 339. — Wings of Anisota I'irginiensis. 332 THE WINGS OF LEPIDOPTERA who make this objection ignore the following taxonomic principle set forth by the writer at the time the establishment of the suborders Jugatse and Frenatae was proposed, viz. : "There will arise, I believe, in a work of this kind a necessity for distinguishing between the essential characters of a group and those characters which are used by the systematist merely to enable students to recognize members of the group. For it seems to me that the essential characters of a group of organisms do not lie necessarily in the presence or absence of any structure or structures, or in the form of any part or parts of the body of the living members of the group ; but rather in the char- acteristic structure of the progeni- tor of the group, and in the direction of specialization of the descendants of this progenitor. "Thus, to use again the illus- tration given above, the Jugatae are essentially characterized as the descendants of those ancient Lepidoptera in which the wings of each side were united by a jugum; and they are also characterized by a tendency towards an equal reduction of the veins of the two pairs of wings. While the Frenatae are essentially characterized as the descendants of those ancient Lepi- doptera in which the wings of each side were united by a frenu- lum; and they are also character- ized by a tendency towards a greater reduction of the veins of the hind wings than of the fore wings, or, in other words, by a tendency towards a cephalization of the powers of flight. The fact that in many of the Frenatas the frenulum has been lost, does not invalidate in the least the truth of this characterization. The loss of the frenulum, however, in certain Frenatas renders necessary the use of some other character or characters by the systematist as recognition characters." The reduction of the radius of the hind wings. — Throughout the sub- order Frenataj the radius of the hind wings is so greatly reduced that it appears to be unbranched. This reduction of the radius is due to two facts ; first, vein Ri coalesces with the subcosta; and second, all of the branches of the radial sector coalesce so as to form a single vein. Fig. 340. — Wings of Bomhyx mori. THE WINGS OF LEPIDOPTERA 333 The coalescence of vein Ri and the subcosta was not suspected until the tracheation of pupal wings was studied; it then became evident. Figure Fig. 341. — Hind wing of a pupa of Picris rapcc. 341 represents the tracheation of a hind wing of a pupa of Pier is rapes. In this wing, trachea Ri extends from the radial fork directly towards the costal margin of the wing until it nearly reaches the subcostal trachea ; it then bends out- ward and extends closely parallel with the subcostal trachea. A single vein is formed about these two trachege; this vein appears to be the subcosta, but the evidence presented by the tracheation of the pupal Vv^ing shows that it is vein Sc+Ri. Figure 342 repre- sents the wings of the adult of the closely allied Pontia protodice, in which it can be seen that a single vein, Sc-|-Ri occupies the posi- tion of the two tracheae just described. In the adult wings of Fig. 342. -Wings of Poiitia protodice. the greater number of the Frenata there is no indication of the coal- escence of veins Sc and Ri of the hind wings; this is the case in Pontia 334 THE WINGS OF LEPIDOPTERA protodice. But in a considerable number of these insects the radial fork is evident; but in these the free part of vein Ri has the appearance of being a cross-vein (Fig. 340). In all pupal wings of the Frenatse the trachea of the radial sector of the hind wings is reduced to an unbranched condition; but in many cases trachea Mi has been transferred to it (Fig. 341) ; this is correlated with the atrophy of the main stem of media in the adult, which is discussed later. The reduction of media to a three-branched condition. — In the Frenatas as in the Jugatas the media of both fore and hind wings has been reduced to a three-branched condition. It is quite possible that this reduction has taken place independently in the two suborders ; for the separation of these suborders doubtless occurred very early in the course of the evolution of the Lepidoptera. For this reason, in attempting to determine the manner in which this reduction has taken place in the Frenatae I submit only data drawn from the study of representatives of this suborder. A study of the method of branching of media in the Frenatae shows that the reduction of this vein is the result of the loss of vein M4 as a distinct vein. There is no reason to believe that this vein has atrophied ; the reduc- tion has doubtless been brought about either by the coalescence of veins M3 and M4 or by the coalescence of veins M4 and Cui. One naturally turns to a study of the tracheation of the wings of pupae to find a solution of the problem ; but in none of the many lepidopterous pupae examined is media more than three-branched ; this suggests the conclusion that veins M3 and M4 have coalesced. But in the course of a renewed investigation of this subject I have found data that lead me to believe that the reduction of media in this suborder is a result of the coalescence of veins M4 and Cui, and that consequently the vein that has been commonly designated as vein Cui is really vein M44-CU1. The nature of this data is well-shown by the tracheation of the wings of Pieris rapes. In the hind wing (Fig. 341) there are, in addition to the tracheae that are the anlagen of the principal veins and their branches, two short trachea; designated in the figure as r-m and m-cu, respectively. Trachea r-m is obviously a vestige of a secondary connection between the medial and radial tracheas. Note that what is here considered as trachea r-m arises from trachea R^ and that trachea Mi arises from trachea r-m a considerable distance from its base ; the letters r-m in the figure are placed at the tip of this trachea. This connection was probably, at first, trans- verse, occupying the position of the cross-vein r-m in more generalized wings, as in the fore wings of Prionoxystus (Fig. 343) and in the hind wing of Packardia (Fig. 348). Correlated with a lessening of the air supply through the main stem of the medial trachea, or with whatever the cause may be that results in the atrophy of the main stem of vein M, the base of THE WINGS OF LEPIDOPTERA 335 trachea Mi migrated alon<( this connection towards trachea R^ and became separated from trachea Mo, receiving its air via trachea R^. At the same time trachea r-m migrated along trachea Rs towards the base of the wing, thus affording a more direct course for the air and becoming longitudinal instead of transverse. This series of events, regarding which there is little room for doubt, as all of the stages in the switching of vein Mi to vein R^ exist in living Lepidoptera, throws light on the question of the significance of trachea m-cu. The position of this trachea, which is closely analagous to that of trachea r-m, indicates that it is a vestige of a secondary connec- tion between the medial and cubital tracheae. A", Fig. 343. — Wings of Prionoxyslus robinicB. It is evident that trachea m-cu is not a vestige of a connection between coalesced tracheae M3 and M4 and trachea Cui, because trachea M3 still retains its primitive connection with the main stem of the medial trachea, which would not be the case if at any time it had become connected directly with trachea Cui. The connection must have been between tracheae M4 and Cui. The two intermediate branches of the medial trachea. Mo and M3 retain their connection with the main stem of this trachea, while on either side a branch has become connected to another trachea, Mi to trachea Rg and M.1 to trachea Cui. In each case the transferred branch has migrated along the secondary connection towards the trachea with which it has become connected. In the hind wing of Pier is rapes the base of trachea Mi has progressed a con- siderable distance towards trachea R^, but is still distinct from it, arising 336 THE WINGS OF LEPIDOPTERA from trachea r-m. In the fore wing (Fig. 333) trachea Mi has reached the radial trachea and coalesces with it for a considerable distance. If this Ki R 2d A Fig. 344. — Tracheation of a fore wing of Anosia. coalescence were to continue to the margin of the wing, vein Mi would be lost as completely as is vein M4. From these facts I conclude that in both fore and hind wings trachea M4 has reached trachea Cui and has coalesced with it completely. In other words the reduction of media is the result of the coalescence of veins M4 and Cur, hence the vein that is commonly designated as Cui in the Lepidoptera is really vein M4+CU1. In the fore wing of Pier is rap(S (Fig. 333), the vestige of the secon- dary connection between the medial and the radial trachea has been lost, although that between the medial and cubital tracheae has been retained. Both of the vestiges are commonly lost ; but one or both of them are retained in many lepidop- terous pupae. That between the medial and cubital tracheae is well- Fig. 345. — Wings of Citheronia regalis. preserved in the pupa of Anosia for example (Fig. 344). Although I firmly believe that the vein in lepidopterous wings that is commonly designated as vein Cui is really vein M4+CU1, for the sake of simplicity, it seems better to designate it ordinarily as vein Cui. This is in THE WINGS OF LEPIDOPTERA 337 accordance with the plan adopted in the discussion of hymenopterous wings, where frequently a compotind vein is designated merely by the term indi- cating its most obvious element, in order to avoid the use of a more oimber- Fig. 346. — Fore wing of Anosia. some terminolog}-. In this chapter, I have used the complete designation M4+CU1 in the few cases where I wish to indicate my conclusion regarding 3dA Fig. 347. — Hind wing of Anosia. the fate of vein M4 in the Lepidoptera; but ordinarily it is designated as vein Cui. The atrophy of the main stem of media. — In a few families, as for example in the Cossidae (Fig. 343) and the Psychidae (Fig. 338) the main 338 THE WINGS OF LEPIDOPTERA Stem of media is well-preserved ; and in some other families as the Megal- opygid^, Eucleidse, and Pyromorphidse, it is preserved in some members of the family and lost in others. The retention of the main stem of media is obvi- ously a character indicating a comparatively generalized condition of those families in which it exists. In very many cases where the main stem of media is lost no trace of it remains; but frequently the position formerly occupied by it is indicated by a faint line or scar; such a line exists in the fore wing of Cither onia regalis (Fig. 345). Less fre- quently vestiges of the basal part of media remain as short _ stimips, projecting into cell '^ '^f^^ R+M, from the outer end ig- 34 • mgs o ac . ^^ ^^.^ ^^^^ _ vestiges of this kind are present in the wings of Anosia (Fig. 346 and 347). The transfer of the branches of media to adjacent veins. — Con-e- lated with the atrophy of the base of media is the coalescence of its branches with the adjacent veins. It follows from this that the extent to which this coalescence has gone is an indication of the degree of departure of a form from the primi- tive type. Compare, for example, the hind wing of Packardia (Fig. 348) with the hind wing of Adoneta (Fig. 349), two genera of the family Eucleidac. In Packardia, where a remnant of the base of media still persists vein Mi is merely connected withthcradial sector by a cross-vein. But in Adoneta, where the base of -' ,j.^^ media of the hind wing is lost Fig. 349. — Wings of Adoneta. THE WINGS OF LEPIDOPTERA 339 vein Ml coalesces with the radial sector for a considerable distance. It is obvious that in this respect, the extent of the coalescence of veins Mi and Rs, Adoneta is the more highly specialized of the two genera. It is an interesting fact that in cases like Anosia (Fig. 346) where the vestiges of the basal part of media are short spurs projecting into the discal' cell, the branches of media are often discontinuous with these spurs. It is obvious that the spurs indicate the positions fonnerly occupied by the branches of vein M and that they have been left stranded upon the discal vein while the functional parts of the branches to which they correspond have moved into new positions. This is especially marked in the case of vein M3 of the fore wing oi Anosia. The union of vein Mi with radius and of vein M3 with cubitus after the atrophy of the base of media is what would be expected. But in which direction would one expect the base of vein M2 to migrate? Occupying an intermediate position between radius and cubitus it may go either way. It is like a stream in the middle of a level plain, a trifle may change its course. And thus we find that in some families it migrates towards cubitus making this vein appar- ently fotir-branched, while in other families it goes towards radius, leaving cubitus appar- ently three-branched. The former condition exists in the Papilionidas (Fig. 350); Pieridas (Fig. 342). This difference may be looked upon as a difference in kind of specializa- tion, and is of high value as indicating a dichotomous division of the line of descent. It is obvious that in a family where vein M2 has migrated far towards cubitus and has thus established its chief source of air supply in that direction, it is not probable that genera will arise in which vein M2 is Fig. 350. — Wings of Papllio polyxenes. the latter, in the 340 THE WINGS OF LEPIDOPTERA more closely vmited to radius than to cubitus. To resume the figure, the plain throtigh which the stream is flowing is an elevated plateau ; a pebble may determine which of two slopes it shall descend ; but when well started down one, it cannot traverse the other. This character, however, must be used with care. In families where the direction of the migration of the base of vein M2 has been established, as in the Saturniidse (Fig. 345), and in the Lasiocampidae (Fig. 351), it is deci- sive. One need not hesitate a moment in determining to which of these two families a genus belongs. But there are other families in which the directions of this migration is not yet fixed; and here the character is of subordinate value. The first anal vein and the anal furrow. — In the Frenatae the first anal vein, when it is present, traverses the cubito-anal sulcus and extends along the bottom of the cubito-anal fold to the margin of the wing. In those Frenatae in which the first anal vein is atrophied a vestige of it remains as a distinct anal fuiTow. An intermediate condition exists in the wings of Bombyx mori (Fig. 340) in which the teraiinal part of the first anal vein is retained ; the anal fuiTow is indicated in the figure by a dotted line. The reduction of the anal area. — Another method of specialization that occurs to a greater or less extent in most of the families of the Frenatae is a reduction in the number of anal veins. In some of the more generalized Frenatae the three anal veins are preserved in both pairs of wings ; but in the far greater number of families a reduction in the number of these veins has occurred in one or both pairs of wings. As a rule the first anal vein is the first to be lost, and this loss takes place by atrophy. Frequently a vestige of this vein persists as a narrow channel. This is the anal furrow of this suborder, and when distinct is represented in the. figures by a dotted line. The second anal vein is the most persistent of the three anal veins. It is frequently retained when no vestige of either of the others remains. The second anal vein frequently appears to be forked at the base. This is due to its coalescence with the third anal vein, or with the first branch of 3d^ 2dA Fig. 351. — Wings of Clisiocampa aniericana. THE WINGS OF LEPIDOPTERA 341 the third anal vein when this vein is two-branched. The distal parts of the two veins coalesce, while the basal parts remain separate; and thus is formed a single vein which is forked at the base. The wings of Caccecia (Fig. 353), illustrate several of the conditions described above. In the fore wing, the first anal vein is represented by the anal furrow, except at the tip where a short vestige of it remains; in the hind wing this vein is well-preserved. In both wings, the second anal vein is forked at the base. In the fore wing the third anal vein is represented only by the hind branch of this fork ; while in the hind wing both branches of the two-branched third anal vein persist; the first branch being joined to the second anal vein and the second branch extending free to the margin of the wing. In a case of this kind, the identity of the first branch of the third anal vein is not obvious ; but studies of the tracheation of the wings of pwpse show clearly that the apparent forking at the base of the second anal vein is due to the coalescence of this vein with the first branch of the third anal vein. This is shown in the figure of the fore wing of a pupa of Pieris (Fig- 333)- While the extent to which the reduction of the anal area has proceeded may merely indicate the degree of divergence from the primitive type, a wing ha\ang two anal veins being more specialized than one having three and less specialized than a wing having a single anal vein, a comparison of the extent of the reduction in the two pairs of wings in different insects may show a difference in kind of specialization indicating a dichotomous division of a line of descent. As an illustration of an application of this principle the separation of the families Papilionidee and Pieridas, which were formerly classed together as a single family, may be cited. In the fore wings of the Papilionidae (Fig. 350) all three of the anal veins are at least partly preserved, while in the hind wings there is only a single anal vein. On the other hand in the Pieridffi (Fig. 342) the anal area of the fore wings is more reduced than the anal area of the hind wings, the former having a single anal vein, the latter two. From this it follows that in tiie Papilionidas the reduction of the anal area of the hind wings preceded the reduction of the anal area of the fore wings; while in the Pieridas the reverse was the case. It is evident there- fore that at the time the separation of these two families occun^ed there had been no reduction of the anal areas of either pair of wings. The difference in the direction of the migi-ation of the base of vein M2 in these two families, already indicated, confirms the conclusion that their separation occurred very early in the history of that division of the Frenatas represented by the butterflies. At the time when it occurred there had been no reduction of the anal areas, and vein Mo had not begun its migra- tion towards either radius or cubitus. This is as generalized a condition of wing structure as exists in any of the living Frenatae. 342 THE WINGS OF LEPIDOPTERA The anastomosis of veins. — In many genera of this order the branches of radius of the fore wings anastomose so as to form one or more closed cells; these have been termed the accessory cells. Sc+R, Fig. 352. — Base of a hind wing of a pupa of Anosia. By an application of the uniform terminology, the homology of the accessory cells can be indicated. Thus in Prionoxystus (Fig. 343), where veins R3 and R4+5 anastomose, cell R3 is divided into two parts, and the accessory cell is evidently ist cell R3. The costa of the hind wings. — Except at the base of the wing in certain members of this order, the costal margin of the hind wings is not thickened appreciably; it may be said, therefore, that the costal vein is either very greatly reduced in length or is lost entirely in the hind wings of the Lepidop- tera. In the hind wings of some pupae, as that of the monarch butterfly, for ex- ample, there is a distinct costal trachea (Fig. 352); but 3<-^^ 2d A ^stA Fig- 353- — Wings of Caccecia cerasivorana. in the hind wing of the adult of this species there is no indication of a costal vein. THE WINGS OF LEPIDOPTERA 343 In sonic members of this order, the extreme base of vein C, that part that I have termed the costal sclerite, is well-developed in the hind wings as well as in the fore wings. This sclerite is represented, but not lettered, in the figure of the wings of Prionoxystus (Fig. 343) ; it is also prominent in the hind wing of Caccecia (Fig. 353). When well-developed in hind wings it serves as a support of the frenulum if a frenuliim is present. The costal sclerite is very prominent in the hind wings of Castnia. The humeral veins, — In some members of the Frenataj in which the frenulum has been supplanted by a broadly expanded humeral angle of the hind wing, this part of the wing is stiffened by one or more accessory veins ; these are termed the humeral veins. Hiuneral veins are present in the Lasiocampidae (Fig. 351); in this family they vary in ntunber. In the butterflies there is commonly a single humeral vein in each hind wing; this vein is represented in Figures 342, 347, and 350. The trachea that preceded it is shown in Figure 352. In butterflies the humeral vein is often forked. It has been suggested that the single humeral vein of the hind wings of butterflies is vein Sci ; but this does not seem to me to be at all probable. The humeral veins of the Lasiocampidae are obviously accessory veins. The fact that there are frequently several of them, the niunber varying in closely allied genera, and their evident function as a support of the second- arily developed expansion of the humeral angle of the wing clearly indicate that they are secondarily developed veins. In the butterflies the conditions are similar to those found in the Lasiocampidae. The frenulum has been supplanted b}' a broadh' expanded humeral angle of the wing, and this area is stiffened by a vein ; the striking difference is that in the butterflies there is a single humeral vein while in the lasiocampids the number of these veins is variable. In no member of the Frenatffi known to me is the primitive forking of the subcosta preserved in the fore wings; even in the most generalized moths, as the Cossida:, the subcosta of the fore wings is simple. It does not seem at all probable that this primitive feature, which is lost in the fore wings, should be preserved in the hind wings, in which there is a marked reduction of the subcosto-radial areas. It is more probable that correlated with the expanding of the humeral angle of the hind wing an accessory vein has been developed to stift'en this part of the wing. The splitting of the radial sector in butterflies. — A remarkable speciali- zation of the radial sector of the fore wings that appears to the distinctively characteristic of the Rhopalocera was pointed out by Headlee ('07). This specialization consists of a splitting of the radial sector which results in vein R4+5 arising from the main stem of the radius near the base of the wing. Owing to the total or nearly complete atrophy of the base of vein R4+5 in wings of adults, this splitting back of it is not obvious if only mature 344 THE WINGS OF LEPIDOPTERA wings be examined. But it is clearly indicated by the tracheation of the wings of pupas ; and after one has learned what has taken place by a study of pupal wings, it is easy to find vestiges of the lost part of vein R4+0 in the wings of adults. The wings of Anosia plexippus illustrate well these facts. Figure 354 is a reproduction of a photograph of a wing of a pupa of this species, and Fig. 354.