THE LIBRARY 
 
 OF 
 THE UNIVERSITY 
 
 OF CALIFORNIA 
 
 PRESENTED BY 
 
 PROF. CHARLES A. KOFOID AND 
 MRS. PRUDENCE W. KOFOID 
 
/ 
 
 THE 
 
 ANATOMY & PHYSIOLOGY 
 
 OF THE 
 
 LOW-FLY 
 
 (Mmca vomitoriaLinnJ 
 
 A MONOGRAPH 
 
 BY 
 
 BENJAMIN THOMPSON LOWNE, M.R.C.S. Eng. 
 
 ILLUSTRATED WITH TEN PLATES. 
 
 JOHN VAN VOORST, Paternoster Row. 
 LONDON. 
 
 MDCCCLXX. 
 
THOMSON & PINDBB, PR1KTBR8, Bolt Court, Fleet Street, London. 
 
L ip 
 
 CONTENTS. 
 
 PREFACE v 
 
 INXRODUCTORY NOTE ox HOMOLOGIES OF INSECTS.. vii 
 
 PART l.-GENERAL OBSERVATIONS ON THE ANATOMY AND 
 PHYSIOLOGY OF THE FLY. 
 
 Section I. Introductory Remarks 1 
 
 SectionIL Development , 6 
 
 Section III. The Integument 9 
 
 Section IV. The Xervous System 13 
 
 Section \. The Wings and Legs , 18 
 
 Section \LT/ie Digestive System 22 
 
 Section VII. The Circulation 23 
 
 Section VIII. The Respiratory System 25 
 
 Section IX. The Fat bodies and Ductless Glands. ... 29 
 
 Section X. The Organs of Special Sense 30 
 
 SeccionXI. The Generative Organs 34 
 
 Section XII. Concluding Remarks 35 
 
IV CONTENTS. 
 
 PART II. DETAILS OF THE ANATOMY OF THE FLY. 
 
 Section I. The Integument of the Head 37 
 
 Section II. The Proboscis 41 
 
 Section III. The Salivary Glands and Ducts 51 
 
 Section IV. The Alimentary Canal and its Appendages. 53 
 
 Section^. The Rectal Papilloc 57 
 
 Section VI. The Integumen t of the Thorax 61 
 
 Section VII. The Thoracic Appendages 65 
 
 Section VIII. The Abdominal Segments 70 
 
 Section IX. The Respiratory Organs 71 
 
 Sectien X. The Dorsal Vessel ... 75 
 
 Section XL The Nervous System 77 
 
 Section XII. The Compound Eyes 84 
 
 Section XIII. The Ocelli 89 
 
 Section ~X.IV. The Antennce 91 
 
 Section XV. The Maxillary Palpi 93 
 
 Section XVI. The Frontal Sac 94 
 
 Section XVII. The Cephalo-Sternum 95 
 
 Section XVIII. The Halteres and Wing Organs 96 
 
 Section XIX. The Folliculate or Ductless Glands 100 
 
 Section XX. The Male Generative Organs 101 
 
 Section XXI. The Female Generative Organs 107 
 
 Section XXII. The Development of the Ovum 113 
 
 Section XXIII. The Formation of the Pupa 116 
 
PREFACE. 
 
 In submiting the following pages to the 
 scientific world, although I do so knowing their 
 many shortcomings, I trust there is sufficient 
 that is new to make amends for any slight 
 inaccuracies, which could, perhaps, hardly have 
 been avoided, since so little has been done to 
 pave the way for the present investigation. 
 
 The Plates, which form no inconsiderable 
 feature in this Work, have been carefully executed 
 on copper, by myself; andj have spared no 
 trouble to make them as perfect as possible. 
 
 I have endeavored to give prior authorities 
 for my statements as often as possible, except 
 when the facts are well known, but much 
 work upon the anatomy of insects is so scattered 
 that it has often not been in my power to discern 
 whether I am entitled to priority of discovery. 
 
1 have divided the text into two parts, in order 
 that the more general facts may be separated from 
 the more special ; and I have further had the 
 most technical portions of the subject printed in 
 smaller type. The correct appreciation of homo- 
 logies has been my aim throughout the entire 
 work, and the whole has been drawn up from my 
 own dissections. I have made no statements 
 merely upon the authority of others, except in 
 one or two instances, in which this is specially 
 stated. 
 