— Fore \vu j[ .III Odd plexippus. Figure 355 illustrates the tracheation of this wing. In this wing, trachea R4+5 separates from the radial trachea near the base of the wing. Figure 346 represents the adult wing of the same species. In this wing a vestige of vein R4+5 is to be seen near the end of the discal cell. Ri R 2d A Fig- 355- — Tracheation of the wing represented in Figure 354. An extended examination of the wings of Lepidoptera made by Headlee led to the conclusion that this splitting back of vein R4+5 has taken place throughout the Rhopalocera, and has not occurred in the Heterocera. THE WINGS OF LEPIDOPTERA 345 (g) COMPARISON OF TERiMIN'OLOGIES OF THE WING-VEINS OF THE LEPIDOPTERA I. PRINCIPAL VEINS British German Redtenbad ler Spider Conistock & Needham Costa, I Costa, C Costal Costale Subcosla, II I Subcosta, So Subcostal Subcostale Radius, III II Radius, R Media, V III Media, M Median Mediana Cubitus, VII IV Cubitus, Cu VIII : Y ist Anal, I St A Internal Submediana IX a 2d Anal, 2d A Interandsader XI 0 3d Anal, 3d A The system chiefly used by British naturalists, excepting those who have adopted the Comstock-Needham system, is based on that used by Herrich-Schaffer in his "Systematische Bearbeitung der Schmetterlinge von Europa" (1843-1856). The German system given here is that adopted by Staudinger and Schatz in their "Exotische Schmetterlinge" (1892). It should be noted that the costal and subcostal veins of the writers on the Lepidoptera who have not adopted the uniform terminology correspond to the subcosta and radius, respectively, of that system. VEINS THAT EXTEND TO THE MARGIN OF THE WINGS British German Redtenbacher Spider Comstock & h Jeedhart I Costa C 12 C II I Subcosta Sc II SC: IIIi III Radius-one Ri 10 SC, III2 II2 Radius-two Ro 9 SC3 III3 II3 Radius-three R3 8 SC4 III4 III Radius-four R4 7 SC5 III5 II5 Radius-five R5 6 OR V, nil Media-one Ml 5 UR v.. III2 Media-two Mo 4 M3 VI 1 1 III3 Media-three Ms 3 Mo VII, IVi Cubitus-one Cui 2 M, VII3 IVo Cubitus-two Cuo IC VIII V ist Anal ist A lb SM IX a 2d Anal 2dA la lA XI & 3d Anal 3d A 346 THE WINGS OF LEPIDOPTERA The significance of the abbreviations given in the column indicating the German system are as follows: C, Costale; SC, subcostale; OR, Oberer Radiale; UR, Unter Radialc; M, Mediana; SM, Submediana; lA, Innerandsader. In the hind wings vein SC+Ri of the uniform terminology is vein 8 of the British system and vein C of the German; vein Rg is vein 7 of the British system and vein SC of the German. The other veins are designated as indicated in the above table, which refers to the fore wings. As is shown on an earlier page, vein Cui of the Comstock-Needham system is in reality vein M4+CU1. The accompanying figures, copied from Sharp ('99), will serve toillus- trate the British and the German systems of terminology of the wing-veins. Fig- 356, I, is after Hampson, and illustrates the British system. Fig. 356, II, is after Staudingcr and Schatz, and illustrates the German system. M' m' Fig. 356. — Figures illustrating the British (I) and the German (II) systems of terminology of the wing-veins of the Lepidoptera. CHAPTER XXIV THE WINGS OF THE DIPTERA (a) THE MORE GENERAL FEATURES OF THE WIXGS OF THE DIPTERA The most distinctive feature of the wings of the Diptera is the fact that only the first pair are developed as organs of flight ; the second pair being greatly reduced in size. The second pair of wings are known as the halteres, they are thread-like, enlarged at the end, and are pnjbably organs Fig. 357. — Wing of Rhyphus. of special sense. They are present in nearh' all members of the order, even when the front wings are wanting. The fore wings are thin, membranous, and usually either naked or clothed with microscopic setae; but with mosquitoes the wings bear a fringe of scale-like seta on the margin and usually also on each of the wing- veins, and in the moth-hke flies (Psychodidae) and in some others the clothing of setffi is very conspicuous. Fig. 358. — Wing of Rhynchocephalus. In the more generalized members of the order, the wing-venation is simple, the only departures from the hypothetical primitive type being the result of a reduction in the number of veins by the coalescence of adjacent veins; this is well-shown by the wings of Rhyphus (Fig. 357). Neither (:M7) 348 THE WINGS OF DIPT ERA accessory nor intercalary veins are ever developed, and only the principal cross-veins are present. In more specialized forms the typical atTangement of the veins has been greatly modified by the approaching and coalescing of the tips of adjacent veins. Remarkable instances of this occur in the Nemistrinidse (Fig. 358). (6) THE METHODS OF SPECIALIZATION OF THE WINGS OF THE DIPTERA In the discussion of the methods of specialization of the wings of the Diptera reference will be made only to the specialization of the fore wings, as a discussion of the structure of the halters does not fall within the scope of this work. Ordinal specializations. — In all Diptera the venation of the wings is more or less reduced, in none has there been a specialization b}^ addition. /?. R, ?d A I si A C»2 ^'^i Fig. 359. — Wings of a caddice-fly. Even in the most generalized forms, with the possible exception of Proto- plasa, media is only three-branched ; and the anal area is always more or less reduced, that is, in none are there three well-developed anal veins. The loss ojvein Mi. — A study of the method of branching of media in the more generalized forms, as, for example, in Rhyphus (Fig. 357), shows that veins Mi and M2 are preserved distinct, and that the third branch of this vein is cither vein M3, in which case veins M4 and Cui have coalesced; or vein M3+4, veins M3 and M4 having coalesced. THE WINGS OF DIPT ERA 349 The truth of the conclusion that in Rhy pints veins Mi and Mo are dis- tinct is made evident by a comparison of the wing of this insect with the fore wing of a caddice fly (Fig. 3 59), in which media is not reduced. In this comparison, the position of the medial cross-vein should be noted. It will be seen that this cross-\-ein divides cell M2; and consequently the two branches of media in front of the cell divided by this cross-vein are veins Ml and Mo. Owing to the fact that in the Diptera a great reduction of wing tracheae has taken place, it is not possible to determine by a study of the tracheation of the wings of pupae whether vein M4 has coalesced with vein M3 or with S-, 2d A Cu, Qiv^' Fig. 360. — Wing of Proloplasa fitchii. vein Cui; our conclusions, therefore, must be based on a study of wings of adults. Among the dipterous wings that are ccmparati\-ely little modified are those of Rhyphus. An examination of a wing of this insect (Fig. 357) would lead one to believe that the three-branched condition of the media is due to a coalescence of veins M3 and Mi and that consequently the vein labeled M3 is really M3H-4. That the reduction of media may have been attained in another way is indicated by a study of the wings of Proloplasa fitchii (Fig. 360). This species is one of the most generalized members of the order Diptera; this is shown by the fact that the subcosta is two-branched and that all of the branches of the radius are retained distinct. The suggestive feature indicating that vein M4 may have coalesced with vein Cui is the presence of a vein extending from near the middle of the length of vein M3 to vein Cui. This vein has the appearance of a cross-vein; but it is quite possible that it is a section of vein M4, the more distal part of this vein having coalesced with vein Cui. Such a coalescence wovild be quite in accord with what commonly takes place in this order. There are many instances where the distal part of a vein has coalesced with an adjacent vein, the coalescence beginning at the margin of the wing and proceding towards the base of the wnng; with the result that the free part of the vein is pulled out of its earlier position and comes to occupy one transverse to the length of the wing; illustrations of this are given later. If this is what has taken 350 THE WINGS OF DIPT ERA place here, the third branch of media is vein M3 and the vein behind it is vein CU1+M4. This, as shown in the preceding chapter, is probably the method of reduction of media that has taken place in the suborder Frenatae of the Lepidoptera; bvit there the coalescence of the two \'eins preceded outward. The figure of a wing of Protoplasa (Fig. 360) is labeled in accordance with this theory. But until more evidence as to the fate of vein M4 is obtained it seems best to omit ^R.+i any reference to this vein in the designation of the veins of the Diptera. Specializations within the order. — While a partial reduc- tion of the anal area and the loss of vein M4 as a distinct vein are characteristic of recent Diptera as a whole, the greater number of specializations ex- hibited by the wings of Dip- tera have arisen within the order and have been earned to widely different degrees in different members of the order. The most practicable way of discussing these specializations is to treat of the modifications of each of the principal veins separately. ^"' "' The costa. — The costa is Fig. 361.— Wings of fungus gnats. marginal. I have no data to give regarding modifications of its extent. The subcosta. — The primitive, two-branched condition of the subcosta is preserved in a very few forms; one of these is Protoplasa (Fig. 360). In most members of the order it is not branched. In some cases the tip of the simple subcosta ends in the margin of the wing, indicating that vein Sc-z is lost; In other cases the tip coalesces with vein Ri, indicating that vein Sci is lost. Frequently the subcosta is greatly shortened. Three widely different conditions of the stibcosta exist in the three wings of fungus-gnats represented in Figure 361; in the first it is greatly shortened, in the second it is tw^o-branched , and in the third it coalesces with radius for the greater part of its length. The radhis. — In a few genera of flies the radius retains its primitive, five-branched condition; the genus Protoplasa of the Tipulida? has already THE WINGS OF DIPT ERA 351 been cited as an illustration of this; and in the Psychodidas (Fig. 362) all of the branches of radius remain distinct. But usually the number of the branches of this vein is reduced by a coalescence of some of the branches of the radial sector. Thus in many families the radial sector is three-branched, in others it is only two-branched Fig. 362. — Wing of a moth-like fly. and in the gall gnats (Cecidomyiidas) it is reduced to a simple, unbranched condition (Fig. 363). As this variation in the number of the branches of this vein is due to a greater or less degree of coalescence among them, it is evident that here is a character of considerable taxonomic importance ser\'ing as it does to indicate degrees of divergence from the primitive type. Not only do we find differences in degree of reduction of this vein, but differences in the method of reduction are also shown. If the wings of Fig. 363. — Wing of a gall-gnat. Leptis (Fig. 364) and of Dixa (Fig. 365) be compared it will be seen that although in each the radial sector is only three-branched, the reduction has been brought about in a different way in the two genera. In Leptis veins R2 and R3 have coalesced; while in Dixa it is veins R4 and R5 that have grown together. This is a difference in kind of specialization, which 352 THE WINGS OF DIPTERA indicates that the two fonns belong to different Hnes of descent. The common progenitor of these two genera had a four-branched radial sector ; in some of the descendants of this primitive forai one method of reduction has taken place, while in other descendants another method has been followed. That this differentiation took place comparatively early in the history of the order is shown by the fact that in all Nematocera that have a three- Fig. 364. — Wing of Leptis. branched radial sector veins Ro and R3 remain distinct; while in those Brachycera that have a three-branched radial sector veins R4 and R5 are separate. It is evident that as a rule an}- reduction in the number of the branches of radius is the result of a coalescence that has preceded outward, the Fig. 365. — Wing of Dixa. method of reduction being the same as that which commonly takes place in the Lepidoptera. But this is not always the case. In some of the fungus- gnats, for example, there is shown the result of a coalescence that began at the margin of the wing and preceded inward. This is illustrated by the wings represented in Figure 361. In the wing shown at a in this figure, vein R2+3 occupies the tisvial position of this vein ; but in the wing shown at b it coalesces with vein Ri, and it is evident that the coalescence begati THE WINGS OF DIPT ERA 353 at the margin of the wing. The result of this coalescence is the reduction of cell Ri to a small quadrilateral area; in the wing shown at c, the coal- escence of veins Ri and R2+3 is complete and cell Ri is obliterated. 2d A*Cu, Fig. 366. — Wing of Thereva. The media. — A comparatively generalized condition of media exists in the wings of Rhyphiis (Fig. 357), which differs from the hypothetical primi- tive type only in the fact that vein M4 has coalesced either with vein Ma or with vein Cui. The three-branched condition of media is preserved in many families of this order and is illustrated by several of the figures given later. In the reduction of media in this order each of the different methods known has taken place. In Dixa (Fig. 365) and in the fungus gnats represented in Figure 361, veins Mi and Mo have coalesced. It is probable that this coalescence proceded outward. In Rhynchocephalus (Fig. 358), veins Mi and Mo coalesce at the margin of the wing and the coalescence is proceding inward. In the gall-gnat represented by Figure 363, only a slight vestige of media remains; here the last stage of the reduction was doubtless due to atrophy. And in the insects discussed in the following R h ' R. Fig. 367. — Wing of Eulonchus. section of this chapter media is reduced by coalescence with an adjacent vein. The coalescence oj veins A/3 and Cu\- — One of the most striking of the modifications of the wings of Diptera is the coalescence in certain forms of 354 THE WINGS OF DIPTERA vein M3 and Cui. The result of this coalescence where it is carried to its extreme limit is the production of a type of wing in which the homologies of the veins concerned is so obscured that they were not understood until the method of study used here was employed. But by beginning with a Fig. 368. — Wing of Pantarbes. 5,, R, /?..3 2d A+Cu, Fig. 369. — Wing of Conops /?.., idA^Cu, Fig. 370. — Wing of Scenopintis. generahzed form in which the homologies of the wing-veins are easily understood and by selecting a series of forms illustrating varying degrees of modification of this form an understanding of the most specialized form is attained. THE WINGS OF DIPTERA 355 In Rhyphiis (Fig. 357) veins M3 and Cui retain their primitive position, extending nearly parallel and ending remote from each other at the margin of the wing. (For reasons given in an earlier paragraph no account is taken of vein M4 in this discussion). In Thereva (Fig. 366) an approxima- tion of the ends of these veins has taken place, which results in a narrowing of the outer end of cell M3. In Eulonchus (Fig. 367) the tips of the two veins coalesce, and cell M3 is thus closed. In Pantarbes (Fig. 368) the twj veins coalesce for the greater part of their length, and cell M3 is completely- obliterated. The coalescence of veins Cuo and the second anal vein. — Quite closely- correlated with the coalescence of veins M3 and Cui there exists a coales- cence of vein Cuo and the second anal vein. In Rhyphus (Fig. 357) these Fig. 371. — Wing of Rhamphomyia. two veins retain their primitive position. In Leptis (Fig. 364) the tips are approximate. In Thereva (Fig. 366) the tips coalesce for a short distatic3. In Conops (Fig. 369) the coalescence is more striking. In Scenopiniis (Fig. 3 70) it is carried still farther. While in Rhamphomyia (Fig. 3 7 1) it ha? pro- ceeded so far that vein Cu2 extends towards the base of the win; and presents the appearance of a cross-vein. The reduction of the number of cells in a wing. — -A reduction of the num- ber of cells in the wings has taken place in the wings of nearly all Diptera. This is the result of two causes : first, the coalescence of veins; and second, the atrophy of veins. The following will serve as illustrations of this process. In the Brachycera, the coalescence of veins R2 and R3 results in the obliteration of cell R2; and in the Nematocera, a similar coalescence of veins R4 and R5 results in the loss of cell Ri. The atrophy of a vein separating two cells results in the uniting of these cells and a consequent reduction in the number of distinct cells. In the wing of Musca domestica (Fig. 372) cells M and ist M> are separated by the free part of vein M3. In the wing of Psilopodius sipho (Fig. 373), there remains only a short spur representing the free part of vein Mi; and in the wing of Dolichopus coquilletti (Fig. 374), the free part of vein M3 is completely 366 THE WINGS OF DIPTERA Sc ^. Cu2+2d A Fig. 372. — A wing of Musca domes tica. Fig. 373. — A wing of Psilopodins sipho. I'^ig- 374- — -A wing of Dolichopus coquilletti. THE WINGS OF DIPTERA 357 atrophied and cells M and ist Mo arc united into a single cell, cell M + ist M2. The anal veins. — In all Diptera known to me there is a greater or less reduction of the anal area; for in none that I have seen are there present three well developed anal \'eins that extend to the margin of the wing. Fig. 375. — Wing of Erax. In most Diptera the first anal vein is wanting as a distinct vein but in many there is a suture-like line, the anal furrow, immediately back of cubitus and closely parallel with this vein ; this is a vestige of the first anal Fig. 376. — A wing Eristalis. vein ; this furrow is represented in several of the figures in this chapter by a dotted line. The first anal vein is retained, however, in certain Asilidae; where, although somewhat shortened, it is a distinct vein extending from the base of the wing to near the point where vein Cu forks (Fig. 375). id A id A Fig. 377. — Wing of Tipnia. 358 THE WINGS OF DIPT ERA The second anal vein is the most persistent of the three anal veins ; it is well-preserved in many families; and is represented in several of the figiires in this chapter. The third anal vein is well-preserved in comparatively few forms, although a vestige of it exists in many. It is well-preserved in Tipida (Fig. 377), and is fairly well-preserved in Stratiomyia (Fig. 378). The arctilus. — There is a well-developed arculus in the Diptera; but owing to the reduced condition of the tracheation in this order the elements that enter into its formation can not be definitely determined. In labeling the cells at the base of the wing, I have assumed that the structure of the Sc /?, R.., R, Fig. 378.- — Wing of Stratiomyia. arculus is the same as in those orders where its composition is well-known. According to this interpretation veins R and M coalesce from the base of the wing to the arculus; and, consequently, the cell behind the coalesced veins R and M is I St cell M (Fig. 378). In many of the Diptera, the base of the free part of the media has migrated along the arculus so far that it appears to arise from the cubitus (Fig. 375)- The spurious vein. — In the Syrphidse there is a longitudinal thickening of the wing between the radius and the media ; this is termed the spurious vein (Fig. 376, s). (c) COMPARISON OF TERMINOLOGIES OF THE WING-VEINS OF THE DIPTERA. It is practically impossible to make in a single table a comparison of the various terminologies of the wing-veins of the Diptera that have been used by the prominent writers on this order. This is due to the fact that none of the older authors succeeded in determining the homologies of the different veins throughout the order; the result is that homologous veins are designated differently in the discussions of different families. This was frankly admitted by Osten-Sacken ('69, p. 32) in the following remarkable statement. THE WINGS OF DIPT ERA 359 "This study of the homologies has two distinct aims in view: the scientific aim of showing that the ground-plan of the venation is the same in all of the families of the order; and the practical aim of adopting a terminology for descriptive purposes. We cannot carry out a terminology on solely theoretical grounds; we will have to vary the details of it accord- ing to the peculiarities of structure occurring in different families, the main plan remaining the same. This is done in all the departments of zoology, and I do not see why the venation of the Diptera should be treated differ- ently." Owing to the lack of uniformity in the terminologies used by different authors and to the fact that even in the writings of a single author homol- ogous veins are designated differently in the discussions of different families, 5-.. . TT— ^ '^ Fig. 379. — Wing of Tabanus (After Williston). it is necessary for the student using any of the works in which the uniform terminology has not been adopted to make a study of the system used in that particular work. The system that has has been most used in this country is that which was adopted by Loew ('62) and modified by Osten-Sacken ('69). Figure 379 from Williston ('08) illustrates this system as applied by him to the wing of Tabamts. In the following table this system and that of Redten- bacher are compared with the uniform terminology. Redtenbacher in his efforts to recognize the concave veins IV and VI of his system was led into inconsistences in his determinations of the homolo- gies of the branches of media. The equivalents given in the following table are based on his figure of the wing of Stratiomys. 360 THE WINGS OF DIPT ERA WILLISTON REDTENBACHER COMSTOCK AND NEEDHAM {Tabanus) (Strationiys) {All families) Costal I Costa C Auxiliary II Subcosta Sc First longitudinal III. Radius-one Ri Second longitudinal I III2 III3 Radius-two Radius-three R3 Third longitudinal < III4 III5 Radius-four Radius-five R4 R5 Fourth longitudinal I IV V Media-one Media-two Mt M2 1 VI Media-three M3 Fifth longitudinal \ VIIi Cubitus-one Cui \ VII2 Cubitus-two CU2 VIII First Anal ist A Sixth longitudinal IX Second Anal 2dA Third Anal 3dA Humeral cross- vein a Humeral cross-vein h Anterior cross-vein b Radio-medial cross-vein r-m Discal cross- vein c Medio-cubital cross-vein m-cu Posterior cross-vein d Medial cross- vein nt (d) COMPARISON OF TERMINOLOGIES OF THE CELLS OF THE WINGS OF THE DIPTERA The writers on the Diptera have made use of very complete terminol- ogies of the cells of the wing. As examples of these I give in the following table those used in three important American works, and indicate the equivalents of the terms used in the uniform terminology. THE WINGS OF DIPTERA 361 LOEW ('62) OSTEN-SACKEN ('69) WILLISTON (08) COMSTOCK AND (Ortalis) (Cladiira) (Tabanus) NEEDHAM I st costal ISt C 2d costal Costal Costal 2dC 3d costal Subcostal Subcostal Sc I St basal I st basal ist basal R Marginal Marginal Marginal R. ist submarginal R2 Submarginal 2d submarginal ist submarginal R.3 2d submarginal R4 I St posterior ist posterior ist posterior Rs 2d basal 2d basal 2d basal M 2d posterior 2d posterior Ml Discal Discal Discal IStM2 2d posterior 3d posterior 3d posterior 2dM2 4th posterior 4th posterior M3 Cu 3d posterior 5th posterior 5th posterior Cui Anal Anal Anal ISt A Axillary Axillary Axillary 2d A Spurious 3dA The difference in the application of the terms first submarginal and second submarginal in the terminologies of Osten-Sacken and Williston arises from the fact that in each case the cell lying next to the marginal cell is termed the first submarginal; but this is not the same cell in the two genera used to illustrate their terminologies. The genus used by Osten- Sacken is a Tipulid; and in all Nematocera that have a three-branched radial sector veins R2 and R3 remain distinct; hence the cell next to the marginal cell is cell Ro. On the other hand the genus used by Williston belongs to the Brachycera ; and in this division of the order veins R2 and R3 coalesce throughout their entire length, and consequently the cell immed- iately behind the marginal cell is cell R3. Fig. 380. — Sphecius speciosus. CHAPTER XXV THE WINGS OF THE HYMENOPTERA (a) THE GENERAL FEATURES OF THE WINGS OF THE HYMENOPTERA The members of this order have four wings (Fig. 380) ; these are mem- branous, and have the wing-venation more or less reduced. In the more generaHzed famihes the reduction of the wing-venation is shght; in the more speciaHzed famihes, it is extreme. The two pairs of wings are similar in texture. The wings of each side are held together by a row of hooks, the hamuli, on the front margin of the hind wing (Fig. 381, h); these hooks fasten to a fold in the hind margin of the front wing, so that the two wings present a continuous surface. The hind wings are smaller than the fore wings, and have a more reduced venation. Some forms are apterous. (6) THE VENATION OF THE WINGS OF THE MORE GENERALIZED HYMENOPTERA The suborder Chalastogastra has been recognized as the more general- ized of the two suborders of this order independently of any consideration of the characters presented by the wings. This conclusion has been con- firmed by studies of the wings; for within this suborder are to be found the most generalized wings that exist among living representatives of the Hymenoptera. In the wings of certain sawflies of the families Xyelidae and Lydidas we find a close approximation in the number of the wing-veins to that of the hypothetical primitive type of wing- venation. But even here the courses of the branches of the forked veins have been greatly modified. These changes have been so great that the determination of the homologies of the wing-veins in this order was one of the most difficult problems of this kind that arose in the course of the study of the wings of insects. (362) Fig. 381. — Wings of Apis showing hamuli. y. THE WINGS OF HYMENOPTERA 363 This determination was made by the writer, from an examination of the wings of adult insects, and pubHshed in a text-book* before the ontogenetic study of wings was undertaken by Dr. Needham and myself. The results of this later study did not modify the conclusions that I had reached from the study of adults alone. In fact, we found, as will be shown later, that the courses of the tracheae of the wings of hymenopterous pupae have not been modified in the same way as have the courses of the veins; and that, for this reason we are still forced to determine the homol- ogies of the wing-veins in this order by a comparative study of the wings of adult insects. An understanding of the hymenopterous type of wing-venation was not derived from a study of the wings of Hymenoptera alone. Even in the most generalized of the living members of this order, the wings are so highly specialized that the homologies of certain veins would never have been suspected but for help from another source. Fortunately the most characteristic method of modification of the courses of the wing-veins in the H3^menoptera is illustrated by the wings of certain Diptera also. And in the Diptera examples of ever}^ stage in the modification of the courses of veins by this method can be seen ; while in the Hymenoptera only the later stages are shown. Reference is made here to the coalescence of veins from the margin of the wing towards the base of the wing. This results frequently in a branch of a longitudinal vein becoming transverse, so that it appears like a cross-vein; and in some cases, where the coalescence has been carried still farther, a branch of a longitudinal vein has been so diverted from its primitive course that it extends towards the base of the wing. Among living Diptera there are preserv^ed examples of all of the stages of this modification of the course of a vein. The series of figures illustrating the coalescence of veins Cu2 and 2d A (p. 355) shows this. With this data before tis let us proceed to an examination of h^onenop- terous wings. The most generalized condition of the venation of the wings that is known in the Hymenoptera occurs in the two genera Pamphilius, of the family Lydidoe, and Macroxyela, of the family Xyelidae. There is preserved in the fore wings of each of these genera of sawflies all of the primitive wing- veins with a single exception ; and as it is not the same vein that has been lost in the two genera, a figure of a typical hymenopterous wing can be made from a study of the two. Figures 382 and 383 represent such a wing; in the former the veins are'lettered; in the latter, the cells. See also Plate X, where the courses of the veins are made more evident by the use of alternating colors. The typical h\mienopterous wing represented in the accompanying figures is a figure of the fore wing of Pamphilius (Fig. 391) except that vein *Comstock: A Majiual for the Study of Insects. (1895), pp. 603-607. 364 THE WINGS OF HYMENOPTERA Ro, which is lacking in this genus, is added. This vein is well-preserved in Macroxyela (Fig. 392) ; but in Macroxyela vein Cuo is lost; the position of the lost forking of the cubitus is indicated, however, by a bend in this vein. In the wings of these sawfiies, the anal furrow and the median furrow are both well-marked, and are in the typical positions; that is, the anal Fig. 382. — The veins of a typical hymenopterous wing (From C. & N.). furrow is immediately in front of the first anal vein and the median farrow in front of the media. The furrows are represented by dotted lines in Figures 382 and 383. In the anal area the three typical veins are preserved ; but they coalesce to a considerable extent, both at the base and near the margin of the wing. In the basal part of the prenal area, the stems of the principal veins are as follows: the costa coincides with the costal margin of the wing and extends from the base of the wing to the stigma (Fig. 382,0);* the subcosta Fig. 383. — The cells of a typical hymenopterous wing (From C. & N.). (Fig. 382, Sc) is well-preserved and is forked; back of the subcosta is a strong stem formed by the coalescence of the other three veins of the preanal *It is impossible to state whether the costa extends beyond the beginning of the stigma or not. Both MacGillivray and Bradley believe that it ends at or shortly before the beginning of the stigma. This conclusion is based on the fact that, while costa in all insect wings, when present, lines the margin of the wing, in the Hymenoptera there is frequently a strip of free membrane of varying width extending from the nodal furrow to the apex of the wing. This strip is quite wide in the wings of the female of Tiphia, for example; and widens into the appendiculate cell in some wings. THE WIISCS OF HYMENOPTERA 365 area ; the cubitus (Fig. 382, Cu) soon separates from this stem, extending in a curve towards the anal furrow; while the radius and the media coalesce for about half their length. In order to make these veins more distinct in the figure the free portion of the media is marked with cross lines ; see also Plate X. When we pass from a consideration of the main stems of the veins to a study of the branches, we meet a much more complicated problem, a problem that could not have been solved by a study of Hymenoptera alone. But a knowledge of the methods of specialization of the wings of Diptera gives, as already stated, a key to an understanding of the wings of Hymen- optera. If the reader will examine the series of figures illustrating the coalescence of veins Cu^ and 2d A in the Diptera (page 3 5 5) , he will find it easy to under- stand what has taken place in the Hymenoptera. In the Hymenoptera, however, both branches of the cubitus coalesce with the first anal vein ; and li^ A Cu. Cu.' Fig. 384. — Wing of Pantarbes. this coalescence has proceeded so far that both branches cross the anal furrow, and end in the anal vein remote from the margin of the wing. It should be noted that vein Cu2 is comparatively rarely preserved in this order. Dr. MacGillivray, in his very extended study of the wings of the Tenthredinoidea, (MacGillivray '06, p. 552), notes its presence in represen- tatives of several genera of the Lydidje ; but in most of the Tenthredinoidea all traces of it are lacking. If the branches of the media be now examined, it will be seen that vein Ml (Fig. 382) extends longitudinally near the center of the distal part of the wing, its primitive course being modified slightly if at all. Vein Mo follows a course similar to the course of this vein in the dipterous gentis Pantarbes (Fig. 384). Vein M3 extends across the anal furrow near the margin of the wing and coalesces with the first anal vein. It is evident that the forces that are causing the branches of the cubitus to migrate along the first anal vein towards the base of the wing are exerting a similar influence on this vein. It is also evident that veins M4 and Cui coalesce at the tip, and that the migration of the united tips of these veins (marked Cui in the figure) 366 THE WINGS OF HYMENOPTERA towards the base of the wing has so modified the course of that part of vein M4 which is still free that it extends towards the base of the wing. This change is very similar to the change in the course of vein Cu2 in the dipter- ous genus Rhamphomyia (Fig. 385). A curious result of this change in the direction of the course of vein M4 is that the cell M4 has been closed and pressed back to the center of the wing ^4 + J Fig. 385. — Wing of Rhamphomyia. (Fig. 382, M4), and now lies in front of the free portion of vein M4 instead of behind it. A somewhat similar modification of cell Ms is found in the dipterous genus Eulonchus (Fig. 386). Let us now consider the courses of the branches of the radius. Here again we can gain help from a study of dipterous wings. Observe in Pantarbes (Fig. 384) the coalescence of the tips of veins R5 and Mi. In the Hymenoptera a similar coalescence of veins R5 and Mi has occurred; but it has proceeded much farther, so that the free portion of vein R5 in Pampkilius Fig. 386. — Wing of Eulonchus. (Fig. 391, Rr.) is remote from the end of the wing and has the appearance of a cross-vein. In the Hymenoptera vein Rr, has been followed in its migration along vein Ml by vein R4, which has now reached a stage in Pamphilius that is quite similar to that reached by vein R.^ in Pantarbes; but like vein R5 it has the appearance of a cross-vein. In the fore wing of the honey-bee (Fig. THE WINGS OF HYMEN OPT ERA 367 387) veins R4 and R5 still retain the appearance of branches of a forked vein. In Paniphilius vein Ri is cur\'ed away from the costal margin of the wing to make room for a stigma (Fig. 391 and Fig. 382, st), and vein R3 ends in Fig. 387. — Wings of a honey-bee (From C. & N.). the costal margin a short distance before the apex of the wing. Vein R2 has been lost in this genus but as it is well-preserved in the closely allied genus Macroxyela (Fig. 392) it is represented in the figure of the typical h^Tiienopterous wing (Fig. 3S2). While the tips of the branches of the radial sector have migrated away from the apex of the wing, the bases of these branches coalesce in the opposite direction; from these two causes results the transverse bracing of the radial area of the wing, which is a very characteristic feature of the venation of the wings of this order. The details of these changes will be made clear by an examination of Figures 388a and 388b. The former represents the primitive Fig. 388. — Diagrams of the radius: a, typical; b, hymenopterous (After C. & N.). mode of branching of the radius; the latter, the radial area of the typical hymenopterous wing. As the radial cross-vein is usually present in the H>Tnenoptera, it is represented in these figures. In the hvTuenopterous type, veins R2+3 and R4+5 of the primitive t>T3e coalesce so far that the 308 THE WINGS OF HYMENOPTERA branches of the radial sector arise from a common stem ; and the tips of all of them have moved away from the apex of the wing, veins R2 and R3 following the costal margin of the wing, and veins R4 and R5 following vein Ml. Reference has been made above to the curving of vein Ri away from the margin of the wing to make room for a stigma (Fig. 383, si). The stigma is bounded in front and at its base by the tip of vein Sc2, which anastomoses with the radius for a considerable distance but separates from that vein at the base of the stigma. In the above account reference has been made only to the venation of the fore wings. In the hind wings of most Hymenoptera the venation has been so greatly reduced that it is exceedingly difficult to determine the homologies of the wing-veins. The suborder Chalastogastra includes those Hymenoptera in which the reduction of the venation of the hind wings is least. The extended study of the wings of this suborder made by MacGillivray ('06) enabled him to prepare a diagram of a typical hymenopterous hind wing, which is of great value as an aid to the determination of the venation of the hind wings of Hymenoptera, this diagram was followed in the preparation of Plate X. The lacking veins are indicated by dotted lines. From the foregoing account it will be seen that even in the most general- ized of living Hymenoptera there exists a highly modified wing-venation. But notwithstanding this it is possible to determine the homologies of all of the wing-veins. Some of the modifications of this primitive hymenopterous type will be discussed in a later section of this chapter. [c) THE TRACHEATION OF THE WINGS OF THE HYMENOPTERA The tracheation of the wings of the Hymenoptera was studied by Com- stock and Needham in order to ascertain if the courses of the tracheas fur- nish any data regarding the homologies of the wing-veins in this order. We had found, as shown in preceding pages, that in the more generalized insects there is a close correlation between the venation and the tracheation of the wings; and that it can be accepted as a firmly estabhshed fact that the courses of the wing-veins of primitive insects were determined by the courses of preexisting tracheae. But we also found that in the Trichoptera there is little correlation between the venation and the tracheation of the wings, a remarkable reduc- tion of the wing-trachese having taken place. A similar reduction of the tracheae of the wings exists in most families of the Di]^tera; and even when a large portion of the tracheae are retained, as in certain Asilids, they afford little aid in determining the homologies of the wing veins. THE WINGS OF HYMENOPTERA 369 Again, in the Hymcnoptera, we found that the courses of the tracheae cannot be depended upon for determining the homologies of the wing-veins, notwithstanding the fact that in the more generahzed members of the order a very complete system of wing-tracheae exist. The wings of a pupa of Fig. 389. — Wings of a pupa of Tremex (From C. & X.). Tremex (Fig. 389) illustrate the extent to which the tracheae are retained in the more generalized members of the order; and the wings of Apis (Fig. 390) illustrate the tracheation of the wings in a more specialized form. • The most striking feature of the courses of the main tracheae in the wings of Tremex is that each extends in a nearly direct line from the base of the wing to near its outer margin ; and that while in most cases the basal part of each lies in the cavity of the vein with which it corresponds, this corres- pondence does not extend to the branches of the principal veins. Fig. 390. — Wings of a young pupa of Apis (From C. & N.). Even in the case of the principal veins, this correspondence is not com- plete; for the basal part of the radial trachea of the fore wing lies in what is without doubt the cavity of subcosta; and in both wings the medial trachea does not extend into the branches of the media. In the course of a study of the development of the wings of the honey- bee, Dr. Needham and I discovered the cause of the lack of con-espondence 370 THE WINGS OF HYMENOPTERA between the tracheation and venation of the wings in the Hymenoptera. An examination of the wings of young pupae of this insect revealed the fact that in this insect the laying out of the wing-venation precedes the trachea- tion of the wing. After the wing- veins reach that stage of development in which they appear as pale bands, the tracheae grow out from the base of the wing into them. Figure 390 represents the wings of a pupa taken at a stage that illustrates this pushing out of the tracheae into the previously formed wing-vein cavities. It is obvious that tracheae developed after the vein-cavities are formed will follow the paths offering the least resistance to their progress ; and it is not to be expected that they will preserve their primitive arrangement under these conditions. This brings us to the conclusion already announced, that in deteiTnining the homologies of the wing-veins in the Hymenoptera we are forced to base our conclusions on a study of the veins themselves; and that a method of study which is of the highest importance in determining the homologies of the wing-veins in many other insects, is of little use here for this s]3ecial purpose. id) METHODS OF MODIFICATION OF THE PRIMITIVE HYMENOPTEROUS TYPE OF WING-VENATION The working out of the details of the modification of the wing-venation in the several families of the Hymenoptera can only be done by those who make extended studies of these families, a work that does not fall within the scope of this essay. I wish merely to indicate, in this chapter, the essential features of the hymenopterous type of wing- venation, the more prominent of the methods by which this type is modified, and to illustrate by a few examples the nature of the results that have been attained in the modifica- tions of the primitive hymenopterous type. If this is done, data will be available that will facilitate more extended studies in this field. Our knowledge of the various modifications of the primitive hymenop- terous type of wing-venation that occur in the more generalized of the two suborders of this order, the Chalastogastra, is very complete. This is due to the labors of Dr. A. D. MacGillivray, who has published the results of a very extended and detailed study, illustrated by nearly one hundred figures of the wings of members of this suborder (MacGillivray '06). This most excellent work, with its great wealth of details, furnishes all the data needed for an understanding of the venation of the wings of the Chalastogastra, and gives an invaluable basis of comparison for the study of the wings of the more specialized suborder, the Clistogastra. I shall not attempt to give an abstract of Dr. MacGillivray's paper; because a mere abstract could not take the place of the complete memoir, THE WINGS OF HYMENOPTERA 371 which should be consulted by anyone making a study of the wing-venation of members of this suborder. But I shall have occasion, in the course of the later discussion, to quote some of his results. Passing to the suborder Clistogastra, we find it comparatively easy to determine the homologies of the wing-veins in many forms, that is in those in which the venation is not greatly reduced, by making use of the data obtained in the study of the Chalastogastra. But there are many forms with greatly reduced wing-venation in which the problem is exceedingly difficult . Much has been done in this field by Professor J. Chester Bradley, but unfortunately only a part of his results has been published. In his monograph of the Evaniidse (Bradley '08), Professor Bradley has figured all of the types of wings known to occur in this remarkable family. In another very extended work, entitled The Wings 0} Hymeyioptera with partictdar reference to the Ichneumon Flies, some of the more perplexing problems that have arisen in the study of the modifications of the venation of the wings of the Hymenoptera are discussed in a very thorough manner; but this work is not yet published. The m.ore important of the methods by which the primitive type of wing- venation has been modified in the Hymenoptera are the following: A reduction in the number of the wing-veins by the atrophy of one or more veins. A reduction in the number of the wing-veins by the coalescence of adjacent veins in one or more areas of the wing. A change in the course of a vein by the coalescence of its base with an adjacent vein. Changes in the courses of veins by the coalescence of the tip of each with an adjacent vein. The formation of serial veins. . The following examples will serve to illustrate the results of the modifi- cation of the primitive type of wing-venation by these several methods. It is an interesting fact that the wings of the most generalized of living Hymen- optera show modifications by each of these methods. The wonderfully modified wings of the more specialized Hymenoptera are merely the result of carrying to an extreme methods of modification already inaugurated in the most primitive Hymenoptera known to us. Reduction by atrophy — In the Hymenoptera one result of the specializa- tion of the venation of the wings is a reduction of the number of wing-veins; in no case are either accessory veins or intercalary veins developed; and even in the most generalized members of the order known to us not all of the wing- veins are retained. The reduction in number of the wing-\'eins, however, is slight in the more generalized members of the order; as already indicated, in some only vein R2 is lacking, in others, only veins Cuo has been lost. This refers to 372 THE WINGS OF HYMENOPTERA Fig- 391- — Wings of Pa)nphilins. The veins are lettered. Fig. 392. — Wings of Macroxyela. The cells arc lettered. THE WINGS OF IIYMENOPTERA 373 the fore wings ; in the hind wings, the reduction of the venation is greater than this in all cases. The hind wing of Pamphilius (Fig. 39 1 ) is an example of the most generalized condition found in the hind wings of Hymenoptera. While the reduction of the venation of the fore wings is slight in the more generalized members of the order, it is extreme in the more specialized Fig. 393. — Wing of an ichneumon-fly. families, as in the Chalcidid^e, where frequently only a vestige of the wing- venation remains. In the Hymenoptera, as in other orders where the specialization of the wing-venation is by reduction, the lessening of the number of wing-veins is the result partly of the atrophy of veins and partly of the coalescence of adjacent veins. The beginning of the atrophy of a vein is illustrated by the subcosta of the hind wing of Pamphilms (Fig. 391), the basal part of which has faded out. In the hind wing of Macroxyela (Fig. 392) the atrophy of the subcosta is complete. In the fore wing of Paniphilins vein Ro has been lost; and in the fore wing of Macroxyela vein Cu2 is lacking. A comparison of these two wings makes evident the fact that in each case a vein has atrophied. rsl A Fig. 394.^\Ving of a braconid. In Phamphilius no vestige of the lost vein Ro remains. In Macroxyela a bend in the cubitus indicates the position of the fonner forking of this vein. In many other cases, however, vestiges of lost veins remain either as short spurs or as faint lines. 374 THE WINGS OF HYMENOPTERA Reduction by coalescence. — An excellent illustration of the reduction in number of distinct wing-veins by the coalescence of adjacent veins is to be seen in the fore wing of an ichneumon-fly (Fig. 393), where the thickened margin of the wing, between its base and the stigma, is formed of the united costa, subcosta, radius, and media. A similar condition exists in the fore wing of a braconid (Fig. 394). In many of the Clistogastra in which the costa remains separate, the subcosta, radius, and media ccalesce from the base of the wing to the stigma. Modification of the course of a vein by coalescence at base. — Perhaps the best illustration of the modification of the course of a vein by the Fig. 395. — Wings of Manoxyela (From MacGillivray). coalescence of its base with an adjacent vein is that of the media of the fore wing; for in this case a very complete record of the change exists among living Hymenoptera. Correlated with the modification of the course of media are certain changes in the relations of the medio-cubital cross-vein; hence the two veins will be treated together. In Manoxyela, as was shown by Dr. MacGillivray (Fig. 395), the main stem of the media retains a cjuite primitive course except that the basal part coalesces with the radius for a considerable distance; for from the point where it separates from the radius it extends directly toward the outer margin of the wing. In this genus, the medio-cubital cross-vein also retains a comparatively primitive position (Fig. 395, m-cu). In Macroxyela (Fig. 392), the medio-cubital cross-vein has migrated towards the base of the wing, the coalescence of the media with the radius has been extended somewhat, and the section of the media between the radius and the medio-cubital cross-vein has become nearU^ transverse. THE ir/A'G-S- OF HYMENOPTERA 375 In 2\pis (Fig. 396), the position of the medio-cubital cross-vein is nearly the same as in Macroxyela; but the coalescence of the media with the radius has extended much farther, with the result that the section of the media that is between the radius and the medio-cubital cross-vein extends backward from the point where it separates from the radius to its junction with the medio-cubital cross-vein; from this point the media extends out- ward as in the more generalized forms. This results in the course of the main stem of media being Z-shaped. In the forms in which the condition just described exists, and this is true of a very large portion of the members of the order, the section of the media that is between the radius and the medio-cubital cross-veins forms with this cross- vein what appears to be a cross-vein, extending from the radius to the cubitus, from which the main stem of the media appears to arise. But it is now obvious that this apparent cross-vein is a serial vein consisting of a section of the media and the medio-cubital cross-vein. In Apis, and in many other genera of the more specialized H\TTienop- tera, the shortening of the cubitus as a result of an extending of the coales- Fig. 396. — Wings of Apis. vSee Fig. 397, for more detailed lettering. cence of veins Cui and ist A has left the cubital end of the medio-cubital cross-AX'in stranded upon vein iM,i. Modification of the course of a vein by the coalescence of its tip with an adjacent vein. — Examples of this kind of modification of the courses of veins have been described in the discussion of the venation of the wings of the more generalized Hymenoptera. Attention was called there to the changes that have taken place in the courses of the branches of the radial sector, the branches of media, and the branches of cubitus. Some of the modifications of the courses of veins which are evident in the more general- ized H^TTienoptera are greater in the more specialized Hymenoptera. Note 376 THE