 Dr. Weismann's elaborate researches on the 
 Development of the Diptera have greatly aided me; 
 and my best thanks are due to Mr. C. Stewart, 
 of St. Thomas's Hospital, for the kind manner 
 in which he has allowed me access to his 
 cabinet, and the readiness with which he has 
 always aided me by his opinion or advice. 
 
 I have to acknowledge also the kindness of 
 Mr. Curtics for the aid he has afforded me in. 
 bringing out the present Work. 
 
INTRODUCTORY NOTE ON THE IIOMOLOGIES 
 OF INSECTS. 
 
 In the following pages, I am aware, that in describing the nn- 
 tennal and optic segments as posterior to tbose composing the 
 mouth, I have done some violence to the ordinarily received 
 views of many distinguished naturalists ; but I think the sequence 
 of the cephalic segments in the fly is too plain to be easily mis- 
 construed. 
 
 Professor Huxley,* states that he believes the upper surface 
 of the head in insects is sternal instead of dorsal, and he founds 
 this belief on the apparent curvature of the embryo ; but I think 
 it possible that the extraordinary development of the lateral parts 
 of segments behind the mouth, arching over and surrounding the 
 mouth segments, may give rise to the appearance 6f curvature. 
 
 My knowledge of the Crustacea is too limited to allow me to 
 give a positive opinion, as to which segment is most anterior in 
 them, but I see no reason why the optic and the antennal seg- 
 ments should not be dorsal plates of segments posterior to the 
 mouth, largely deA'elopecl and arching over it, whilst the corres- 
 ponding ventral plates are either wanting or amalgamated with 
 the ventral portions of other segments. Such a supposition ap- 
 pears to me to be as likely to be correct as that their dorsal plates 
 are absent, whilst the ventral plates occupy the dorsal region, and 
 unite by their anterior border (made posterior by their flexion) 
 with the anterior edge of a dorsal" plate belonging to another 
 segment. 
 
 As the names I have given to the various plates, forming the 
 head of the fly, depend upon their relations with each other, and 
 are mainly borne out by the distinctness of the four plates in 
 each segment, and the manner in which they articulate with each 
 
 The agamic reproduction of Aphis, Linn. Soc. trans, vol. xxii. 
 
"NOTE ON THE HOMOLOGIES OF INSECTS. Till 
 
 other ; as well as by the fact that the ganglia of the mouth seg- 
 ments (my oesophageal and oral ganglia) are anterior to the ganglia 
 of the eyes and antenna: and that the optic and antennal seg- 
 ments ore developed from iniaginal discs, contiguous with those 
 of the thorax. I give them for what they are worth and must 
 leave those, who know the homologies of the arthropods generally 
 better than I do, to judge whether the facts are sufficient to jus- 
 tify my conclusions. It is true Dr. Weismann has named certain 
 lobes, which he has found in the embryo larva, behind the an- 
 tennal and optic segments, maxillary and mandibular segments ; 
 but I do not think any connection can be traced between these 
 lobes and the organs of the mouth. I confess, however, I have 
 never been able to distinguish them. 
 
 With regard to the thorax if any substantial homology can be 
 traced between it and any of the segments of the Crustacea, I 
 should be inclined to look upon it as the homologue of the mid- 
 dle-body* of the latter from its late development in the fly, and 
 from its bearing the ambulatory legs. 
 
 I have stated, inadvertently in a note, on page 3, that " seven- 
 teen segments is the number assigned on theoretical grounds to 
 all true insects" but no sufficiently stable theory has, pehaps, 
 been formed to justify the expression : and I find, according to 
 Gerstacker, that the Orthopterahave eleven abdominalsegments, 
 forming an exception to the number I have given. 
 
 I find I agree with M. Duthiers, in describing the opening of 
 the sexual organs between the eighth and ninth abdominal seg- 
 ments, but did not see the great importance of the observation 
 at the time I made it. Lastly, strangely enongh, I have made use 
 of an expression in the second paragraph on page 7 which is iden- 
 tical with one used by Sir John Lubbockf in his paper on the 
 development of Ephemera, although I did not know at the time, 
 that it had been previously used. 
 