WINGS OF HYMENOPTERA for example the course of \'ein Mi in Apis (Fig. 396). Here this vein extends towards the base of the wing, with the result that the free part of it appears to be a continuation of the cubitus. Serial veins. — Perhaps the most astonishing and perplexing of the results that have followed the changes in the courses of veins in the wings of the Hymenoptera are those that have resulted in the formation of serial veins ; that is veins that appear to be simple veins but which in reality are compound veins composed of sections of two or more ^'eins joined end to end. Serial veins are discussed in Chapter III (page 69). In addition to the serial vein described there, vein ^n & M2 of the braconid wing, the follow- ing examples may be cited. In the fore wing oiApis (Fig. 396), the cubitus appears to extend from the base of the wing to a point near the middle of the wing. But, as shown above, only the basal part of this vein is cubitvis, the distal part being vein M4. If reference is made to this vein as a whole, it should be designated as vein Cu & M4. In the case just cited the fomiation of the serial vein is the result in a change in the course of one element of it. Another method of the formation of serial veins is by the atrophy of a section of a vein, which leaves the terminal portion of the vein stranded upon some other vein. Vein m & M2 of the braconid wing was formed in this manner. A remarkable instance of the atrophy of a section of a vein and the consequent formation of a serial vein that has the appearance of being a simple one is that described by MacGillivray ('06) as the switching of the base of the radial sector. This example of the formation of a serial vein is of especial interest; for this switching of the base of the radial sector has occurred in the wings of nearly all members of the suborder Clistogastra, and is found in only a few members of the suborder Chalastogastra. A very complete series illustrating the switching of the radial sector was found by MacGillivray and is shown in Figure 397. Before studying this series, however, let us examine the fore wing of Paniphilius (Fig. 391) and note the point of origin of the radial sector and the position of the radial cross-vein in what we have accepted as a typical hymenopterous wing. An examination of this wing and of the very extended series of wings of Chalastogastra figured by MacGillivray, shows that the radial sector separates at or before the base of the stigma, and that the radial cross-vein extends from near the middle of the stigma or from a point be^'-ond the middle, in a more or less nearly transverse direction, to the radial sector. In Figure 397, a, 6, and c show a part of the fore wing of three members of the Cephidae, one of the more specialized families of the Chalastogastra ; and d represents the same part in Oryssus, of the Oryssidae, the family in which the greatest reduction of the wing-venation found in the Chalasto- THE WINGS OF HYMENOPTERA 377 C^Ii^M gastra occurs. The other two figures, e and/, represent this part of the wing in two members of the suborder Chstogastra, Pelopazus and Apis respectively. In Macrocephiis satyr us (Fig. 397> a) ^he direction of both the basal part of the radial sector and of the radial cross-vein is oblique; and the radial cross-vein has the appearance of being a longi- tudinal vein branching from the stigma, and not that of a cross-\'ein, a foreshadowing of its function in more special- ized forms. In Janus cyuoshati (Fig. 397, h), a short section of the base of the radial sector near the stigma has faded out. In Janus abbreviatus (Fig. 387, c), a longer section of this vein is lacking. And in Oryssiis abieiinus (Fig. 397, c?),all of the radial sector proximad of the radial cross-vein is lack- ing; and this cross-vein appears to be the base of the radial sector. Thus is fonned a serial vein, consisting of the radial cross- vein and the rem- nant of the radial sector, which has the appearance of a simple xem. This is the m cu Fig. 397. — The switching of the base of the radial sector (From MacGillivra}')- condition that exists in all of the Chstogastra and is ilhistrated by PelopcBus cementarius (Fig. 397, £^) and Apis mcllifica (Fig. 397,/)' Figures 398 and 399, representing two entire wings, will illustrate two stages in this switching of the base of the radial sector. In Janus abbrevia- tus (Fig. 398), the atrophy of the section of the radial sector between the point where it is joined by vein r-m and the stigma has begun; in Odon- taulacMS editus (Fig. 399) the atrophy of this part of the radial sector is complete. Correlated with the series of changes just described are changes in some adjacent parts that merit description, and which were also pointed out by Dr. MacGillivray. The parts refeiTcd to are the radio-medial cross-vein and that part of the radial sector that lies between the radio-medial cross- vein and the radial cross- vein. 378 THE WINGS OF IIYMENOPTERA In Oryssus (Fig. 397, ci), a form in which there has been a great reduction of the wing-veins, these parts are both wanting. But in the other forms shown in Figure 397 these parts are retained. In the first three (Fig. 397, a, b, c) these two parts taken together constitute a shghtly bent vein; in the fifth (Fig. 397, e) they form a perfectly straight vein ; and in the last (Fig. 397, /), one that is slightly bent. In the last two forms, and in nearly all other members of the Clisto- Fig. 398. — A fore wing of Janus abhreviatus. gastra, there is no trace of that section of the radial sector extending from the radio-medial cross-vein to the stigma ; hence there is nothing to indicate the compound nature of the vein composed of the radio-medial cross-vein and that section of the radial sector between this vein and the radial cross- vein. For this reason this serial vein was formerly believed to be merely the radio-medial cross-vein. It should now be designated as vein r-m The switching of the base of the radial sector and the changes described in the two preceding paragraphs were not understood until the publication of Dr. MacGiUivray's paper in 1906. As Comstock's Manual for tlie Study Sc^R'rM Fig. 399. — A fore wing of Odontaulacus cdilus. of Insects, in which was made the first attempt to apply the vmifonn terminology of the wing-veins to the wings of Hymenoptera, was published in 1895. the figures of wings of the Clistogastra in that work are, in some respects, incorrectly lettered. THE WINGS OF HYMEXOPTERA 379 (e) AN ILLUSTRATION OF THE REDUCTION OF THE WING-VENATION IN HYMENOPTERA Having determined the homologies of the wing-veins in the more generahzed members of the Hymenoptera and having attained an under- standing of the various ways in which the typical wing-venation is modified in the more specialized forms, it is now comparatively easy to trace the Fig. 400. — Fore wing of Aulacinus Jusiger (.After Bradley). Fig. 401. — Fore wing of Evajiia appendigastcr (After Bradley). Fig. 402. — Fore wing of Acanthinevania principis. homologies of the wing-veins in all of the families of this order in which the venation of the wings is not greatly reduced. There are, however, certain families in which the various methods of modifying the primitive type of wing-venation have progressed so far that it is exceedingly difficult to determine the nature of the result and the steps by which it has been reached. It is in this field that lies the greater amount of the work that remains to be done to complete our understanding of the 380 THE WINGS OF HYMENOPTERA wing-venation of the H\TTienoptera, an understanding which will doubtless make available much data of great value in working out the relationships of the various divisions of the order. The only way in which the identity of the remaining veins in a wing in which the venation is greatly reduced can be satisfactorily determined is by comparison with allied foiTns in which the reduction has not been earned so far. And the more complete the series that can be found illus- Cu&Cui istA Fig. 403. — Fore wing of Semceodogaster barlicensis. trating the successive steps in the modification of the wing-venation the more satisfactory will be the conclusion. As an illustration of this method of study and of the results that can be obtained by it I include here a series of figures taken from Professor Bradley's monograph of the Evaniidos (Bradley '08). Figure 400 represents the venation of the fore wing of Anlacinus fusiger, (^^^ + Cui Fig. 404. — Fore wing of Hyptiu. a member of the subfamily Aulacinae. This is selected as the first of the following series as it is the most generahzed wing of the family. Note that veins R4, R5, and all of vein M2 are present. It is in the subfamily Evaniinai that the most extensive modification of the wing-venation is found; for this reason the remaining members of this series of figures have been selected from those illustrating this sub- family. THE WINGS OF HYMEN OPT ERA 381 Figure 401 represents the venation of the fore wing of Evania appeiidi- gaster; in this wing the modifications of the evaniid type as illustrated by Aulacinus are not great. Veins R4 and Rj are lost, and also the transverse part of vein Mo. The coalescence of media with radius has progressed Fig. 405. — Fore wing of Evaniellns. farther; this has resulted in a striking change in the direction of the cross- vein m-cit. The reduction of the wing-venation by the atrophy of veins which is well under way in Evania has been carried farther and farther in the succes- sive forms represented in this series of figures (Fig. 402, 403, 404, and 405) until in the last, EvauieUiis, only two veins are left ; these are the costa and veinSc + R M- M. CHAPTER XXVI THE TEACHING OF THE UNIFORM TERMINOLOGY OF THE WING-VEINS OF INSECTS SUGGESTIONS TO TEACHERS Much use is now made in systematic entomology of the characters presented by the wing-veins of insects. It is important, for this reason, that the student should acquire, early in his study of insects, a clear knowl- edge of the fundamental type of wing- venation, of the ways in which this type has been modified in the different orders of insects, and of the uniform terminology used in describing the wing-veins. The following outline of a course of study in this field is offered in the hope that it may be of service to teachers ; it is based on an outline that has been in use in the entomological laboratory of Cornell University for many years, where the work indicated in it has formed a part of the introductory courses in entomology. These courses have been planned to meet the needs of two classes of students: first, those students who are specializing in entomology; and second, those students who take only a general lecture course and the accompanying practicmns in this subject. The students who are specializing in entomology are furnished mounted specimens of the wings to be studied whenever it is practicable to do so, and are required to make drawings of them representing the wing- venation. In those cases where wings are not available for use in the laboratory, figures of the wings in which the veins are not lettered are furnished. In either case the student determines the homologies of the wing-veins and records his conclusions by lettering the veins in his figures. Each figure is criticised by the instnictor before another wing or figure is issued for study. This is of prime importance; for without such criticism an error made in one figure is likely to be repeated in the next, and an incoiTcct conclusion becomes established in the mind of the student. The course outlined here is merely an introductory course in which the student becomes grounded in the fundamental principles of the study of wing- venation. The wings to be studied in it have been selected with a view to illustrate the various ways in which the primitive type of wing- venation has been modified in the course of the evolution of recent forms. This course of study, or one of similar scope, should be completed before the student is encouraged to undertake an investigation of the wing-venation of a particular group of insects, as a family or an order; he should be impressed with the fact that the interpretation of venational characters must be based upon a knowledge of the various ways in which the wing- (382) THE WINGS OF INSECTS 383 venation has been modified in other groups of insects than that which he is studying; and that the easiest way to gain this knowledge is by a study of a carefuhy selected series of wings illustrating these modifications. Having obtained the essential preliminary knowledge, the student can then proceed profitably with his investigation. He should first determine which are the most generalized wings to be found in the group of insects that is being studied, the wings that most closely resemble the hypothetical primitive type; and when this has been done, he should determine the various ways in which the more specialized wings ha^x been modified. Fig. 406. — Fore wing of Pleronidea ribesii. In planning the work to be done in practicums accompanying a general introductory^ lecture course, one is hampered by the limited amount of time that can be devoted to each division of the more general subject. It is obviously impracticable to require the students to make many drawings of wings; on the other hand, it is essential that they should have practice in determining the homologies of the wing-veins in a considerable number of /s/ ^ Fig. 407. — Fore wing of Sirex. wings, if they are to gain a knowledge that will enable them to make use of the analytical tables in the text books in which wing-venational characters are used. We have met this difficulty by requiring these students to make draw- ings of only one or two wings, just enough to give them a little experience in the study of actual wings, and then furnishing them with figures of wings in which the veins are not lettered. These figures are issued one at a time, and each is to be properly lettered before another one is issued. Lettered figures of most of the wings used in this course are given in the preceding pages; the others, two in number, are represented by Figures 406 and 407. 384 THE WINGS OF INSECTS MATERIALS NEEDED FOR THIS COURSE A wall chart showing the hypothetical tracheation of a wing of the primitive nymph. This is a copy of Figure 411 of the following outline and is used in an introductory lecture in which the fundamental principles, briefly indicated in the introduction to the outline, are more fully explained. Mounted wings to be studied by the students specializing in entomology. It is well to have several sets of these, so that more than one student can be studying the same kind of wing at the same time. If the mounts are carefully used, they will serve for many successive classes. Mounted wings for use in practicums. A limited number of kinds will be required; but there should be as many mounts of each kind as there are students in a section of the class, as all will need the same kind of wing at the beginning of the practicimi. Sets of printed figures of wings for use in practicums, where the students do not have the time necessary for making original drawings of all of the kinds of wings studied, and for use in other cases when desired. These sets can be obtained of The Comstock Publishing Company, Ithaca, N. Y. Each figure is printed on a separate sheet. Figure 408 is a copy of one of these figures. A printed outline, one for each student, of the laboratory work in this subject. This outline consists of a reprint of the following pages of this chapter. The preceding pages, being merely suggestions for the use of teachers, are not included in the outline of laboratory work. Copies of this otttline can be obtained also of the publishers of this volume. Red ink for marking the lines representing media and its branches in the figures of wings. The use of red ink for this purpose adds greatly to the clearness of the figures. In those cases where the stem of media or any of its branches coalesce with another vein, this fact should be indicated by drawing a red line, indicating the course of media, closely parallel with the black line indicating the course of the vein with which media coalesces. Red ink is also used for lettering the cells of the wings. Blank paper uniform in size with that upon which the figures are printed for use in making the notes called for in the outline. This will permit the keeping together of the drawings and notes for reference. THE WINGS OF INSECTS 385 i f ^ r \ \ -1 ^/^ \[ / / \ •^ J 1 / ■ "1 \ K y\ /*"^ 1 '■^ "^ •f H X \ f M s -^ J^" \j / A / 1 ^W i 1 /*■ W^>\J ri V\ / / 1? i >W i/ I \ 1 tt ^+j "» I I 1 ft 0 5, I 7 • S § 0 ^^ i-. 0 1 \ , / ¥ Fig V . 408.— y A.n exair pie of the figures used in laboratory work. OUTLINE OF LABORATORY WORK IN THE STUDY OF THE VENATION OF THE WINGS OF INSECTS* IXTRODUCTIOX In form an insect's wing is a large, membranous app2ndage, which is thickened along certain lines. These thickened lines are termsd the veins or nerves of the wing; and their arrangement is described as the venation or neuration of the wings. It has been found that the venation of the wings of closely allied insects is very similar, and that great differences in this respect exist between insects remotel}' connected. Hence, the wings afford excellent characters for use in the classification of insects. In fact, as slight differences in \'enation are easily observ^ed, the wings being spread out like an open page, these differences are probably the most available characteristics of winged insects for taxonomic work. It is important, therefore, that the student of entomology should learn early in his course the more important facts regarding this subject. A careful study of the wings of many insects has shown that the fuida- mental type of venation is the same in all of the orders of winged insects. But this fact is evident only when the more primitive or generalized mem- bers of different orders are compared with each other. In most of the orders of insects the greater number of species have become so modified or specialized as regards the structure of their wings that it is difficult at first to trace out the primitive type. Note. — The student should have a clear idea of the significance of the terms genera- lized and specialized, which are now much used in biology. Generalized indicates a primitive condition, a nearness to ancestral forms. Thus, the most generalized member of a group (as a family or an order) is that memb2r which most clearly resembles the ancient progenitor of that group. Specialized, on the other hand, indicates remoteness from the primitive type, an adaptation to more special conditions of existence. Thus, the most specialized member of a group is the one that departs most widely from the ancient progenitor of that grouj). These terms are used in a com"parative sense; thus, a highly specialized form may be regarded as generalized when compared with forms that are still more highly specia- lized. This agreement in the important features of the venation of the wings of the generalized members of the different orders of insects is still more evident when the wings of nymphs and pupx are studied. It has been demonstrated that in the development of the wings of generalized insects the longitudinal wing-veins are formed about preexisting tracheae. In the course of the development of the wing, these tracheae grow out into the wing-bvid, and later the wing-veins are formed about them. *The material for this outline has been drawn largely from The Elements of Insect Anatomy, by John Henry Comstock and Vernon L. Kellogg, published by the Com- stock Publishing Company, Ithaca, N. Y. (387) 388 THE WINGS OF INSECTS The wings of nymphs and pupae are broad at the base, and consequently the tracheae that precede the wing-veins are not crowded together as are the wing-veins at the base of the wings of adults. For this reason the identity of the wing-veins can be deteraiined more surely in the wings of nymphs and Fig. 409. — Hypothetical tracheation of a wing of the primitive nymph. pupae than they can be in the wings of adults. This is especially true where two or more veins coalesce in the adult wing while the tracheae that precede these veins are distinctly separate in the immature wing. A study was made of the tracheation of the wings of nymphs and pupae of representatives of most of the orders of insects, and, assuming that those features that are possessed by all of these must have been inherited from a common ancestor, a diagram was made representing the hypothetical tracheation of a nymph of the primitive winged insect, (Fig. 409). In this diagram the tracheae are lettered with the abbreviations used in designating /?, R.^, Fig. 410. — A wing of Khyphus, with the veins and cells lettered. the veins that are formed about them in the course of the development of the wing. The diagram will serve, therefore, to indicate the typical \'ena- tion of an insect wing, excejDt that the trachea; are not crowded together at the base of the wing as are the vein s in the wings of adults. THE WINGS OF INSECTS 389 Figure 410 represents the wing of an adult fly of the genus Rhyphus. This wing is comparatively generalized, but in several respects it depa-ts from the primitive t>'pe; the radius has been reduced to a three-branched condition; the media is also reduced to a three-branched condition; only vestiges of the first and third anal veins remain ; these are represented by dotted lines; and a part of the stem of media is atrophied. By studying a wing of Rhyphus and the accompanying figure (Fig. 410) the student can gain a good idea of the type of the wings of insects bslonging to the order Diptera, and have a standard with which to compare wings of insects of other orders. Longitudinal veins and cross-veins. — The veins can be grouped under two heads: first, loigiiudmal veins, those that normally extend lengthwise the wing ; and second, cross-veins, those that noTTnally extend more or less nearly transversely. In Figure 410, three of the cross- veins are indicated by arrows, near the middle of the wing; two other cross-veins are repre- sented near the base of the wing. All other veins represented in this figure are longitudinal veins. The insertion of the word normally in the above definitions is impartaat; for it is only in comparatively generalized wings that the direction of a vein can be depended upon for determining to which of these two classes it belongs. A little later the student will study wings in which the direction of some of the longitudinal veins has been so modified in the course of specialization that they extend transversely {i. e., cephalo-caudad), and some cross- veins extend in a longitudinal direction {i. e., proximo-distad) . Simple veins and branched veins. — Veins are either simple or branched. The veins lettered Sc and 2d A in Figure 410 are simple veins; between these there are three branched veins. In the case of branched veins the entire vein, including all of its branches is often referred to as a single vein. Thus the third vein in the wing Rhyphus, counting the thickened, cephalic margin of the wing as the first vein, is termed the radius or vein R; and by this expression we include both the main stem of the vein and its three divisions. On the other hand, each division of a branched vein is oiten termed a vein. Tiius tlu first division of the radius, counting from the cephalic margin of the wing, is termsd radius-one or Vein Ri, and the second division, radius-twj or vsin Ri, and so on till all are numbered. Note. — In the most generalized flies known to us, the radius is five-branched. But in most flies some of the branches of this vein coalesce so that the number of apparent branches is less than live. In Rhypus veins R^ and Ry coalesce so as to appear as a single vein. In order to indicate that this apparently simple vein is composed of two veins, and in order that homologous veins in different insects shall bear the same designation, this compound vein is termed radius-two- plus-three or vein R0+3. In the same way, what appears to be the third branch of the radius in Rhyphus is really the fourth and fifth coalesced, and is, therefore, designated as radius-four- plus-five or vein R[+a. The tracing out of the homologies of the branches of veins is often ver^" difficult; but it is of the greatest importance in determining the relationships of different genera or of families. 390 THE WINGS OF INSECTS Names of the longitudinal wing-veins. — There have been many different sets of names appHed to the veins of wings. Not only have the students of each order of insects had a peculiar terminology, but in many cases different writers on the same order have used different sets of terms. This condition of affairs was incident to the beginning of the science, the period before the correspondence of the veins in the different orders had been worked out. But now the time has come when it is practicable to apply a uniform terminology to the longitudinal wing-veins of all orders ; and the following set of terms has been proposed for that purpose : Costa. — The vein extending along the cephalic or costal margin of the wing is the costa. Subcosta. — Immediately caudad of the costa and extending parallel with it, is a vein, which is usually simple in flies; this is the subcosta (Fig. 410, Sc). Radius. — Immediately caudad of the subcosta there is a vein which in generalized insects is always branched; this is the radius. In Rhyphus, the radius is three-branched (Fig. 410, R\, R0+3, and Ri+h-) The radial sector. — When the radius preserves its primitive mode of branching, it separates at its first fork into two unequal parts; the first of these is vein Ri ; the other gives rise to the remaining four branches of the radius (Fig. 408). This second part of the radius, including its branches, is termed the radial sector or vein R^. Media. — Traversing the middle of the wing there is a longitudinal vein which is always branched in generalized insects; this is the media. In Rhyphus the media is only three-branched (Fig. 410, Mi, Mi, and M3) and a part of its stem is atrophied. Cubitus. — The third and last of the branched veins in flies is the cubitus. This vein is two branched in Rhyphus (Fig. 410, Cui and Cuo^. Anal veins. — Caudad of the cubitus there is in Rhyphus a single well- developed vein ; this is termed an anal vein . As in more generalized insects there are three anal veins, and as this is the second of the series, it is desig- nated the second anal vein (Fig. 410, 2d A). Vestiges of the first and third anal veins persist in Rhyphus; these are indicated in the figure by dotted lines. Designation of the longitudinal wing-veins by numbers. — Several writers have designated the longitudinal wing-veins by numbers. In Comstock's Manual for the Study 0} Insects both the names given above and numbers are used. The following table indicates the correspondence of the names and numbers. Costa = vein I. Cubitus = vein VII. Subcosta = vein II. ist anal vein = vein VIII. Radius = vein III. 2d anal vein = vein IX. Media = vein V. 3d anal vein = vein X. It will be observed that in the above table the numbers IV and VI are omitted. At the time the Manual for the Study of Insects was published, it was believed that THE WINGS OF INSECTS 391 two other longitudinal veins (the so-called premedia and postmedia) were present in certain orders of insects, and the numbers IV and VI were applied to these veins. It has since been determined that this conclusion was based on an error. Certain writers number the wing veins without omitting the numbers IV and VI, designating the media as vein IV and the cubitus as vein V. There are still other writers who do not regard the costa as a true vein, and, there- fore, designate the subcosta as vein I. The result is that there are three distinct systems of numbering the wing-veins,^in addition to several old systems which were applied to single orders. It seems better, therefore, to designate the wing-veins by names, and use abbreviations of these names in lettering figures. Names of the cross-veins. — In the wings of certain insects, as the dragon-flies, the May-flies, and others, there are many cross-veins; it is impracticable in cases of this kind to name them. But in several of the orders of insects there are only a few cross-veins, and these have been named. Figure 411 represents the hypothetical primitive type of wing- Fig. 411. — Hypothetical type of wing-venation with the named cross- veins added. venation with the named cross-veins added in the positions in which they normally occur; these are the following: The humeral cross-vein. — This extends from subcosta to costa near the humeral angle of the wing (Fig. 411, h). The radial cross-vein. — This extends between the two principal divisions of radius, i. e. from vein Ri to vein R^ (Fig. 411, r). The sectorial cross-vein. — This extends between the principal divisions of the radial sector, i. e. from vein Ro-t-s to vein R4-1-5 or from vein R3 to vein R4 (Fig. 411. s). The radio-medial cross-vein. — This extends from radius to media, usually near the center of the wing, (Fig. 411, r-m). When in its typical position this cross-vein extends from vein R^-f-o to vein Mi-^2- TJie medial cross-vein. — This extends from vein M2 to vein M3 (Fig. 411, m). This cross-vein divides cell Mo into cells ist M2 and 2d Mo; see Figure 410, where the cells are lettered. 392 THE WINGS OF INSECTS The medio-cuhital cross-vein. — This extends from media to cubitus (Fig. 411, m-cn). The arctilus. — In many insects there is what appears to be a cross-vein extending from the radius to the cubitus near the base of the wing. This has been termed the ar cuius by writers on the Odonata, and the use of this term has been extended to all orders in which there is a similar arrangement of the veins in this part of the wing. The arculus is designated by the abbreviation ar. Usually when the arculus is present the media appears to arise from it. The fact is, the arculus is compound, being composed of a section of the media and a cross-vein. The structure of this part can be clearly seen in the Odonata (Fig. 412). In Rhyphus (Fig. 410) the arculus appears as a simple cross-vein extending from the radius to the cubitus, and a part of the base of the media is atrophied. R^M Fig. 412. — The arculus of a dragon-fly. That part of the arculus which is a section of media is designated as the anterior arculus (Fig. 412, a a), and that part formed by a cross-vein, the posterior arculus (Fig. 412, p a). Designation of the cells of the wing. — The thin spaces of the wings which are bounded by the veins are called cells. In descriptions of wings it is often desirable to refer to one or more of the cells. It is necessary, there- fore, to have a terminology of the cells of the wing, as well as of the wing- veins. Having named the wing-veins, the simplest possible method of desig- nating the cells of the wing is to apply to each the abbreviation of the name of the vein that forms its cephalic (front) margin. It should be borne in mind, however, that by modifications of the typical arrangement of the wing-veins, a vein that normally forms the cephalic margin of a cell may bear a very different relation to it; and this must be taken into account if we are to apply the same term to homologous cells throughout the insect series. The cells of the wing fall naturally into two groups: first, those on the basal part of the wing; and second, those nearer the distal end of the wing. The former are bounded by the principal veins ; the latter, by the branches of the forked veins; a corresponding distinction is made in designating the cells. Thus the cell lying behind the main stem of the radius and on the basal part of the wing is designated as cell R; while the cell lying behind radius-one is designated as cell R\. THE WINGS OF INSECTS 393 It should be remembered that the coalescence of two veins results in the obliteration of the cell that was between them. Thus when veins R2 and Rz coalesce, as in Rhyphus (Fig. 410), the cell lying behind R2+3 is cell i?3, and not cell Ri+.u cell Ri having been obliterated. When one of these principal cells is divided intcj two or more parts by one or more cross-veins, the parts may be numbered, beginning with the proximal one. Thus in Rhyphus (Fig. 410) cell Ml is divided by the medial cross-vein into two parts, which are designated as cell 15/ M2 and cell 2d M-i, respectively. There are many cases where two or more adjacent cells have been united by the atrophy of the vein or veins separating them. A com- pound cell thus formed is designated by a combination of the terms applied to the elements of the compound cell. When, for example, the stem of media is atro- phied, the cell resulting from the combination of cells R and M is designated as cell R -F M. The application of this system of naming cells of the wing is an easy matter Fig. 413. — Wings of Bombyx mori. in those orders where the wings have few veins ; but in those orders where many secondary veins are developed it is more difficult of application. In the latter case we have to do with areas of the wing rather than with separate cells. Thus, for example, it will be seen later that in certain Neuroptera the area R2 is divided by several longitudinal veins, which are connected by many cross-veins, the area Ro (which is strictly homologous with cell R2) being composed of a large number of secondary cells. The corrugations of the wings. —The wings of comparatively few insects present a flat surface; in most cases we find that the membrane is thrown into a series of folds of corrugations. This corrugating of the wing in some cases adds greatly to its strength. This is well shown by the wings of 394 THE WINGS OF INSECTS dragon-flies; and in most orders, the costal margin of the wing is strength- ened by a fold between the costa and the radius, the subcostal fold. In other cases, the corrugations are the result of a folding of the wing when not in use; this is well shown in the anal area when this part is broadly expanded. It rarely happens that there is occasion to refer to individual members of either of these classes of folds, except, perhaps, to the one that has just been designated as the subcostal fold. The furrows of the wings. — There are found in the wings of many insects one or more suture-like grooves in the membrane of the wing; these are termed the furrows of the wing. The following furrows have received distinctive names. They occur chiefly in the fore wings. The anal furrow. — The anal furrow is usually either between the cubitus and the first anal vein or it is coincident with the first anal vein which it may supplant in forms in which the venation of the anal area is reduced. This is the case in the Lepidoptera and the Diptera, and is well-shown in the wings of Bombyx mori (Fig. 413), in which a vestige of the first anal vein is preserv^ed near the margin of the wing. The median furrow. — This is a longitudinal furrow which is usually between the radius and the media. ^ It is well-marked in many of the Fig. 414.— Diagram of a wing showing Hemiptera, where it separates the margins and angles. . embolium from the remanider ot the coirum; and in the Hymenoptera its course is marked by a series of weak spots (bullae) in certain veins. The nodal furrow. — This is a transverse suture beginning at a point in the costal margin of the wing, corresponding to the nodus of the Odonata, and extending towards the inner margin of the wing. It crosses a varying number of veins in different orders of insects. The axillary furrow. — The axillary furrow is a suture-like line extending from the base of the wing to the inner margin; it ends at the axillary exci- sion (a notch near the base of the wing) when this is present. Margins of wings. — ^An insect's wing is more or less triangular in outline ; it, therefore, presents three margins; the costal margin (Fig. 414, a-b)\ the outer margin (Fig. 414, b-c)\ and the inner margin (Fig. 414, c-d). Angles of wings. — The angle at the base of the costal margin (Fig. 414, a) is the humeral angle; that between the costal margin and the outer margin (Fig. 414, b) is the apex of the wing; and the angle between the outer margin and the inner margin (Fig. 414, c) is the anal angle. THE WINGS OF INSECTS 395 IDENTIFICATION OF THE \VIXG-VEINS AND OF THE CELLS OF THE WINGS IN DIFTERA The Diptera constitute a hit^hly specialized order of insec ts in which only one pair of wings has been presen'ed as organs of flight, and in most cases the venation of the remaining pair of wings exhibits striking variations from the primitive type. But in some members of the order the venation of the wings is nearly typical; and, as varying degrees of departure from the typical form exist, by studying a carefully selected series of wings one can obtain a knowledge of the manner in which the more striking variations from the ]:)rimiti\'e t}'pe have been evolved. In the wings of Protoplasa fitchii, one of the most generalized members of the order Diptera, the variations from the typical form are comparatively slight (Fig. 415). Note that while the costal trachea in the wings of nymphs and pupas extends parallel with the costal margin of the wing, but at some distance from it, as shown in Figure 409, in the wings of adults the costal vein coin- cides with the costal margin of the wing (Fig. 415, C). In Protoplasa fitchii the subcosta is forked and the radius is five- branched; these are two generalized features rarely found in the Diptera, usually the subcosta is not forked and the number of the branches of the radius is reduced. In all Diptera known to me except Protoplasa fitchii, media is only three- branched. In lettering his drawings, the student may omit any reference to vein M4, as is done in the figure of the wing of Rhyphus above. 2d A Cu, Fig. 415. — A wing of Protoplasa fitchii. In most Diptera the first anal vein has atrophied ; but there is usually a vestige of it, which is a suture-like line parallel with \-ein Cu; this is kno^^^l as the anal furrow; it is represented in the figure of the wing of Rhypluis by a dotted line. In Rhyphus there is also a vestige of the third anal vein, represented in the figure by a dotted line. Directions for the study of wings. — Make a drawing of the wing, based upon a carefvil stud)- of it with a compound microscope, using a low power. The drawing should be first made with a pencil; after it has been criticised by the teacher, the lines should be inked; ink the lines representing media 396 THE WINGS OF INSECTS with red ink. Make the drawings on a sufficiently large scale so that each vein can be represented distinctly, and on paper uniform in size with that used for the printed figures of wings used in this course. Letter each vein and cell of the wing, using black ink for the veins and red ink for the cells. Note the more important features of its venation, and especially the more important departures from the primitive type of the order as indicated by the generalized form first studied. In the Diptera the wing of Rhyphus (Fig. 410) may be used as a generalized type, although in certain respects other wings will be found to be more generalized. The following are some of the more important points to be noted: Whether the subcosta is simple or forked at the tip; the number of the branches of the radius; in this connection determine which of the radial cells has been obliterated by the coalescence of branches of the radius (study Fig. 409); the position of the radio-medial cross-vein; the number of the branches of the media; the division or not of cell M2; the presence or absence of cell M3 ; the courses of the branches of the cubitus ; the extent of the anal furrow, which is a vestige of the first anal vein; and the extent and course of the second anal vein. Wing of a tabanid.— A specimen of one of the horse-flies, Tabanus, will be given the student for examination. Observe the subcostal fold, and note that this corrugation stiffens the wing. Make a drawing of a mounted Tabanid wing, which will be furnished on application to the instructor. Note that in the mounted wing the sub- costa is more or less concealed by the radius, although the two veins are distinct, as was seen in the unmounted specimen. Represent these two veins as slightly separated in your drawing. Note a case of coalescence of veins not exhibited by Rhyphus. In the description of this wing, state in what respect it is more general- ized than that of Rhyphus, and in what respect it is more specialized. Wing of an asilid. -A wing of a robber-fly of the genus Erax will be used as an example. There is a spur projecting from one of the branches of radius in this wing. This is a secondary development. Such spurs are not uncommon in the Diptera ; there is one near the base of the radial sector in the wings of Proioplasa fitchii, (Fig. 415). Not all spur-like veins are secondary develop- ments; in many cases a spur is a vestige of a vein that is partly atrophied. Wing of abombyliid. — The example used is a wing of Pautarhes, one of the bee-flies. Wing of a scenopinid. — The wing used is that of a common window-fly, Scenopinus. Wing of an empidid. — The wing used is that of Rkaniphomyia, one of the dance-flies. THE WINGS OF INSECTS 397 Wing of a muscid. — The wing used is that of the common house-fly, Miisca domestica. Wings of dolichopodids. — The wings of two of the long-legged flies will be used. The first belongs to the genus Psilopodius; the second to the genus Dolichopus. Wing of a syrphid. — The wings of a fly of the genus Eristalis will be used. Note the \-ein-like structure between the radius and the media ; this is termed the spurious vein. Wing of Dixa. — The wing of a dixa-midge is used as an example of the venation of the midge-like flies. If a wing of Dixa is not available, use one of a mosquito. In the midge-like flies the number of the branches of the radial sector is reduced in a way different from that seen in the families previously studied. Compare with the asilid, the tombyliid, and the scenopinid. REVIEW" OF WORK ON WINGS OF DIPTERA In the preceding studies of wdngs of Diptera illustrations have been seen of A'arious ways in which these wings have been modified in the course of their evolution. It will clarify the subject if some of the more important of these methods of modification of the primitive type be studied separately. Reduction of the radial sector. — Note the follow ing facts : — In Protoplasa fitchii (Fig. 415) the radial sector is not reduced, being four-branched as in the hypothetical typical form of this vein. In Tabanns the radial sector is only three-branched. This is the result of the coalescence of two of the branches of this vein. In Dixa also the radial sector is reduced to a three-branched condition by the coalescence of two of its branches; but in Dixa not the same branches coalesce as in Tabanns. These two genera represent two kinds of specialization, which indicates that they belong to different lines of descent. The common progenitor of these two genera had a four branched radial sector; in some of the descen- dants of this primitive form one method of reduction has taken place, while in other descendants another method has been followed. That this differentation took place comparatively early in the history of the order is shown by the fact that in all Nematocera that ha\-e a three- branched radial sector veins R2 and R3 remain distinct; while in those Brachycera that have a three-branched radial sector veins R^ and R5 are separate. The Nematocera and the Brachycera are the two chief divisions of the order. In both divisions of the order the reduction of the radial sector is carried farther in many cases. Thus in Rhyp'ius the radial sector is only two- branched (Fig. 410). Two methods of coalescence of veins. — There are two methods of coalescence of wing-veins. By one method the coalescence proceeds out- 398 THE WINGS OF INSECTS wards, the point of separation of two veins moving nearer and nearer to the margin of the wing until it is reached and the two veins are completely united. The existence of this method of coalescence has been demon- started by studies of series of allied forms in which the successive stages are shown ; such series are easily found in the Lepidoptera, and it was probably by this method that the number of the branches of radius was reduced in Rhyphus and the other Diptera studied in this course. By the other method the coalescence begins at the margin of the wing and procedes towards the base of the wing; two series illustrating this method of coalescence are indicated below. Coalescence of veins Cuo and 2d A. — Study your drawings of the follow- ing named wings: — In the wing of Rhyphus veins Cui; and 2d A are widely separated at the tip. In the wing of Paniarhes veins Cuo and 2d A are approximate at the margin of the wing but are still separate. In the wing of Erax these two veins coalesce for a short distance at the margin of the wing. In the wing of Scenopinus the coalescence of these two veins has progressed to a considerable distance. In the wing of Rhamphomyia the coalescence of these two veins has progressed so far that vein Cuo extends towards the base of the wing. Coalescence of veins M3 and Cui. — Arrange your drawings of the wings of Tahanits, Erax, and Pantarhes in the order named and note the successive stages in the coalescence of veins M3 and Cui and in the obliteration of cell Ms. The uniting of cells. — An-ange your drawings of the wings of Musca domestica, Psilopodiiis sipho, and DoHchopus coquilletti in the order named. Note that in Musca cells M and ist M2 are separated by the free part of vein M3; in Psilopodins the free part of vein M3 is partly atrophied, only a short spur remaining ; and in DoHchopus the free part of vein M3 is entirely lost and consequently cells M and ist M2 are completely united. IDENTIFICATION OF THE WING-VEINS AND OF THE CELLS OF THE WINGS IN LEPIDOPTERA As the wings of Lepidoptera are covered with scales, it is difBcult to determine the nature of their venation without specially preparing them for this puri3osc. After a student has become familiar with the type of venation characteristic of the order, he can usually determine the course of any particular vein by jmtting a drop of chlorofonn on the part of the wing to be examined; this will render the veins more distinct for a few seconds. Or the scales can be removed from a small part of the wing with a small, THE WINGS OF INSECTS 399 artist's, sable brush. But when a very careful study of the venation of a wing is to be made, it should be bleached and mounted on a card or on a glass slip, in order that it may be studied with a compound microscope. The following is the method of bleaching wings : — 1 . Remove the wings carefully so as not to break the frenulum if there be one;* it is well to remove the patagium first. f 2. Dip the wings in alcohol in order to wet them. 3. Immerse them for an instant in hydrochloric acid (muriatic acid). Use for this purpose dilute acid, one part acid to nine parts water. 4. Put them in Labaraque solution with the upper surface of the wings down, and leave them there till the color has been removed from the scales. If a wing bleaches slowly, the process can be hastened by dipping it in the dilute acid and returning it to the Labaraque solution from time to time. This solution can be procured of most druggists. It deteriorates if left exposed in strong sunlight. If it cannot be obtained, use an aqueous solu- tion of chloride of lime. 5. When a wing is bleached, put it in alcohol and leave it there till after it floats. This is to wash off the Labaraque solution. The wing can then be mounted on a card. But it is better to mount it as described below. 6. Transfer the wing to a clearing mixture, if it is to be mounted in balsam, and leave it there five or ten minutes. This is to remove any water there may be on it. A good clearing mixture can be made by mixing two parts by measure of carbolic acid cr}^stals and three parts of rectified oil of turpentine. 7. Put the wing on a glass slip with considerable clearing mixture under it to avoid bubbles; put Canada balsam on top, and cover with a cover glass. In the case of small wings, it is best to transfer them from one solution to another, and to the glass slip by means of a camel's-hair brush. J Wings bleached and mounted in this w^ay make an important addition to a collection. The slides should be carefully laVjellcd; and the insect from which the wings were taken should be kept with the slide. It is our prac- tice to remove always the wings from the right side, and then to mount the slide in the collection at the right of the insect from which the wings were taken. Uniformity in this respect adds greatly to the appearance of the collection. Wings of a hepialid. — Figure 416 represents the wings of a hepialid, Pielus labyrinthccus, one of the most generalized members of the order Lepidoptera. It is introduced here to show that there exists among living Lepidoptera forms in which the venation of the wings is not greatly changed *i:hc frenulum is a strong spine or bunch of bristles borne by the hind wing at the humeral angle in most moths. jThe patagia are scale-like appendages at the base of the fore wings. jin the case of very small wings, as those of Tineids, the very fine veins are more distinct when mounted in glycerine-jelly than when mounted in Canada balsam. 