 * 1 use this term as it is used by F. Muller in his "Facts for Darwin " 
 translated by Dallas. 
 
 f Linn. Soc. trans, vol, xxv. page 489. 
 
PART I. 
 
 GENERAL OBSERVATIONS ON THE ANATOMY AND PHYSIOLOGY 
 OP THE FLY, 
 
 Section I. Introductory Remarks. 
 
 rPHE typo, or plan, upon which insects are organised is so 
 
 L very different from the type of the vertebrata, that if we 
 
 wish to understand the Anatomy of the Fly, it is necessary 
 
 to bear in mind the chief characteristics of the class to which 
 
 DESCRIPTION OF PLATE I. FBONTISPIECE. 
 
 This plate represents a dorsal view of the whole of the internal organs 
 as they appear in situ, when the integument of the back is removed, 
 except on the right side of the thorax, where the great dorsal muscles 
 have also been removed to show the deep layers. Commencing at the 
 head, the following organs are displayed : the frontal sac occupies the 
 frontal region and is tinted yellow : on each side of this sac are the 
 antennal nerves, and the sides of the head are occupied by the great eyes, 
 which are seen in section : between these the cephalic nerve centre with 
 its main tracheal trunks occupies the greater part of the head ; it is cut 
 through horizontally near the eyes to exhibit the arrangement of the deep 
 tracheae. The left side of the thorax is occupied by a mass of muscles, 
 longitudinal near the centre, and vertical at the side ; the central muscles 
 are more especially used to depress the wings in flight, and the lateral are 
 partly elevators of the wings and partly flexors and extensors of the legs 
 being inserted into their basal joints. The tracheal vessel, which occupies 
 the posterior portion of this side is connected with the posterior thoracic 
 spiracle. On the right side, the chyle stomach and convoluted salivary 
 gland occupy the portion nearest the mesial line. Immediately external 
 to the latter is the main lateral trachea collapsed, except its posterior 
 portion which remains filled with air ; it communicates with the posterior 
 thoracic spiracle. The nerve to the halter is seen crossing this portion of 
 the tracheal trunk. The anterior tracheal vessel exhibited on this side is 
 
2 ON THE ANATOMY OF THE FLY. 
 
 it belongs, and to investigate its relations to other insects, not 
 only in its perfect state, -but in its various stages of development. 
 
 Insects are articulate animals, organisms in which the ex- 
 ternal integuments fulfil the same purposes in the economy 
 as the osseous skeleton does in the vertebrata ; and, just as 
 the typical structure of the latter is most fully exemplified in 
 the skeleton, so the integumental system in the articulata affords 
 the most obvious indications of the type upon which they are 
 organised. 
 
 In the lowest members of this great division of the 
 Animal kingdom we find the body composed of a series of rings 
 or Annuli, united to each other by a membrane, each ring 
 having a separate nerve centre, and being similar to every 
 other ring in the body of the animal, except the first, 
 which is modified so as to form a distinct head very low in 
 the scale of life. 
 
 As we ascend, we find organs of sense developed upon the 
 anterior rings, however, still in a very rudimentary form; 
 an alimentary canal traverses the whole series, and a decided con- 
 centration of the nervous system takes place about its anterior 
 extremity. If we add to these characters the presence of a 
 mouth furnished with a pair of hooks, which not only serve the 
 purpose of teeth, but also assist the animal materially in locomo- 
 tion, and are in fact, one of three pair of modified limbs which 
 
 from the anterior spiracle; it passes forward into the head and com- 
 municates behind with the lateral trachea. Two small pulmonary sacs 
 are also seen communicating in front with this trunk. The muscles of th c 
 deep layer are probably all more, or less concerned in flight. The ab- 
 domen is occupied in the mesial line by the dorsal vessel, on each side of 
 which are the large abdominal pulmonary sacs; behind these the con. 
 volated bile tubes are colored red. The salirary glands may be seen 
 crossing the pulmonary sacs. The sides of the abdomen are filled with the 
 mature ovaries, and its apex is occupied by the rectum and rectal papillae. 
 