400 THE WINGS OF INSECTS from that of the hypothetical primitive type. The most striking modifica- tion of the primitive type is the fact that in both fore and hind wings there are only three separate branches of media. Vein M4 coalesces either with vein M3 or with vein Cui. As there is doubt regarding the fate of vein M4 the student in lettering his drawings of lepidopterous wings may omit any reference to this vein. Wings of Prionoxystus. — Make a drawing of each wing. Note that the fore and hind wings differ greatly in venation. This is Fig. 416. — Wings of Pielus labyrinthecus. true of the wings of all Lepidoptera except one small family, the Hepialidae, of which Peilus labyrinthecus, the wings of which are represented in Figure 416, is a member. Letter the veins of the fore wing. Note the anastomosis of veins R3 and R4+5. In the hind wing veins vSc and Ri coalesce in the outer half of the wing and the radial sector is not branched ; these facts have been determined by studies of the tracheation of pupal wings. With this information before you, letter the veins of the hind wing. THE WINGS OF INSECTS 401 Wings of the Monarch Butterfly, Anosia plexippus. — Make a drawing of each wing. Study the fore wing first. Figure 41 7 is a reproduction of a photograph of a fore wing of a pupa of this species. In this figure the developing wing- veins appear as pale bands and the trachea? as dark lines. A study of the tracheation of this wing will aid in the determination of the wing-veins. In one respect the branching of the radial trachea differs markedly from that of the hypothetical type ; this feature is distinctively characteristic of the fore wings of butterflies, it does not exist in moths. Determine the significance of the three short spurs that project into the distal end of the large cell near the middle of the fore wing, the discal cell. Letter the veins of the hind wing, and the spurs that project into the discal cell. Note the \-estige of the first anal vein at the base of cubitus. Fig. 417. — Wing of a pupa of the Monarch butterfly. The reduction of media. — In a few of the more generalized families of the Lepdioptera the main stem of media is preserved ; this condition exists in the wings Prionoxystus, already figured. But in most Lepidoptera the base of media is wanting in the wings of adults, although the medial trachea may be well-preser\'ed in the pupae ; this condition exists in the wings of the monarch butterfly already studied. Correlated with the loss of the stem of media its branches become more or less closely united with radius and cubitus. In the fore wing of the Monarch Butterfly the base of vein M3 has moved towards vein Cui and away from its former position indicated by a spur projecting into the discal cell. WIXGS OF NEUROPTERA In the orders Diptera and Lepidoptera the more generalized members of each order possess the maximtmi number of wing-veins found in the order. It is those wings in which the maximum number of wing- veins exist that most closely resemble the hypothetical primitive t}'pe of wing- venation ; 402 THE WINGS OF INSECTS the variations from this type are the result of either the coalescence of veins or the atrophy of veins or of both of these processes. This fact is expressed by the statement that in these orders the wings are specialized by reduction. In certain other orders of insects, of which the Neuroptera is one, the wings that most closely resemble the hypothetical primitive type are those that have few wing-veins compared with the wings of other members of the order. In these orders the wings are specialized by addition. In those wings where the specialization has been by addition there are usually many cross- veins; these, as a rule, are inconstant in number and position, consequently, except in the case of a few of the more important Jf f rr ^7-7-77-7^77^. Fig. 418. — Wings of Polystoecholes punctatus. ones, which are not discussed in this brief introductory course, no effort is made to name them. Accessory veins. — The added longitudinal veins are of two types, each characteristic of different orders of insects. In the order Neuroptera, the added longitudinal veins are of the type that has been designated as accessory veins; they are \^eins that have been developed as branches of the primitive longitudinal veins. The wings of Polysioechotes punctatus (Fig. 418) can be taken as an illus- tration of wings that have been specialized by addition. In these wings there are few cross-veins compared with what is usually the case in highly specialized neuropterous wings; but these wings illustrate well a high development of accessory veins. THE WINGS OF INSECTS 403 Study Figure 418 carefully and letter with a pencil veins Sc, Ri, and the stem of Rg. Submit the lettering to the Instructor for criticism and make changes if necessary. There are two types of accessory veins, which arc designated as the marginal and the definitive respectively. 3^^ A 2d A CU2 Cuia Fig. 419. — Wing of a pupa of Corydalus. The marginal accessory veins are twig-like branches, that are the result of bifurcations of veins that have not extended far back from the margin of the wing. Many such short branches of veins exist in the wings of Poly- stoechotes piinctatiis. The ntraiber and position of marginal accessory veins are not at all constant ; they differ in the wings of the two sides of the same individual. The definitive accessory veins differ from the marginal accessory veins in having attained a position that is comparable in stability to that of the primitive branches of the principal-veins; for this reason it is practicable Fig. 420. — Wing of a nymph of a cockroach. to designate them individually. In Figure 41 8 one of the definitive acces- sory veins is designated as Ro^- Accessory veins are added to the principal veins in two ways: first, in some insects they are added distally by successive splittings of the tip of a principal vein, thus forming a regular series; and second, the number of 404 THE WINGS OF INSECTS accessory veins may be increased in an irregular manner by interpolation, i. e. by the splitting of various members of a series of accessory veins. Illustrations of the adding of accessory veins distally are to be seen in Corydalus and allied genera. The presence of fine twigs at the tip of trachea R2 in the pupal wing of Corydalus (Fig. 419) indicates the method of increase, which is doubtless as follows: the branches have been added one after another to the tip of tra- chea R2, there being a migration of the base of each accessory trachea towards the base of the wing, thus making room for the addition of new branches. In this case the first accessory vein is the proximal one. It is the oldest accessory branch of the radial sector that is lettered R2a in Figure 418. The successive branches are Rob, R2C, R2d. etc. In the wing of a n^Tnph of a cockroach represented by Figure 420 there are many accessory tracheae branching from the front side of the radial trachea. From the presence of the fine twigs near the apex of the wing, it is evident that accessory tracheae are being added distally. It is also evident that the number of accessory tracheae is being increased by the splitting of some of these accessory tracheae, i. e. by ainterpoltion. In cases of this kind it is impracticable to designate the accessory veins individually. The pectinate type of radial sector. — The order Neuroptera differs from all other orders of living insects in the fact that except in a few cases the radial sector has been so modified that it is of the form known as pectinate or comb-like; that is, it consists of a supporting stem upon which are Fig. 421. — Diagrams of several types of radius. THE WINGS OF INSECTS 405 borne a greater or less number of parallel branches. This type of radial sector is well-illustrated in the wings of Polystoechotes punctatus (Fig. 418). The transformation of a t\TDical dichotomously branched radial sector into one that is pectinately branched is usually produced by a splitting apart of veins R4 and R3 so that they arise separately from the supporting stem of the pectinate veins thus formed. In the accompanying series of diagrams (Fig. 421), the first diagram represents the manner of branching of the typical radius in which the radial sector is dichotomously branched. The second diagram represents a radius in which the radial sector has' become pectinate by the splitting apart of veins Ri and R5, so that they arise separately from the supporting stem of the pectinate vein thus formed. The third diagram represents a radial sector in which vein R2 bears two accessory veins labeled R2a and R2b- The wings of Sisyra. — A figure of the wings of Sisyra flavicornis will be furnished the student for study. Label the veins of both fore and hind wings excepting the cross-veins and the marginal accessory veins. The radial sector in these wings is an example of the simplest type of a pectinate radial sector. A wing of Chauliodes. — A figure of a fore wing of a pupa of Chauliodes will be fiu*nished the student for study. Label the tracheae. Note that the forming cross-veins are not preceded by tracheae. In what important respect does the radial sector of this insect differ from that of Sisyra. A wing of Corydalus. — A figure of a front wing of Corydalus conmtus will be furnished the student for study. Label the veins. Compare the radial sector in the wings of Sisyra, Chauliodes, Corydalus, and Polystcechotes. WINGS OF EPHEMERIDA Figure 422 represents the venation of a fore wing of a May-fly; this figure is introduced here merely to facilitate the discussion of two features of the wings of the Ephemerida ; we will not enter upon the study of the homologies of the wing-veins of members of this order in this course. Concave and convex veins. — Examine the wings of a May-fly and note that the wings are fan-like in form, due to a very regular series of corruga- tions. Each wing-vein follows either the crest of a ridge or the bottom of a furrow. A vein that follows the crest of a ridge is termed a convex vein, and one that extends along the bottom of a furrow, a concave vein. In Figure 422 the convex veins are marked with a plus sign and the concave veins with a minus sign. Intercalary veins. — In the order Ephemerida the wings have been specialized by addition; but in this order the added longitudinal veins arise in a way that is very different from the manner in which the accessory veins 406 THE WISGS OF INSECTS of the Neuroptera are de\^eloped- In the Ephemerida the added longitudinal veins are de\'eloped in each case as a thickened line more or less nearly midway between two preexisting veins; for this reason they are termed intercalary veins. Fig. 422. — A fore wing of a May-fly. The convex veins are marked + ; the concave veins — . When it is desirable to refer to a particular intercalary vein it is done by connbining the initial I, indicating intercalary, with the designation of the area of the wing in which the intercalary vein occurs. For example, in the wings of most May-flies in which the venation is not reduced, there is an intercalar\' vein between veins Cui and Cu2, i. e. in the area Cui. This intercalary vein is designated as ICui. This and other intercalary veins are represented in Figure 422. IDENTIFICATION OF THE WIXG-VEIXS AND OF THE CELLS OF THE WINGS IN HYMENOPTERA The determination of the homologies of the wing-veins of Hymenoptera is very difficult, as even in the most generalized of the living members of the order the venation of the wings departs widely from the primitive type. In the Hymenoptera, as in the Diptera and in the Lepidoptera, the specialization of the wings is by a reduction in the ntmiber of the wing- veins, the more generalized forms possessing the maximum nimiber of wing- veins found in the order. In the more generalized families the reduction of the wing-venation is slight; in the more specialized families, it is extreme. The most characteristic method of modification of the wings of Hymen- optera is by the coalescence of veins from the margin of the wing towards the base of the wing. This results frequently in a branch of a longitudinal vein becoming transverse, so that it appears like a cross-vein ; and in some cases, where the coalescence has been carried still farther, a branch of a longitudinal vein has been so diverted from its primitive course that it THE WIXuS OF IXSECTS -107 extends towards the base of the wing. Both of these eonditions has been reached in the most generahzed of H\-ing Hxniienoptera. The existence of this method of speciaHzation in the H\niienoptera and the extent to which it has been cairied in this order were tirst recognized after an understanding of the methixis of modi lica lion of tlie wings of the Diptera had been attained. Fortunately among the Diptera there are to be found examples of all degTccs of this method of coalescence of veins; reference to some of these will be made later. In the M\-niono])tora. the hind wings are extremely nioditied, even in the Fii:^. 4J,V — Willies of Panil^hilius. The \rins ixw \v\.W\\\\. most generalized members of tlu' order; on this accoimt only fore wings of Hymeno])lera will be stiulird in this introductory lunirse. Among till' more generalized wings oi llymenoi)lera are those of (he genera Pamphilins and Macroxycla, (wo genera of saw-(lies. in eai"ii of these genera there is i)n'ser\-ed in (he fore wings all of the primidve wing- veins with a single excei)tioii; and as in each it is a dilTeren{ \ein (hat has been lost from that which is lacking in the other, b\- slndying wings of the two genera all of the wing-\'eins can be obser\cd. Figtu'e 423 represents (he wings of /'a;;//'///7//^s•,■ in (his ligm'c (he \'eins arc lettered. Figure 42.1 represents (he wings of a Macrowrla: in this figm-e the cells are lettered. In i\uiif>ltiliiis vein R- of the fore wings is lacking, biU (his x't-in is i)rt"st'nt in .'l/(j(:r('.v.vc/(i.- on (he otlu-r hand, in Macroxycla \ein (Hi- of (he fore wings is lacking, but this \ein is pri'Si'nl in Pa»if>liiliiis. 408 THE WINGS OF INSECTS From a study of the fore wings of these two genera a diagram of a typical hymenopterous wing can be made. Figures 425 and 426 represent such a diagram ; in the former the wing- veins are lettered, and in the latter, the cells of the wing. This diagram represents the venation of the fore wing of Pamphilius, except that vein R2, which is lacking in this genus, is added. The cell lettered S in Figure 426 is the pterostigma or stigma; it is cell Sc2, but is designated as the stigma, on account of its usually being opaque in many genera of this order. In the wings of these sa^vfiies the anal furrow and the median furrow are Fig. 424. — -Wings of Macroxyela. The cells are lettered. both well-marked; the anal furrow is immediately in front of the first anal vein, and the median furrow is in front of the media. The furrows are represented by dotted lines in the figures. In the anal area (i. e., that portion of the wing back of the anal furrow) the three typical veins are preserved; but they coalesce to a considerable extent, both at the base and near the margin of the wing. In the basal part of the pre-anal area {i. e., that portion of the wing in front of the anal furrow) the stems of the principal veins are as follows : the costa coincides with the costal margin of the wing (Fig. 425, C); the sub- costa (Sc) is well preserv^ed and is forked; back of the subcosta is a strong THE WINGS OF INSECTS 409 Stem formed by the coalescence of the other three veins; the cubitus (Cu) soon separates from the stem, extending in a curve towards the anal furrow; while the radius and the media coalesce for about half their length. In order to make these veins more distinct in the figure the free portion of the media is marked with cross lines. When we pass from the consideration of the main stems to a study of the branches, we meet a much more complicated problem, a problem which Fig. 425. — The veins of a typical hymenopterous wing; a fore wing of Pamphilius with vein R2 added. could not have been solved by a study of Hymenoptera alone. But a knowledge of the methods of specialization of the wings of Diptera gives a key to an understanding of the wings of H^Tnenoptera. We will study first the branches of the cubitus. Spread out before you your drawings of the wings of the following insects, and arrange them in the Fig. 426. -The cells of a typical hymenopterous wing; a fore wing of Pamphilius with vein R2 added. order named: a Bomb}-liid, a Scenopinid, and an Empidid. Now study the figure of a wing of Rhyphus (Fig. 410) and note that while in Rhyphns veins Cu2 and 2d A retain their primitive position, in the three wings named above these two veins exhibit varying degrees of coalescence. A similar method of specialization has taken place in the Hymenoptera, but in this order both branches of the cubitus coalesce with the first anal 410 THE WINGS OF INSECTS vein; and this coalescence has proceeded so far that both branches cross the anal furrow and end in the anal vein remote from the margin of the wing. (See Fig. 425). It should be noted that vein Cu2 is rarely preserved in this order, even in the more generalized forms. In Macroxyela (Fig. 424) the position of the fork of the cubitus is indicated by a bend in this vein. If the branches of the media be now examined, it will be seen that vein Ml (Fig. 425) extends longitudinally near the centre of the distal part of the wing, its primitive course being modified slightly if at all. Vein M2 follows a course similar to the course of this vein in the Bombyliid; so also does the medial cross-vein (Fig. 425, m). A comparison of the position of cells Mi, ist Mo, and 2d Mo in the Bombyliid and in the typical hymenopterous wing (Fig. 426) is very instructive. Returning to Pampkilius (Fig. 425), we see that vein M3 coalesces with the first anal vein, crossing the anal furrow near the margin of the wing. It is evident that the forces that are causing the branches of the cubitus to migrate along the first anal vein and towards the base of the wing are exerting a similar influence on this vein. It is also evident that vein M4 and Cui coalesce at the tip, and that the migration of the united tips of these veins (marked Cui in the figure) towards the base of the wing has so modi- fied the course of that part of vein M4 which is still free that this part of this vein extends towards the base of the wing. This change is very similar to the change in the course of vein Cuo in the Empidid. A curious result of this change in the direction of the course of vein M4 is that the cell M4 has been closed and pressed back to the centre of the wing (Fig. 426, M4), and now lies in front of the free portion of the vein M4 instead of behind it. Let us now consider the courses of the branches of the radius. Here again we can gain help from a study of dipterous wings. Observe in the Bombyliid (Pantarbes) the coalescence of the tips of veins R5 and Mi. In the Hymenoptera a similar coalescence of veins R5 and Mi has occurred; but it has proceeded much farther, so that the free portion of vein Rj in Pampkilius (Fig. 425, R5) is remote from the end of the wing and has the appearance of a cross-vein. In the Hymenoptera vein R5 has been followed in its migration along vein Ml by vein R4, which has now reached a stage in Pamphilins that is quite similar to that reached by vein R5 in Pajitarbes. But like vein R5 it has the appearance of a cross- vein. From this it will be seen that the vein marked Mi in Figure 425 is really compound, as it includes the tips of veins R5 and Ri. In Pamphilins vein Ri is curved away from the costal margin of the wing to make room for a stigma (Fig. 426, S), and vein R3 ends in the costal margin a short distance before the apex of the wing (Fig. 425). Vein R2 THE WINGS OF INSECTS 411 has been lost in this genus, but is well-preserved in Macroxyela and is, there- fore, represented in the figure. While the tips of the branches of the radial sector have migrated away from the apex of the wing, the bases of these branches coalesce in the oppo- site direction ; from these two causes results the transverse bracing of the radial area of the wing, which is a very characteristic feature of the vena- tion of the wings in this order. The details of these changes will be made clear by an examination of Figures 427a and 427b. The former represents the ~Z!I^^^*^~~~-^ primitive mode of branch- p^m^m^^ — ■" "x'L^ ' ^\/?! ing of the radius; the latter, the radial area of the typical hymenopterous wing (Fig. 425). In the hymen- opterous type veins R2+3 and R4+5 of the primitive type coalesce so far that the branches of the sector arise from a common stem; and the tips of all of them have mo\'ed away from the apex of the wing, veins Ro and R3 following the costal margin of the wing; and veins R4 and Rs fol- lowing vein Ml. In the Hymenoptera the radial cross-vein is frequently preserved ; it is marked r in figure 425. In descriptions of wings of Diptera and of Lepidoptcra, in those cases where compound veins are fomied by the coalescence of adjacent \-eins, the compound vein is designated by a term that indicates its composition. Thus in Rhyphits (Fig. 410), the vein formed by the coalescence of veins R2 and R3 is designated as vein R2+.3; and in the wing of Eiilonchus (see your drawing of this wing) the tmited tips of veins Cui and Mg is designated as vein Cui + M3. When it is desired to indicate the composition of a compound xexn it can be readily done by the application of this system. But in descriptions of hymenopterous wings, where a compound vein may be formed by the coalescence of several veins the logical carrying out of this plan would result in a very cumbersome terminology, one that it is impracticable to use in ordinary descriptive work. In such cases the com])ound vein is designated by the term indicating its most obvious element. Thus, for example, in the fore wing of Pamphilius, where veins M4, Cui and Cuo coalesce with the first anal vein, the united tips of these veins is designated as vein ist A, Fig. 427.- -Diagrams of the radius: b, hymenopterous. a, typical; 412 THE WINGS OF INSECTS the first anal vein being its most obvious element (Fig. 425), although it is really vein M4 + Cui + Cu2 + ist A. The more important methods by which the primitive type of wing- venation has been modified in the Hymenoptera are the following : (a) A reduction in the number of the wing-veins by the atrophy of one or more veins. The loss of vein Ro in the fore wing of Pamphilins (Fig. 423) and of vein Cuo in the fore wing of Macroxyela (Fig. 424) are illustra- tions of this. (6) A reduction in the number of the wing-veins by the coalescence of adjacent veins in one or more areas of the wing. Examples of this will be indicated later. Such a reduction has nearly occurred in the fore wing of Pamphilius where the first and second anal veins coalesce for the greater part of their length. (c) A change in the course of a vein by the coalescence of its base with an adjacent vein. The course of media in the fore wing of Pamphilius (Fig. 423) has been modified somewhat by its coalescence with radius; this modification has been carried much farther in some foiTns to be studied later. (d) A change in the course of a A^ein by the coalescence of its tip with an adjacent vein. The changes in the direction of the branches of radius, media, and cubitus described above illustrate this. (e) The formation of serial veins. Examples of this will be presented later. The strident who has followed this discussion and has understood it will be prepared to make original investigations of the venation of the wings of Hymenoptera. But no student should take up the work indicated below before everything in this discussion is clear to him. Study the fore wings of the insects named below. It is not best to attempt to determine the homologies of the veins of the hind wings at first, owing to the great reduction of wing-veins that has taken place in these wings. The directions for the study of wings given on page 395 will apply here except that Figures 425 and 426 instead of Figure 410 will be used for comparison. Examples of comparatively generalized hymenopterous wings. — The fore wing of a saw-fly and the fore wing of a siricid will be used as examples of comj^aratively generalized hymenopterous wings. In neither of these wings is the venation as generalized as in Pamphilitis or in Macroxyela. A fore wing of Pteronidea ribesii. — A fore wing of the currant saw-fly, Pteronidea ribesii, is selected for the study of an actual hymenopterous wing; a mounted wing will be furnished for study; printed figures of other wings to be studied will be issued as needed. Make a drawing of the wing. THE WINGS OF INSECTS 413 Note the total atrophy of certain veins and the partial atrophy of others. Where veins have become very weak but are still visible this fact can be indicated by the use of dotted lines. Determine the homologies of the veins and of the cells of the wing by an application of the knowledge gained by the study of the typical hymenop- terous wing. A fore wing of a siricid. — A printed figure of a fore wing of a horn-tail of the genus Sirex will be furnished for study. Note that near the apex of the wing there is a short, compound vein formed by the coalescence of the tips of two veins, one of which extends back from the costal margin of the wing. The space between this com- pound vein and the costal margin of the wing is termed the appendicidate cell. It may be lettered ap in your figure. Letter each of the veins indicated by arrows in your figure and the tips of the anal veins. Letter the cells of the wing. An example of a change in the course of a vein by the coalescence of its base with an adjacent vein. — Note the course of media in the fore wing of Sirex. The coalescence of the base of this vein with radius has been carried so far that now media follows a Z-shaped course. Examples of the specialization of wings by the atrophy of veins. — Even in the most generalized liATnenopterous wings known one or two of the wing-veins have been lost. We will now study wings in which the atrophy of veins has proceeded farther than in those already studied. A jore wing of Janus ahhreviatus. — Letter the veins in your figure of the fore wing of this species. A jore wing of Odontatdacus editus. — In this wing the anal furrow and the axillary furrows are distinct, the position of each of these furrows is indi- cated by a dotted line. Compare j^our figure of this wing with that of Janus ahhreviatus and indicate by penciled, dotted lines the probable former position of each of the veins that are wanting. What is the probable composition of the single vein that is preser\^ed in the anal area ? Letter this anal vein with the term indicating its most obvious element. The switching of the base of the radial sector. — Compare the figures of the wings of Jauns and Odontaidaciis with that of the typical h}Tnenop- terous wing and describe the change that has taken place in the support of the base of the radial sector of Odontaulaciis . This change is known as the switching of the base of the radial sector; it has taken place in all of the Clistogastra, the more specialized of the two suborders of the H}Tnenoptera. The formation of serial veins. — In the wings of many Hymenoptera there exist what appear to be simple veins that are really compound veins 414 THE WINGS OF INSECTS composed of two or more veins or sections of veins joined end to end with no indication of the point of union; such veins are termed serial veins. There are two serial veins in the fore wing of Odontaulacus editus; one of these consists of a part of the radial sector and the radial cross-vein; the other, of another part of the radial sector and the radio-medial cross-vein. The former is designated as vein r 8c R^; the latter as vein r-m & R^. The sign & is used in these designations instead of + , as the latter is used to indicate compound veins formed by the coalescence of veins side b}^ side. Complete the lettering of the veins in the figure of the wing of Odontau- lacus editus. The reduction in the number of the wing-veins by the coalescence of adjacent veins. — This method of specialization by reduction is well-shown in many Hymenoptera of which the following is an example. The fore wing of an Ichneumon-fly. — ^A figure of a fore wing of an Ichneu- mon-fly, Exetastes fascipennis, will be provided for study. Compare your figures of the wings of Sirex, Odontaulacus and an Ichneumon-fly and deter- mine the composition of that part of the thickened costal margin of the wing of an Ichneumon-fly that extends from the base of the wing to the stigma ; letter this vein. Additional examples of specialization by reduction. — Continue the study of the fore wing of an Ichneumon-fly. The Ichneimion-flies belong to a series of families, the parasitic Hymen- optera, in which vein R5 is usually lost. In Odontaulacus editus, which also belongs to this series of families, there is a small vestige of this vein retained; but in the Ichneumon-flies it is lost completely. With this information in mind proceed to letter the veins in the figure of a fore wing of an Ichneumon- fly. A wing of a braconid. — A figure of a fore wing of a braconid, RJiogas parasiticus, will be provided for study. Letter the veins in this wing. Compare the figures of a wing of an Ichneumon-fly and a wing of a braconid and state in what features each wing is more generalized than the other. THE TRACHEATION OF THE WINGS OF NYMPHS AND OF PUPJE It has been found that, in the course of the development of the wings of the more generalized insects, the tracheas which traverse the principal veins are developed before the veins appear, and that later the veins are developed about these trachese. It is evident, therefore, that much light can be thrown upon questions regarding the homologies of wing-veins by studies of the tracheae which precede them; and the following suggestions are given to aid students who wish to make such studies. If a living pupa or nymph be placed in formol (4%) the tissues of the wings will be rendered translucent in a short time. In the case of very THE WINGS OF INSECTS 415 delicate insects only a few hours are required for this, but with larger ones with more opaque wings it is necessary to leave them in the formol for several days, or even for several weeks. While the formol renders the tissues translucent, it does not soon penetrate the tracheae, which are, therefore, left filled with air, and appear as dark lines when the wing is examined with transmitted light. Just after molting some wings are translucent, but there are few so clear that a short stay in formol will not make them clearer. In order to study wings prepared in this way, they are removed from the body and mounted in glycerine-jelly, care being taken to cool the mount quickly so that the jelly will not penetrate the tracheae. In this way most beautiful objects can be prepared, which will show the minutest ramifica- tions of the tracheae. Not only can the tracheae that precede the wing veins be studied in this manner, but, if the wings be taken at the right stage, the forming veins will appear as pale bands when viewed by transmitted light. This is due to the fact that at this time the veins are merely cavities, filled with hTtiph, and are more translucent than the spaces between them, which are occupied by tissue. Unfortunately, however, this distinction is only temporary in most specimens. As a rule, the entire wing becomes transparent in a few hours after it is mounted in the glycerine-jelly. It is necessary, therefore, to make drawings or photo-micrographs promptly, in order to keep a record of the courses of the veins. On the other hand, the tracheae, as a rule, stand out more sharply twenty-four hours after mounting, because of the clearing effect of the glycerine- jelly upon the tissue of the wing. But the making of drawings or photo-micrographs of the tracheae should not be delayed long; for the trachete soon become filled with the jelly, and are then practically invisible. The preparation of specimens. — Collect living n\Tnphs or pupae, place them in fonnol (4*^^), and leave them for a time, as indicated above. The formol will make the wings of the insects more translucent ; but it will not remove dark colors from chitin. It is well, therefore, to select, at first, the paler species for study. When ready to mount a wing, spread a drop of melted glycerine-jelly on a slide and allow it to cool. Dissect off the wing to be studied, taking with it just enough of the thorax to include the basal attachments of the tracheae. The dissection may be made under water; but the wing should be removed from the water promptly, so that the tracheae may not become filled with water. Place the wing upon the solidified glycerine-jelly on the slide; and lower upon it a heated cover-glass, which will caitse the jelly to melt enough to envelop the wing. 416 THE WINGS OF INSECTS Cool the mount quickly on ice, a marble slab, or some other cold object. Rapid cooling is imperative, for in melted glycerine- jelly the tracheae soon become filled, and the smaller ones are then invisible. It is imperative, also, that the wings be handled with care. Being sac- like structures, the tracheae are almost free within them, and a slight pinch with forceps in the middle of the wing may throw all of its tracheae out of place. It is better to lift the wing by its thoracic attachments or upon a section lifter. Not every nymphal wing is fitted for this study. Just before molting, and especially just before the last molting, the wing becomes so crumpled within its old sheath that the course of its tracheas can be followed only with difficulty. The method of study. — So far as is practicable, the studies of the tracheation of the wings of nymphs and of pupag will be original investiga- tions. For this reason, no particular species is suggested for study. The student will select the most available material, and will endeavor to make an addition to our knowledge of this subject. Among the more available subjects for the beginner in this line of work, are the pupae of moths and the nymphs of Orthoptera. The former illus- trate specialization by reduction; the latter specialization by addition. During the winter, when it is difficult to collect Orthoptera, the nymphs of stone-flies (Plecoptera) may be used instead. These can be found under stones in the beds of streams. In many insects the costal trachea is wanting or is but slightly devel- oped. It does not follow, therefore, that the trachea nearest the costal margin of the wing is the costal trachea. In most cases, the radial trachea can be identified easily, and it will serve as a starting point for the determi- nation of the homologies of the other principal tracheae. In most orders of insects the longitudinal veins can be distinguished from the cross-veins by the fact that the cross-veins are not preceded by tracheae. In some of the many-veined insects, as Odonata, the cross-veins, as well as the longitudinal veins, are preceded by tracheas; there being, in these insects, a great multiplication of tracheas. On the other hand, in the Trichoptera, Diptera, and most Hymenoptera, a great reduction of the tracheal system has taken place. It is not well, therefore, for the student to begin his studies of this subject with members of either of these orders. In those orders where a specialization of wing-veins by addition has taken place, the accessory longitudinal veins are preceded by tracheas. Finally, it should be remembered that it is not safe to base conclusions upon the study of a single insect; a large series, representing as many genera and families as is practicable, should be investigated. BIBLIOGRAPHY The following list includes only the titles of the books and papers to which references have been made in the preceding pages. These works are chiefly those bearing on the development of the uniform terminology of the wing-veins of insects. No effort has been made to include those papers in which this terminology has been used without modification. Adolph, G. Ernst. ('79). "Ueber Insectenfiugel." Nova Acta der Ksl. Leop. -Carol. - Deutschen Akademie der Naturf. Vol. 41, pp. 215-291, with six plates. Albarda, Herman. ('91). "Revision des Rhaphidides." Tijdschr. v. Entom. Vol. 34 (1891), pp. 65-184, with eleven plates. Amans, p. C. ('85). "Comparisons des organes du vol dans la serie animalc." Ann. Sci. Nat. Ser. 6 Zool. 19 (1885), pp. 9-222, with eight plates. Brauer, F. ('68). "Verzeichniss der bis jetzt bekannten Neuropteren im Sinne Linne's." Verb. K. K. Zool-bot. Ges. Wien. 1868, pp. 359-416; 711-742. Brauer, F. ('85). "Systematische-zoologische Studien." Sitzb. der Kais. Akad. der Wissensch. Vol. 91, pp. 237-413. Bauer, F. and Redtenbacher. ('88). "Ein Beitrag zur Entwickelung des Flugelge- aders der Insecten." Zool. Anz. Vol. 11, 1888, pp. 444-447. Berlese, a. ('09). "Gli Insetti, loro organazzazione sviluppo, abitudini e rapporti coll 'uomo." vSocieta Editrice Libraria, Milano, 1906. Bettin, Cornelius. ('13). "An Interesting Feature in the Venation of Helicopsyche, the Molannidae, and the Leptoceridas." Ann. Ent. Soc. Am. Vol. 6, pp. 65-73, with eight text figures. Bradley, J. Chester. ('08). "The Evaniidas, ensign-flies, an archaic family of Hymenoptera." Trans. Am. Ent. Soc. Vol. 34, pp. 103-194, with eleven plates. Bradley, J. C. ('13). "The Siricidae of North America." Jour, of Entom. and Zool. Vol. 5, pp. 1-30, with five plates. Brodie, p. B. ('45). "A history of the fossil insects in the secondary rocks of Eng- land." London, 1845. Brongniart, Charles. ('93). "Recherches pour servir a I'histoire des Insectes Fossiles." Saint-Etienne, 1893. Burmeister, H. ('78). "Examen special des Escales." Description Physique de la Rcpublique Argentine 5me tome (Lepidopteres) ire parte, pp. 21-28. Buenos Ayres. BuscK, August. ('14). "On the Classification of the Microlepidoptera." Proc. Entom. Soc. Wash. Vol. 16 (1914), pp. 47-54. Calvert, Philip P. ('13). "The fossil odonate Phenacolestcs, with a discussion of the venation of the legion Podagrion." Proc. Acad, of Nat. Sci. Phil. 1913. pp. 225- 272, with one plate. Chabrier, J. (1820). "Essai sur le vol des insectes." Mem. du Mus. d' Hist. Nat. Vol. 6 (1820), pp. 410-472, with four plates. CoMSTOCK, J. H. ('92a). "Report of lecture before the California Zoological Club, Jan. 30, 1892." Zoe. Vol. 3, pp. 84-86. CoMSTOCK, J. H. ('92b). "The descent of the Lepidoptera; an application of the theory of natural selection to taxonomy." Proc. Amer. Assoc, Adv. Sci. Vol. 41, p. 199. COMSTOCK, J. H. ('93). "Evolution and Taxonomy; an essay on the application of the theory of natural selection in the classification of animals and plants, illustrated (417) 418 BIBLIOGRAPHY by a study of the evolution of the wings of insects and by a contribution to the classification of the Leipdoptera." In Wilder Quarter Century Book, 1893, pp. 37-1 13, with three plates. CoMSTOCK, J. H. ('95). "The Venation of the Wings of Insects." In "The Elements of Insect Anatomy" by John Henry Comstock and Vernon L. Kellogg. Ithaca, 1895, Chapter 7, pp. 75-91. Comstock, J. H. and A. B. ('95). "A Manual for the Study of Insects." Ithaca, N. Y., 1895. pp. X + 201, with six plates and 798 wdcts. Comstock, J. H. and Needham, J. G. ('98-'99). "The Wings of Insects." A series of articles on the structure and development of the wings of insects, with special reference to the taxonomic value of the characters presented by the wings. Re- printed from The American Naturalist, with the addition of a table of contents. 124 pages, 90 figures. Ithaca, N. Y., 1899. (The articles appeared originally in The American Naturalist, Vol. 32, (1898), pp. 43, 81, 231, 237, 240, 243, 249, 253, 256, 335- 413, 420, 423, 561, 769, 774, 903; Vol. 33, (1899), pp. 118, 573, 845, 851, 853, 858.) Comstock, J. H. ('01). "The wings of the Sesiidae." (In Monograph of the Sesiidae by Wm. Beutenmiiller. Memoirs Amer. Museum Nat. Hist. Vol. i, p. 220.) Cramfton, G. ('16). "The phylogenetic origin and the nature of the wings of insects according to the paranotal theory." Jour of the N. Y. Entomological Society. Vol. 24 (1916), pp. 1-39. Davis, K. C. ('03). "Sialididas of North and South America." N. Y. State Museum bulletin, 68., pp. 442-487. Desneux, J. ('04a). "A propos de la phylogenie des Termetides." Ann. Soc. Ent. Belg., Vol. 48, pp. 278-289. Desneux, J. ('04b). "Isoptera" Genera Insectorum. Fascicule 25. Dewitz, H. ('81). "Ueber die Flugelbildung bei Phryganiden und Lepidopteren." Berl. Ent. Zeit. Vol. 25. DiMMOCK, Geo. ('83). "The scales of Coleoptera." Psyche Vol. 4, pp. 3-1 1 ; 23-27; 43-47; 63-71. Enderlein, G. ('00). "Die Psocidenfauna Perus." Zool. Jahrb. Syst. 14, pp. 133- 160. pis. 8-9. Enderlein, G. ('02). "Eine einseitige Hemmung.sbildung bei Telea polyphemus vom ontogenetischen Standpunkt." Zool. Jahrb. Anat. XVI, pp. 571-614, pi. 40-42. Enderlein, G. ('03). "Die Copeognathen des Indo-Australischen Fauengebietes." Annales historico-naturales Musei Nationalis Hungarici. Bd. i, pp. 179-344, with twelve plates. Enderlein, G. ('05a). "Monographie der Coniopterygidae." Zool. Jahrb. Vol. 23. Abt. f. Syst. Enderlein, G. ('05b). "Die Plecopteren Feuerlands." Zool. Anz. Vol. 28, pp. 809- 815. Enderlein, G. ('05c). "Monographie der Conopterygidge." Zool. Jahrb. Jene. Abt. f. Syst. 23 (1806), pp. 173-242. Enderlein, G. ('09). "Klassification der Plecopteren, sowie Diagnosen neuer Gattungen und Arten." Zool. Anz. Vol. 34 (1909), pp. 385-394. Enderlein, G. ('10). "Klassifikation der Mantispiden nach dem Material des Stettiner Zoologischen Museums." Stett. Entomolo. Zeit. Vol. 71, (1910), pp. 341-379- Enderlein, G. ('12). "Embiidinen, Monographisch Bearbcitet." Collections Zoolo- giques, Selys Longchamps. Fasc. 3, 1912. Froggatt, Walter W. ('13). "White Ants." Farmers' Bulletin, No. 60 Depart- ment of Agriculture, New South Wales, Nov., 1913. BIBLIOGRAPHY 419 FuNKHOUSER, W. D. ('13). "Homologies of the Wing Veins of the Membracidae." Ann. Ent. Soc. Am. Vol. 6, pp. 74-97, with five plates. Ganin, M. ('76). "Materialien zur Kenntniss der postembryonalen Entwicklung- sgeschichte der Insecten." (Russian). Abdruck bei Hoger in Jahresber. der Anat. u. Phys. von Hoffman und Schwalbe. Vol. 5 und in Zeit. f. wiss. Zool. vol. 28. GoNiN, J. ('94). "Rcchcrches sur la metamorphose des Lepidopteres." Bulletin de la Societe Vaudoise des Sciences Naturelles. Vol. 31 (1894), pp. 89-139, with five plates. Grote, a. R. ('96a). "System der nordamerikanischen SchmetterHnge." Mitt, aus dem Roemer-Museum, Hildesheim No. 7, 1896. Grote, A. R. ('96b). "Die Satumiiden." Mitt, aus dem Roemer-Museum, Hilde- sheim, No. 6, 1886. Grote, A. R. ('97). "The British Day Butterflies, and the Changes in the Wings of Butterflies." Proc. South Lond. Entom. and Nat. Hist. Soc. (1897). Haase, Erich. ('91). "Zur Entwicklung der Flugelrippen der Schmetterlinge." Zool. Anz. Vol. 14 (1891), pp. 116-I17. Haase, E. ('93). "Untersuchungen liber die Mimicry." Stuttgart, 1893. Hagen, Dr. ('70). "Ueber rationelle Benennung des Geaders in den Fliigeln der Insekten." Stett. Ent. Zeitung, Vol. 31, pp. 316-320, with one plate. Hancock, J. L. ('02). "The Tettigidae of North America." Chicago, 1902. Pub- lished by special grant of Mrs. Frank G. Logan. Handlirsch, Anton ('06). "Revision of American Paleozoic Insects." Proc. U. S. Nat. Mus. Vol. 29, pp. 661-820. Handlirsch, Anton. ('o6-'o8). "Die Fossilen Insekten und die Phylogenie der Rezenten Formen." Leipzig 1906-1908, i vol. text, pp. IX + VI -f- 430; i vol. plates, pp. XL and 51, plates. Headlee, Thomas J. ('07). "A study in butterfly wing- venation, with special regard to the radial vein of the front wing." Smithsonian Miscellaneous Collections, Vol. 48, pp. 284-296, with five plates. Hinds, W. E. ('03). "Contribution to a monograph of the insects of the order Thy- sanoptera inhabiting North America." Proc. U. S. Nat. Museum, Vol. 26, pp. 79-242, with eleven plates. Holmgren, Nils. ('09). "Termitenstudien. i. Anatomische untersuchungen." Kungl. Svenska Vet.-Ak, Hand. Vol. 44, No. 3 (1909), pp. 1-215. Holmgren, Nils. ('ii). "Termitenstudien, 2. Systematik der Termiten." Kungl. Svenska Vet. — Ak, Handl. Vol. 46, No. 6, pp. 1-86, with six plates. Horvath, G. ('13). "Etude morphologique sur la construction de I'elytre des Cicadides." Trans. 2d. Int. Cong. Entom. Vol. 2, pp. 422-432. Jones, Paul R. ('12). "Some new California and Georgia Thysanoptera." Bull. U. S. Dept. of Agri. Bureau of Ent. Tech. Series No. 23, Part i, 1912. Jurine, L. (1820). "Observationes sur les ailes des Hymenopteres." Mem. Reale Accad. Sci. Torino, Vol. 24 (1820), pp. 177-214, with six plates. Kellogg, V. L. ('94). "The Taxonomic Value of the Scales of the Lepidoptera." Kans. Univ. Quar. Vol. 3, No. i, (1894). Kellogg, V. L. ('95a). "The Ephemeridae and Venation Nomenclature." Psyche, Vol. 7, pp. 1 17-126. Kellogg, V. L. ('95b). "The Affinities of The Lepidopterous Wing." The Amer. Nat. Aug. 1895, pp. 709-717. Kellogg, V. L. ('08). "American Insects." New York, Henry Holt and Company, 1908, pp. XIV + 694. Kempers, K. J. W. ('99-'o9). "Het Adersysteem der Kevervleugels." Tijdschrift- voor Entomologie, Vol. 41, verslag, p. 31; Vol. 42, pp. 180-208; Vol. 43, pp. 172- 420 BIBLIOGRAPHY 190; Vol. 44, pp. 13-38; Vol. 45, pp. 53-71; Vol. 51, pp. IX-XVI; Vol. 52, pp. 272-283. Klapalek, F. ('12). "Perlodidas monographische Revision." Coll. Selys Long- champs. Fasc. 4, pp. 1-66. Bruxelles. Kruger, Edgar. ('98). "Ueber die Entwicklung der Fliigel der Insekten mit beson- derer Beriicksichtigung der Deckfliigel der Kafer." Inang.-Diss. Gottingen, 1899; Biol. Ctrbl. V, 19, No. 23 u 24, pp. 797-783. KuHNE, Otto. ('15). "Der Tracheenverlauf im Flugel der Koleopterennymphe." Zeit fur wiss. Zool. Vol. 112, pp. 691-718, with two plates. Lameere, a. ('08). "La paleontologie et les metamorphoses des insectes." Ann. Soc. Ent. Belg. Vol. 52, pp. 127-149. Landois, H. ('71). "Beitrage zur Entwicklungsgeschichte der Schmetterlingsflugel in der Raupe and Puppe." Zeit. f. wiss. Zool. Vol. 21, pp. 305-315. Leonard, M. D. ('16). "The immature stages of two Hemiptera — Empoasca obtusa Walsh (Typhlocybidae) and Lopidea rohinice Uhler (Capsidse)." Ent. News, Vol. 2^, pp. 49-54, with two plates. LoEW, H. ('62). "Monographs of the Diptera of North America." Part I, Smith- sonian Misc. Coll. (1862). LoWNE, B. T. ('92). "The anatomy, physiology, morphology, and development of the blow-fly." 2 Vols., London, 1890-1895. McClendon, J. F. ('06). "Notes on the True Neuroptera." Ent. News, 1906, pp. 116-121. MacGillivray, Alexander Dyar. ('06). "A Study of the Wings of the Tenth- redinoidea, a Superfamily of Hymenoptera." Proc. U. S. Nat. Mus. Vol. 39, pp. 569-654, with twenty-four plates. MacGillivray, A. D. ('12). "The Pupal Wings of Hepialus Thule." Ann. Ent. Soc. Am. Vol. 5, pp. 239-245. M'Lachlan, R. ('77). "On Notiothaiinia Reedi, a remarkable new genus and species. of Neuroptera from Chili, pertaining to the family Panorpidae." Trans. Ent. Soc. Lond. 1877, p. 427. Marshall, Wm. S. ('15). "The development of the hairs upon the wings of Platy- phylax designatus Walk." Ann. Ent. Soc. Am. Vol. 8 (1915), pp. 153-160. Marey, M. ('69). "Reproduction mecanique du vol des Insectes." C. R. Ac. Sci. Vol. 68, pp. 667-669. Mayer, A. G. ('96). "The development of the wing scales and their pigment in butter- flies and moths." Bull. Mus. Comp. Zool. Vol. 29, pp. 209-236, with seven plates. Meinert, F. ('80). "Sur 1' homologie des Elytres des Coleoptcres." Entomologist Tidskrift, 1880, p. 168. Melander, a. L. ('02a). "Two new Embiidas." Biol. Bull. Vol. 3, pp. 16-26. Melander, a. L. ('02b). "Notes on the structure and development of Emhia texana." Biol. Bull. Vol. 4, pp. 99-118. Mercer, W. F. ('00). "The development of the wings in the Lepidoptera." Jour. N. Y. Ent. Soc. Vol. 8, pp. 1-20, with five plates. Metcalf, Z. p. ('13a). "The Wing Venation of the Jassidte." Ann. Ent. Soc. Am. Vol. 6, pp. 103-115, with eight plates. Metcalf, Z. P. ('13b). "The Wing Venation of the Fulgoridie." Ann. Ent. Soc. Am. Vol. 6, pp. 341-352, with six plates. Metcalf, Z. P. ('17). "The wing venation of the Cercopidas." Ann. Ent. Soc. Am. Vol. 10, pp. 27-34, with two plates. Meyrick, E. (12). "Lepidoptera Hetcrocera Fam. Micropterygidas." Genera Insectorum, Fascicule, 132. BIBLIOGRAPHY 421 MiYAKt. T. ('13). "Studies on the Mecoptera of Japan." Jour. College of Agricul- ture, Imperial University of Tokyo, V^ol. 4, pp. 265-400, with ten plates. Morgan, A. H. ('12). "Homologies in the Wing-Veins of May-Flies." Ann. Ent. Soc. Am. Vol. 5, pp. 89-106, with five plates. MouLTON, Dudley. ('71)- "Synopsis, Catalogue, and Bibliography of North Ameri- can Thysanoptera." Bull. U. S. Dept. of Agri. Bureau of Ent. Tech. Series No. 21, 1911. Nakahara, Wako. ('15). "On the Hemerobiinagof Japan." Annotationes Zoologic«e Japonenses, Vol. 9, 1915, pp. 11-48. Navas, Longings S. J. ('12). "Neuroptera Fam. Nemoptcridae." Genera Insectorum, Fas. 136. Needham, J. G. ('00). "Some general features of the metamorphosis of the Flag- Weavil, Mononychus vidpeculus Fabr." Biol. Bull., Vol. i, pp. 179-191. Needham, J. G. ('03). "A Geneologic Study of Dragon-fly Wing- Venation." Proc. U. S. National Museum, Vol. 26, pp. 703-764, with twenty-four plates. NuTTALL, G. H. F. and Shipley, A. E. ('01). "The Structure and Biology of Ano- pheles." Jour, of Hygiene, Vol. I, p. 475. Osten-S.\cken, R. ('69). "Monographs of the Diptera of North America." Smith- sonian Miscellaneous Collections 219, 1869. Pack.