INTRODUCTORY REMARKS. 3 
 
 exist in the mouths of insects,* we shall have an idea of 
 the larva of the fly, commonly known as the gentle or maggot : 
 which, although it resembles one of the lowest of the articulata, 
 except in the nervous system, as will be seen afterwards possesses 
 the power of development into one of the highest of its forms. 
 
 The integumental system of this larva consists of seventeen 
 segments,! of which the three anterior alone have lost more or 
 less of their original annular form. These, both in the larva and 
 perfect insect, compose the mouth, and in both are provided with 
 three pair of modified limbs, although the existence of the first 
 pair is less certain in the larva, owing to the very rudimentary 
 
 - Each segment in the lowest articulata is normally furnished with 
 two pairs of lateral appendages or rudimentary limbs, one pair placed 
 above the other, the superior being dorsal and the inferior ventral ; at 
 least, such is their arrangement in annelides. Both pairs are 
 much modified in the higher forms, and are often entirely suppressed. 
 
 The segments themselves may be said to consist typically of four plates, 
 a ventral, a dorsal, and a lateral plate on each side ; the superior appen- 
 dages being placed between the lateral and dorsal, and the inferior 
 between the lateral and ventral plates. In insects the wings, when they 
 exist, represent the dorsal appendages, being placed between the superior 
 and lateral plates. It seems to me probable that the spiracles and their 
 tracheae are likewise lateral appendages. If such be the case, I think it 
 quite likely, from the number of lateral appendages, that more than 
 three segments really exist in the thorax, especially as the anterior 
 thoracic spiracles are placed between two segments and the four posterior 
 legs occupy a similar position; so that possibly these may be latera' 
 appendages of undeveloped segments. This is certainly the case if the 
 spiracles with their tracheae are homologues of lateral appendages, like the 
 wings and legs. This appears to me not unlikely from their position^ 
 and further from the consideration that the lateral appendages of many 
 annelides are both respiratory and motor organs, ( M. Quatrefages Hist 
 Nat. des Anneles Nouv, Suites a Buffon, Tom. i. p. 19). And we must 
 remember the tracheae of insects are merely modified processes of the skin 
 developed inwards instead of outwards. 
 
 t This is the typical number assigned on theoretical grounds to all 
 true insects, which always have the body in the perfect state divided into 
 three part*, the head, thorax and abdomen. The head consists of five 
 
 B 2 
 
4 ON THE ANATOMY OF THE FLY. 
 
 condition of this segment, and the extreme modification it under- 
 goes in the development of the perfect insect^ The fourth and 
 fifth segments are united above in the larva as well as in the 
 perfect form, and each bear a pair of sense organs ; the 
 antenna and the eyes. The sixth segment of the larva bears 
 the anterior spiracles or breathing pores ; but, together with the 
 next two, the seventh and eighth, and in fact all the succeeding 
 ones, is unprovided with legs. These three segments correspond 
 to the thorax of the perfect insect, which is however not only 
 provided with six legs, but also with a pair of wings, appendages 
 of the seventh segment. The eighth or last thoracic segment is 
 likewise usually supplied with wings in insects, but in the Dip- 
 tera, the class to which the fly belongs, they are suppressed and 
 their place is occupied by a pair of small organs, called halteres 
 or balancers. Of the remaining nine segments which form the 
 abdomen of the perfect insect, four are modified and converted 
 into the hard parts of the sexual organs, which are more or less 
 internal. 
 
 Although the segments of the larva and perfect insect cor- 
 respond exactly in the manner above indicated, nevertheless the 
 head and thorax of the fly, from the fourth to the eighth seg- 
 ment inclusive, are not developed from the corresponding larval 
 segments. 
 
 It may be here observed that the histological elements, that 
 is, the elementary tissues of the perfect fly, cells, fibres, &c., 
 
 segments, or according to some, of six, making eighteen in the whole body; 
 but as the segments of the head of the fly, both in the larva and perfect 
 insect are very distinct, I am inclined to the former view, The thorax 
 constantly bears three pairs of legs and consists of three distinct segments ; 
 unless, as may appear probable from the position of the legs and spiracles, 
 the thorax has six, when the whole number of segments in the insect 
 would amount to twenty, the number found in the highest scolopendrife 
 (Scutigera. Lam. ), where the head consists of five segments, and the 
 body of fifteen, which each bear a pair of legs, but are united above, so 
 that only eight appear on the dorsum, 
 
INTRODUCTORY REMARKS. 5 
 
 differ in no important character from those of the highest 
 animals. The chief difference is seen in the larger size of the 
 gland cells. The muscular fibres are described as larger 
 than in vertebrates. This has probably arisen from the larva 
 and not the perfect insect, having been examined. All 
 the tissues in the larva, except the nerve cells and fibres, 
 are four or five times as large as in the perfect insect, but this 
 seems to depend rather upon rapidity of growth than upon 
 any inherent difference. 
 