\rd, a. S. ('95). "On a rational nomenclature of the veins of insects, especially those of the Lepidoptera." Psyche, Vol. 7, 1895, pp. 235-241. Pancritius, p. ('84). "Beitrage zur Kenntniss der Fliigelentwickelung bei den Insecten." Inagural-Dissertation, Konigsberg (1884). Patch, Edith M. ('09). "Homologies of the Wing Veins of the Aphididae, PsyUidas, Aleurodidas, and Coccidae." Ann. Ent. Soc. Am. Vol. 2, pp. 101-129, with six plates. Pierce, W. Dwight. ('09). "A monographic revision of the twisted winged insects comprising the order Strepsiptera Kirby." Bulletin of the United States National Museum, No. 66. Powell, P. B. ('o4-'o5). "The development of wings of certain beetles and some studies of the origin of the wings of insects." Jour. N. Y. Ent. Soc. Vol. 12 (1904), pp. 237-243; Vol. 13, (1905), pp. 5-22, with seven plates. Redtenbacher, Josef. ('86). "Vergleichende Studien iiber das Fliigelgeader der Insecten." Ann. des k. k. nat. Hofmuseums, Bd. I, Heft 3, 1886, pp. 153-232, with twelve plates. Rohwer, S. a. and Gahan, A. B. ('16). "Hormismolog\' of the Hymenopterous wing." Proc. Ent. Soc. of Wash. Vol. 18, pp. 20-76. Ris, F. ('96). "Die schweizer-Arten, der Perlidengattung Dictyopteryx." Mt. Schweiz ent. Ges. Vol. 9, pp. 333-313. Rosen, Kurt von. ('13). "Die fossile termiten: cine kurze zusammemfassung der bis jetzt bekannten funde." Trans. 2nd. Intern. Congress of Entomology, 1913. Schaeffer, C. ('89). "Beitrage zur Histologic der Insekten. I. Die Bauchdriisen • des Raupen. II. Ueber Blutbildungsherde bei Insektenlarven." Zool. Jahrb. von Prof. Spengel. Abth. f. Anat. u. Ontogenie. Vol. 3, pp. 611-652. ScuDDER, S. H. ('80). "The Devonian Insects of New Brunswick." Anniv. Mem. Boston Soc. Nat. Hist. (1880), pp. 1-41, with one plate. ScuDDER, S. H. ('90). "The fossil insects of North America, with notes on some European species." New York, 1890, 2 v. I. The pretertiary insects. — II. The tertiary insects. Scudder, S. H. ('93). "The Songs of our Grasshoppers and Crickets." Ent. Soc. Ont. 23d Ann. Report (for 1892), pp. 62-78. 422 BIBLIOGRAPHY Sellards, E. H. ('04). "A Study of the Structure of Paleozoic Cockroaches, with Descriptions of New Forms from the Coal Measures." Am. Jour. Sci. 4th ser. Vol. 18, (1904), pp. 113-227. Sellards, E. H. (06, '07, '09). "Types of Permian Insects." Am. Jour. Sci. Vol. 22, p. 249-258; Vol. 23, pp. 345-355; Vol. 27, pp. 151-173- Semper, C. ('57). "Ueber die Bildung der Fliigel, Schuppen und Haare bei den Lepidopteren." Zeit. f. wiss. Zool. Vol. 8, pp. 326-339, with one plate. Sharp, D. ('99). "Insects." The Cambridge Natural History, Vol. 6, London, 1899. Shelford, V. E. ('13). "Noteworthy variations in the elytral tracheation of Cicindela (Coleop.)." Entomological News, Vol. 24 (1913) pp. 124-125. Shelford, V. E. ('15). "Elytral tracheation of the tiger beetles (CicindeHdae)." Trans. Am. Micro. Society, October, 191 5. SiLVESTRi, F. ('09). "Die Fauna Siidwest-Australiens, Isoptera." Bd. II, Lie- ferung 17, 1909. Ergebnisse der Hamburger Sudwest-australischen Forschungreise 1905 (Michaelsen und Hartmeyer). Smith, R. G. "Evolution of the Venation in the anal area of the wings of Insects." A thesis presented to the Faculty of the Graduate School of Cornell University, 19 14. (not published). Snodgrass, R. E. ('09). "The Thorax of Insects and the Articulation of the Wings." Proc. U. S. Nat. Mus. Vol. 36, pp. 511-595, with thirty plates. Snodgrass, R. E. ('loa). "The Thorax of the Hymenoptera." Proc. U. S. Nat. Mus. Vol. 39, pp. 37-91, with sixteen plates. Snodgrass, R. E. ('lob). "The Anatomy of the Honey Bee." U. S. Dept. of Agri. Bureau of Entomolog\^ Tech. Series, No. 18. Speyer, a. ('70). "Zur Genealogie der Schmetterlinge." Stettin. Entom. Zeit. 1870, pp. 202-223. Spuler, a. ('92). "Zur Phylogenie und Ontogenie des Fliigelgeaders der Schmetter- linge." Zeit. fur wissen. Zoologie, Vol. 53, pp. 597-649, with two plates. Spuler, A. ('95). "Beitrag zur Kentniss des feiner Baues und der Phylogenie der Flugelbedeckung der Schmetterlinge." Zool. Jahrb. Vol. 8, pp. 520-543, with one plate. Straus-Durckheim, H. (1828). "Considerationes generales surl 'anatomic comparee des animaux articules." Royal Institute of France. Strickland, H. E. ('40). "On the occurrence of a Fossil Dragon-fly in the Lias of Warwickshire." Mag. Nat. Hist. Vol. 4, N. S. (1840), pp. 301-303. SuLC, K. ('12). "Ueber Respiration, Tracheensystem und Schaumproduktion der Schaumcikadenlarven (Aphrophorinse-Homoptera)." Zeit. f. wiss. Zoologie, Vol. 99 (1912, pp. 147-188). Tillyard, R. J. ('14). "On some problems concerning the development of the wing- venation of Odonata." Proc. Linn. Soc. New South Wales Vol. 39, pp. 163-216. Tillyard, R. J. ('16). "Studies in AustraHan Neuroptera." "No. i. — The wing- venation of the Myrmeleonidce." Proc. Linn. Soc. New South Wales, Vol. 40 (1916) , pp. 734-751, with two plates. "No. 2— Descriptions of new genera and species of the Famihes Osmylidae, Myrmeleontidas, and Ascalaphidae." /. c. Vol. 41, pp. 41-70, with six plates. "No. 3 — The wing-venation of the Chrysopida?." /. c. Vol. 41, pp. 221-248, with two plates. Tillyard, R. J. ('17a). "Further researches upon the problems of llie radial and zygopterid sectors in the wings of Odonata, and upon the formation of bridges." Proc. Linn. Soc. New South Wales, Vol. 41 (1916), pp. 871-887. Tillyard, R. J. ('17b). "The wing-venation of Lepidoptera (Preliminary report.)" Proc. Linn. Soc. New South Wales, Vol. 42 (1917), pp. 167-174. BIBLIOGRAPHY 423 TiLLYARD, R. J. ('17c). "Mesozoic Insects of Queensland." "Xo. i. — Planipennia, Trichoptera, and the new Order Protomecoptera." Proc. Linn. Soc. New South Wales, Vol. 42 (191 7), pp. 175-200, with three plates. Tower, W. L. ('03a). "The Origin and Development of the Wings of Coleoptera." Zool. Jahrb. Abth. fur Anat. und Ontogcnie. Vol. 17, pp. 517-572, PI. 14-20. Tower, W. L. ('03b). "Colors and Color-patterns of Coleoptera." Decennial Pubs. of Univ. of Chicago, 1903, Vol. 10, pp. 33-70. Ulmer, George. ('07). "Trichoptera." Genera Insectorum, Fascicule 60. Verson, a. ('90a). "La formazione delle ali nella larva del Bombyx mori." Publ. R. Staz. Bacol. Padova. Verson, A. ('90b). "Der Schmetterlingsflugel und die sogen. Imaginalscheiben desselben." Zool. Anzeiger, Vol. 13. Voss, F. ('05). "Ueber den Thorax von Gryllus domesticus, mit besonderer Beruck- sechtigung des Fliigelgelenks und dessen Bewegung." Zeit. f. Wiss. Zool. Vol. 78, 1905- Weele, H. W. van der. ('08). "Ascalaphiden, Monographisch Bcarbeitet." Coll. Zool. Selys Longchamps, Bruxelles, 1908. Weele, H. W. van der. ('10). "Megaloptera Monographic Revision." Coll. Zool. du Baron Edm. de Selys Longchamps. Fasc. V, Bruxelles, 1910. Weismann, a. ('64). "Die nachembryonale Entwicklung der Musciden nach Beo- bachtungen an Mnsca vomitoria und Sarcophaga carnaria." Zeit. f. wiss. Zool. \'oL 14. Weismann, A. ('66). "Die IMctamorphosc von Corethra pLumicornis." Zeit. f. wiss. Zool. Vol. 16. Williston, S. W. ('08). "Manual of North American Diptera." Xew Haven, James T. Hathaway, 1908. Wood-Mason, J. ('83). "A contribution to our knowledge of the Embiidte, a family of Orthopterous Insects." Proc. Zool. Soc. Lond. 1883., pp. 628-634, ^^'ith one plate." WoODWORTH, C. W. ('06). "The Wing Veins of Insects." Univ. Cal. Pub. Agri. Exp. Station, Tech. Bull. Ent. Vol. i, 1906. INDEX Figures in bold-faced type refer to pages bearing illustrations. Acanthaclisis, 205 Acanthiidae, 292 Acanthinevania princips, 379 Accessory cells, 342 Accessory vein of Enderlein, 248 Accessory veins, 71, 99, 146, 402 Acridid nymph, 17, 128 Acroneiiria, 243 Acroschismus hiibhardi, 301 Actias liina, 33 Aculeatae, 324 Aculei, 323 Acutalis, 278 Adolph, G. Ernst, 4 Adoneta, 338 Adventitious veins, 76 Aelothrips nasturtii, 267 Aenigmatodes danielsi, 77 Agallia 4-punctata, 281, 282 Albarda, H., 173 Albardia fiinatu, 207 Aleurodes, 289 Aletirodidae, 289 Aleuropleryx, 213 Alula, 55 ambient vein, 81 Amblychila, 299 Anal area, 58, 144 Anal crossing, 237 Anal furrow, 58, 131, 340, 394 Anal loop, 239 Anal triangle, 239 Anal vein, the, 67 Anal veins, 237,390 Anastomosis of veins, 76, 342 Anax Junius, 13, 43, 44, 47, 113, 234 Androconia, 322, 323 Angles of wings, 54, 394 Anisoptcra, 24 Anisola virginiensis, 331 Annandalia, 179 Anosia plexippus, 337, 342, 344, 401 Antcnodal cross-veins, 235 Anterior arculus, 392 Anterior notal wing process, 55 Anterior tuberosity, 57 AnthercBa pernyi, 18 Anther cea roylei, 33 Anthony, Needham and, 47 Aphid, wings of an, 269 Aphidid:e, 285, 291 Aphis, 286 Apis mellifica, 23, 40, 362, 369, 375, 377 Aphrophora salicis, 28 Apochrysa crwsiis, 210, 211 Apochrysa maisumurce, 21 1 Apochrysida;, 210 Appcndiculate cell, 81 Apterygogenea, 52 Archasia belfragei, ijg Arculus, 20, 78, 236, 358, 292 Argynnis, 5 Articulation of the wings, 55 Ascalaphus italicus, 207 Aspidothorax, 94 Atrophy, Reduction by, 371 Atdacinus fusiger, 379 A uslrolestes, 47 Axillaris, 65, 263 Axillary cord, 54 Axillary excision, 60 Axillary furrow, 59, 394 Axillary membrane, 55 Axillary sclerites, 56 Axillary vein, 248 Balaga micans, 200 Banks, Mr. Nathan, 183 Basal anal area, 239 Basal anal cell, 247 Bathytaptus falcipennis, 97, 98, 103 Becquerelia Grehanli, 107 Berotha insolita, 186 Berothidae, 186 Betten, Ur. Cornelius, 310, 31 1 Bittacomorpha, 36, 37 Blastophaga, 52 Blattida;, 123 Bleaching wings, method of, 399 Bombyx mori, 58, 1 16, 332 Boyeria irene, 234, 235 Bnichvnemunis longipalpus, 201 Braco'nid, Wing of a, 373, 414 Bradley, J. C, 364. 371 Brauer, Fr., 52, 193 Bridge, ormation of the, 230 British system, 346 Brongniart, C, 185, 305 Bulte, 81 Busck, A., 323 Cabbage butterfly, development of wings of, 115 Caccecia, 341, 342 Caddice-fly, 22, 308 CcBtiis, 214 Calyptcres 55 Campteroneura reticulata, 97, 98, 102 Cantharis, 300 Capiiia, 252 Capniida?, 244, 251 Castnia, 343 Celithemis el isa, 112 Cells of a typical hymenopterous wing, 364 Ccrambycid pupa, 298 Ceresa, 278 (425) 426 Index Ceresa borealis, 275 Ceresa dicer os, 280 Chaitophorus populicola, 286, 287 Chalcopteryx rutilans, 80, 233 Chapman, R. N., 26, 27 Chapman, T. A., 313, 3I7 ChauUodes, 21, 22, 32, 73, 405 ChauUodes pectkornis, 149, 170 Cicada, 60, 270, 271, 272, 291 atheroma regalis, 336 Chermes abietis, 287, 288 Chermes pinifolicp, 287, 288 Chermesinas, 287 Chief branches of the wing-veins, 65 Chief cubito-anal cross-vein, 239 Chirotoneles, 40 Chirotonetes albomanicatus, 222, 223 Chloroperla, 251 Chloroperla cydippe, 255 Chrysopa nigricornis, 190, 191 Chrysopa plorabunda, 189 Chrysopa signata, 189 Chrysopidse, 189 Clavus, 292 Climaciella brunnea, 174 Clisiocawpa americana, 87, 116, 340 Clothing of the wings, 64 Clothoda nobilis, 262, 265 Coccidae, 54, 290 Cockroach, 73, 403 Coleoptera, 121, 297 Comparison of terminologies of the wing- veins of the Odonata, 228; Aphididae, 289; Lepidoptera, 345; Diptera, 358 Concave veins, 81, 405 Coniocomposa, 213 Coniopterygidse, 212 Conioplcryx, 213 Conocephalus, 128, 129 Conops, 61, 354 Convex veins, 81 , 405 Conwenlzia, 213 Cordulegaster diastalops, 232 Cordnlegastcr sayi, 227, 236, 240 Corium, 292 Corodentia, 121, 258 Corrugations of the wings, 57, 393 Corydalinae, 170 Corydaloides, 94 Corydalus cornutus, 12, 72, 153, 154, 157, 171, 403, 405 Corydalus primitivus, 155 Costa, 65, 390 Costal cross-veins, 78 Costal hinge, 59 Costal sclerite, 57 Costal trachea of Orthoptera, 1 24 Costo-radial group of trachea?, 17,216 Croce filipennis, 208, 209 Cross-veins, 76, 77, 79, 131, 235, 389, 391 Cryptoleon nebidosutn, 205 Cubital area, 239 Cubital fork, primary, 165; secondary, 165 Cubitalstamm, 263 Cubital supplement, 240 Cubito-anal excision, 81 Cubito-anal fold, 57 Cubito-anal group of trachea, 17, 216 Cubito-anal loop, 240 Cubito-anal sulcus, 57 Cubitus, 65, 390 Cuneus, 292 Dactvlopius 290 Damsel-fly, 224 Definitive accessory veins, 72, 146, 403 Dermaptera, 121, 295, 296 Desmocenis palliatus, 297 Desneux J., 135, 136 Development, of wing-veins, 12; of the wings of larvae, 115; of the wings of nymphs, no Diamphipnoa, 252 Diaphanoptera Munieri 94, 95 Diastatomma Hasina, 228, 229, 238, 242 Dictyoneura libelluloides, 100, 104 Dictyoneuridae, 93 Dictyopteryx, 248 Diedrocephala coccinea 281 Dilar americanus, 185 Dilar nohircc, 185 Dilar turcius, 185 Dilaridae, 184 Diptera, 22, 122, 347, 395 Directions for the study of wings, 395 Discal cell, 82 Discal vein, 82 Dixa, 352 Dolichopns coqitiUetti, 355, 356 Donaconethis abyssinica, 264 Doubleday, Edward, 5 Draeculacephala mollipes, 281 Dragon-fly, 224 Earwig, 295, 296 Elytra, 54 Eiiihia, 265 Embia sabulosa, 263, 264 Embia texana, 263 Embiidina, 121, 262 Embolium, 292 Enderlein, Dr. Gunther, 18, 33, 65, 157, 213, 244, 248, 260 Epeorns, trachea; of, 39, 40, 45; wings of, 74, 216, 217 Ephemerida, 121,214, 220, 405 Erax, 357 Eristalis, 357 Erythrothrips arizona, 268 Etibleptus danielsi, 89 Euclca cippits, 320 Enlonchus, 363, 366 Eurvlhomopteryx aniiqua, 19, 77, 98, 103, 106, 107 Eusthenia spectabilis, 247, 251 Evauiti appcndigaster, 379, 381 Evaniclliis, 381 Index 421 Evolution, of the costa, 92; subcosta, 92; radius 95; media, 99; cubitus, 104; anal veins, 107 Evolution and Taxonomy, 8 Exjianded humeral angle, 63 Faltentheil, 7 Fan-like wings, 53 Fibula, 61 ; oi Corydalus, 63; oi Mnemon- ica, 315 of RhyacophUa, 63, 312 First Anal Vein, 65 First radio-medial cross-vein, 167 Fitch, Dr Asa, 304 Fixed fan-like type, 53 Fixed hairs, 323 Flugelrippen, 4 Folding fan-like type, 53 Forbes, Dr W.T.'M.,3I5 Fossil insects, 85 Frenata;, 8, 326 Frenulum, 61 , 330; loss of, 331 Frenulum hook, 61, 330 Froggatt, Walter W., 136 Fungus gnats, 350 Funkhouser, W. D., 271, 274 Furrows of the wing, 58, 394 Gall-gnat, 351 Genesis of the Uniform Terminology, i Geologic Time and Formations, Chart of, 84 German system of terminology, 346 Glossonotus, 278 Ccera, •^13 Gomphus descriptiis, 24, 225, 226, 227, 237, 240 Gonin, J., 1 14 Grabcr, Ur. V., 2 Gradate veins, 166, 204 Gripopterygida;, 244, 250 Gripopleryx tesselln'a, 250 Gypona 8-lineata, 281, 282 Haase, Dr. Erich, 6 Hadentomum americanum , 94 Iladroneura bohemica, 101 Haftfeld, 323 Hagen, Dr. II. A., 2, 5, 224; Plate by, 3 Halteres, 54 Hampson, Sir G. F., 346 Hamuli, 60, 61 Handlirsch, A., 52, 88, 89, 185, 238, 293 Ilarmosics rcflcxnliis, 59, 294 Heinemann, H. von, 8 Helicons, 213 Heliocharis, 241 Heliria, 278 Hemelytra, 54, 292 Hemerobiid group of families, 145 Memerol )iida?, 1 80 Ilcnicrobins hiaiiidi, 158, 159, 164, 177, 181, 182 Hcpialida,>, 318 Hepiaius, 9, 82 Hepialus sylvinus 323 Hepiaius thide, 327, 328 Ileptagenia interpuuclata, 218 Herrich-SchafFer, 345 Hetaerina, 241 Heteroptera, 121, 292 Hexagenia, 219 Hippodamia ij-pii>ictala, 297 Holmgren, Xils, 135, 137, 140 Ilolognatha, 244 Ilomaloneura punctata, 104 Homoptera, 121, 269 Ilomothetus fossilis, 86 Honey-bee, 367 Humeral cross-vein, 78, 391 Humeral suture, 143 Humeral veins, 81 , 343 Hydromanicus, 312, 313 Hymenoptera, 122, 362, 406 Hypostigmatic space, 83 Hypothetical primitive type, 15, 16, 64, 388 Ilyptia, 380 Ichneumon-fly, 69, 373 Increase of the numljer of wing-veins, 70 Inocellia longicornis, 171, 172 Intercalary veins, 71, 74, 232, 405, 406 Interpolated sectors, 233 Interradial nexus, 206 I so genus sp., 245 Isoptera, 121, 132 Isolated front branch type of media, 100 Italian loop, 240 Itliotieftdva, 176, 182 Ithone fusca, 175, 176 IthonidcC, 175 Janus ahbreviatiis, 377, 378, 413 Janus cynosbati, 377 Jassida;, 280, 282, 291 Jones, P. R., 267 Jugatie, 8,325 Jugum, 61, 63, 327 Kellogg, V. L., 321- 32.3 Kempers, K. J. VV., 300 Klapalek, Fr., 246 Kruger, E., 299 Kuhne, O., 300 Labidarge dibapha, 82 Lamccre, A., 88 Landois, H., 4, 5, 320 Lanthus parvulus, 237 Leonard, M. D., 112 Lepidoptera, 2^, 122, 319, 398 Leptis, 352 Leptocerus, 313 Leplophlebia, 219 Leslcs, 43, 44, 45, 47 Lestes rcctangularis, 231, 2^-^2 Lelhocerus, t,^ 428 Index Leucotermes, 134, 135 Leucoternies fiavipes, 141, 142, 143 Leucotermes, unknown species of, 142 Limnophilidas, 38 Lithomantis carbonaria, 90 Loew, H., 359 Lontamyia, 187 Longitudinal veins, 389 Lopidea robinice, 111, 112 Lycocerciis Goldenbergi, 91 Lycomorpha constans, 320 Lydidae, 362 McClendon, J. F., 189 MacGillivray, Dr. A. D., 78, 327, 328, 364, 368, 370, 376, 377 McLachlan, R., 185, 305, 313 Macrocephus satyrus, 377 Macroxyela, 363^372, 408 Manoxyela, 374 Mantichora, 299 Mantispidse, 174 Marginal accessory veins, 72, 147, 403 Marginal dots or dashes, 167 Margins of wings, 54, 394 Marshall, Wm. S., 324 Mastolermes darwiniensis, 132, 133, 134, 135, 137, 143 Maver, A. G.,320 Mayer, Paul, 88 May-fly, 406 Mecoptera, 121,302 Media, 65, 390 Medial cross- vein, 78, 391 Median furrow, 59, 394 Median nexus, 206 Median plates, 56 Medio-cubital cross- vein, 78, 392 Medius, 65 Megalomus moestus, 164, 183 Melanoplus, 42, 49 Membracidae, 274, 291 Membranule, 239 Mercer, W. F., 114 Merope tuber, 302, 303, 304, 305 Mesonemura Maaki, 249 Metcalf, Z. P., 271, 280, 282, 283 Metrnpator pussillus, 96, 97, 101 Meyrick,E.,3i3 Micropterygidae, 313, 317 Micropterygina, 313 Micropteryx unimaculella, 323 Micrutalis, 278 Mixatermes lugauensis, 94 Mnemonica, 313, 314, 316 Monarch butterfly, 401 Monohammus, 36 Moth-like fly, 351 Moulton, D., 268 Musca domestica, 356 Musical organs, tracheation of, 131 Myiodactylida;, 193 Myiodactyliis osmyloides, 193 Myiodaclylus pubescens, 194 Myrmecia, 82 Myrmeleon, 197 Myrmeleonid group of families, 145 Myrmeleonid, Tracheation, 198, 199 Myrmeleonidas, 196 Nakahara, Waro, 185 Named cross-veins, 19 Needham, J. G., quoted, 9, 76, 225, 231, 235, 237, 242, 299, 304; on Myr- meleonid Venation, 203 Nemoptera sinuata, 208, 210 Nemopteridce, 208 Nemoura, 13, 20, 108, 244, 250, 252, 253 Nemouridce, 244, 249. 231 Neosticta canescens, 231 Neuroptera, 121, 145, 401 Neuroplynx appendictilatiis, 149, 151 Newman, E., 175 Nodal furrow, 59, 273, 394 Nodus, 230 Notiobiella, 179 Notiothauma Reedi, 302, 305, 306 Nymphes myrmeleonides, 195, 196 Nymphidse, 195 Oblique vein, 197, 230 Odonata, 22, 24, 121, 224, 228 Odontaulacus editus, 377, 378, 413 Oecanthus, 130 Ogcogasler tesselata, 206, 207 Oligoneuria, 214 Oligoloma saundersi, 262, 265 Oliverina extensa, 208, 209 Oncomelopia undata, 281 Origin of Wings, 87 Orthemis ferriiginea, 240 Orthoptera, 120, 123 ' Oryssus abietinus, 377 Osmylidaj, 192 Osmylus hvalinatus, "ji, 151, 152, 153, 154, 192 Osmylus tessellatus, 151 Osten-Sacken, R., 359 Outline of Laboratory Work 387 Packard, A. vS., 7 Packardia, 338 Pal pares aesrhnoides, 164, 204 Paleontological data, 85, 88 Pamphilius, 363, 372, 407, 409 Panorpa,2>Q2,iQi Panlarbes, 354,365 Paolia Gurleyi, 97, 102 Paoliavetusla, 97, 101, 102, 105, 107 Papilio Machaon, 6 Papilio polyxenes, 339 Pajnlionida^, 341 Paraxetios eberi, 301 Parnassiiis sminllieus, 321 Patch, E. M., 271 285, 289, 290 Pelopceus ccmentarius, ZTJ Pcntatomid;e, 293 Perissoncura, 31 1 Index 429 Perla, 249 Perlidse, 244 251 Petalura gigantea, 232 Phanosloma, 313 Philaenus lineatus, 28, 271 Phryganeina, 307 Phvlfldromia germanica, 31, 126 Pielus Labyrinthecus, 326, 327, 399 400 Pier is rapce, 1 14, 324, 333 Pieridae, 341 Planates, 205 Platephemera antiqiia, 86 Plat hem is lydia, 43 Plecoptera, 121, 243 PoliopteniiS, 96 Poll opt 01 us elegans, 108, 109 Polystaxhotes punclatus, 150, 187, 188, 402 Polystcechotida;, 187 Pontia protodice, 333 Postanalfield, 137 Posterior arculus, 392 Posterior lobe, 60 Posterior notal wing process, 55 Posterior tuberosity, 57 Postnodal cross-veins, 236 Powell, P. B., 87, 299 Preanal area, 58 Preparation of specimens, 415 Principal wing-veins, 64 Prionoxystus robinicc, 75, 335, 400 Protentomon, 88 Prolohcrmes davidi, 155, 156 Protoplasafilchii, 349, 395 Protosialis, 168 Psectra, 179 Pseudofouquea cambrensis, 99, 103, 106 Pseudo-cubitus, 190 Pseudo-halteres, 54 Pseudo-media, 190 Psilopodius siplio, 355, 356 Psocus, 13, 14,258,259 Psychopsid, 189 Psychopsida;, 188 Psychopsis, 188 Psylla floccosa 284, 285 Psyllida;, 283,291 Pteronarcella badia, 252 Pteronarcys, 32, 45, 246 Pteronarcys dorsata, 243, 256 Pteronarcidee 244 252 Pteronidea ribesii, 383, 412 Pterostigma, 81, 247, 408 Pterygogenea, 52 Piter, 20b Pupal tracheoles of the wing, 1 16 Quadrangle of the Zygoptera, 240 Radial cross- vein, 78, 391 Radial cuneate area, 162 Radial planate, 206 Radial sector, 66, 390; in butterflies, 343; division of the stem of, 180; numbering of the branches of, 156; pectinate type of, 404; suppression of the dichotomy of, 147, 148; suppression of the stem of, 180; switching of the base of, 376, 413; two types of, 95 Radio-medial cross-vein, 78, 391 Radius, 65, 390; Diagrams of several types of, 148, 404; Diagrams of the typical and of the hymenopterous, 367, 411 Raphidia adnixa, 172, 173 Raphidiidaj, 171 Rapisma viridipennis, 176, 177 Reaumur, 320 Recurrent vein, 166 Redtenbachcr, Josef, 2, 4, 5, 6, 11, 65, 85 Reduction of the number of wing- veins, 67 Re xavius japonic us, 1 86 Rhamphomyia, 355, 366 Rhogas parasiticus, 414 Rhyacophila, 37, 38, 311 Rhyacophila fiiscula, 308, 309, 311 Rh\7ichocephalus, 347 ;?/rv/>/7i<5, 68, 79, 347, 388 Ris', F., 248 Rosen, Kurt von, 1 35 Sabatinca, 314 Saltatorial Orthoptera, 127 Samia cecropaea, 33 Saussure, H. F. de, 5 Scales, 319 Scenopinus, 354 Schaffer, 320 Schizoneiira americana, 287 Schizoneiira rileyi, 285, 286 Scott, G. G., 43 Scudderia, 129, 131 Second Anal Vein, 65 Secondary anal vein, 238 Secondary longitudinal veins, 71 Sectoral cross- vein, 78, 391 Selys-Longchamps, Baron de, 5, 224 Semidalis aleurodiformis, 212, 213 Sertiaeodogaster barticensis, 380 Semper, Carl, 4, 320 Serial veins, 69,376, 413 Sharp, David, 2, 313, 346 Shelford, V.E.,299 Stalls, 13 Stalls infumata, 168, 169 Sialida;, 168 Sialina;, 168 Silvestri, F., 135, 136 Siphlurtis, 220 Sir ex, 383, 413 Siricid, 413 Sisyra, 405 Sisyraflavicornis, 149, 158, 177, 178 Sisyridas, 177 Sin ilia camel us, 279 Smith, Miss Lucv W., 244, 256 Smith, Mr. R.C., 189 Spangbergiella, 283 430 Index Spaniodera ambulans, 91 Specialization by addition, 70 Specialization by reduction, 67 Spermophorella, 187 Sphecius speciosus, 362 Spreitentheil, 7 Spuler, Dr. Arnold, 6, 7 Spurious vein, 358 Squamae, 55 Stacheln, 323 Staudinger and Schatz, 345 Stenobiella, 187 Stenodictya, 96 Stenodictya lobata, 18, 90, 93, 100, 108 Stenophylax, 39 Stenopsocus, 260 Steps in the Specialization of Wings, 118 StJmiopis, 328, 329 Stigma, 81, 229, 408 Stratiomyia, 358 Stratiomys, 359 Strepsiptera, 54, 121, 301 Strickland, H.E., 238 Stygne Rcemeri, 98, 99, 104, 105, 106, 107 Subcosta, 65, 39; three types of the, 93 Subcostal area, 79 Subcostal fold, 57 Subnodus, 230 Subquadrangle of the Zygoptera, 241 Subtriangle, 239 Sulc, Karel, 28, 271 Supertriangle, 237 Supplements, 234, 235 Supplemental anal loop, 240 Symmathetes contrarins, 202 Sympherobiidas, 178 Sympherohius amicidus, 179 Synthemis, 235 Systellognatha, 244 Tahanus, 68, 359 Teaching of the Uniform Terminology, 382 Tegmina, 54 Tegula, 54 Telamonanthe pulchella, 279 Telmona, 278 Teratembia genicidata, 265 Terminology of the cells of the wing, 79 Termopsis, 135, 143 Termopsis angiisticollis, 132, 138, 139, 140, 141 Terrestrial Trichoptera, 313 Thelia bimaculata, 270, 275, 276, 277, 278 Therva, 353 Third Anal Vein, 65 Thoracic supports of the wings, 55 Thryidopteryx ephemerceformis, 62, 330 Thysanojjtcra, 121, 267 Tillyard, R. J., 43, 47, 48, 49, 83, 157, 175, 182, 189, 225, 231, 237, 240 Tiphia, 364 Tipula, 357 Tomalares clavicornis, 200 Tower, W. L., 87, 299, 322 Trachea, transverse basal, 216 Tracheae, basal connections of, 25, 27; typical condition, 29; in a cockroach, 31; Pleronarcys, 32; Chaidiodes, 32 Anthercea, 33; notonectid nymph, 34 corisid nymph, 34; Lethocerus, 35 Monohamus, 36; Bittacomorpha, 37 Rhyacophila, 38; Limnophilidae, 38 Stenophylax, 39; Epeorus, 40; Chirolo- netes, 40; Apis, 41; Melanoplus, 42; Anax, 44; Lestes, 44; Blattidas, 124; Saltatorial Orthoptera, 127; Membra- cid£e, 274 Tracheae, Limitations to the value of, 24 Trach cation of the Wings, 12, 19, 414; eccentric, 23; increased, 21; simpler type of, 21 ; variations in, 20 Tracheen, geknauelten, 4 Tracheoles, distinction between tracheag and, 113; larval tracheoles of the wing, 116 Transverse basal trachea, 17 Transverse cord, 82, 247 Tremex, 23, 369 Triangle of the Anisoptera, 236 Trichoma, 187 Trichoptera, 22, 122, 307 Trigonal fork, 206 Trigonal planate, 206 Truss cell, 203 Tuberosities of the base of the wing, 57 Ultdodes hyalina, 150, 206 Vanduzea, 278 Vanduzea arquata, 278, 279 Vanessa, 4 Veins of the anal area, 66 Veins of a typical hymenopterous wing, 364 Veins Cu2 and the second anal vein, coalescence of, 355 Veins M3 and Cui, coalescence of, 353 Vein M4, loss of, 348 Vena media, 65 Verson, A., 87 Walsh, B.D., 224 Weele, Dr. H. W. van der, 155 Westwood, J. O., 239, 304 Williston, S. W., 359 Wing-buds, 114 Wing process of the pleurum, 55 Wings, different types of, 53 ; fundamental structure of, 53 ; presence or absence of, 52 Wood-Mason, J., 262 Xantholobus trilineatus, 279 Xenoneura antiqiiorum, 85 Xiphiditim, 127 Xyelidae, 362 Zygoptera, 24 j!!lllll!i!!!|!l!!S lit iMlMll J! I iilli I Hi! ii ! ! mm i ill il ill ■ 1 ! !i ■ ! Ill IPIIiJ! I'il iili I !l!j il III III III I ■■I ililliilli i i I ' II! Ill ! ( ! II !'l 1 i i i ' 'I i II I , II I I I I llllllili i III ili!ii i < I IIH 1 1 1 1 II I 111 iij i Iilli , ill III il ill!