 The striated muscular fibre in the fly varies from 1 -3000th of an inch in 
 diameter in the muscles of the halteres and abdomen, to l-1000th in the 
 pharyngeal muscles and those of the legs, but in the larva several of the 
 muscles have fibres as large as 1 -200th of an inch in diameter. 
 
 All the muscular fibres of the alimentary canal and dorsal vessel 
 exhibit striae. The only non-striated muscles I have been able to detect 
 are the great longitudinal dorsal muscles of the thorax ; these consist of 
 long clear fibres less than 1 -5000th of an inch in diameter, swollen at 
 intervals and containing nuclei. Numerous slightly granular cells lie 
 amongst them. No myolemma exists, although the tracheal vessels appear 
 to separate them into bundles ; so that they are soft aud easily broken 
 up into a mass of cells and fibres. 
 
 Nuclei exist permanently in the myolemma of the striated muscles, and 
 appear in rows when they are treated with acetic acid. In many the fibres 
 branch, and the branches are as large as the primitive fibres. The muscles 
 of the abdomen and crop are flat bands of fibres with well marked striae ; 
 those of the crop terminate in pointed and often in branched extremities. 
 The muscles are either inserted into the integument directly or through the 
 medium of apod ernes, a clear transparent tendinous structure intervening^ 
 which appears to be continuous with the myolemma. The whole muscular 
 tissue is firm, but pale. 
 
 The gland cells of the salivary tubes are 1-1 000th of an inch in 
 diameter ; those of the rectal papillae are l-500th of an inch in diameter ; 
 the conical epithelium of the stomach has a long diameter of 1 -1500th 
 of an inch. The liver cells are polygonal, filled with pigment and oil 
 globules, with a diameter of about l-1500th of an inch : all have well 
 marked nuclei. Adipo?e tissue does not exist, but large quantities of fat 
 are laid up in the folliculate glands and liver cells, especially in the former. 
 
 Nerve cells vary from l-5000th to l-2500th of an inch in diameter ; 
 t hey are irregular in form, often being branched. The nerve fibres vary 
 very much in size ; those of the muscles branch, and end in loops ; others, 
 
6 ON THE ANATOMY OF THE FLY. 
 
 especially in the organs of special sense, end in nerve cells. The nerves 
 of the ganglia of the head and thorax are rounded, the fibres of which 
 they consist being enclosed in a neurilemma; but those of the sympathetic 
 system appear to have no neurilemma : they spread out into flattened 
 branched expansions, and have numerous branching nerve cells scattered 
 amongst them. 
 
 Section II. Development. 
 
 Perhaps the most startling difference between the vertebrate 
 and the articulate divisions of the animal kingdom is seen in 
 the manner of their development. In insects, as in many other 
 articulata, this takes place by a process termed Metamorphosis, 
 the creature not only being entirely and suddenly changed in ap- 
 pearance, but also in habits of life. Usually there are three 
 periods in the existence of an insect, the first characterised by rapid 
 growth and a comparatively small exercise of animal functions, 
 the second by a period of apparent quiescence, during which 
 growth is changed into development, and lastly one in which 
 the animal functions are seen in their highest perfection ; these 
 are called respectively the larval, pupal, and imaginal states. 
 
 In the larval state, where growth is rapid, the hard in- 
 tegument is usually shed periodically, its place being supplied by 
 new layers deposited beneath it, just as the cuticle in vertebrates is 
 renewed by younger epithelial cells, or as the old bark of trees 
 is replaced by younger layers. The larva of the fly forms an ex- 
 ception and does not shed its skin. After the full growth is 
 attained, the insect commonly assumes a different form, called a 
 pupa or chrysalis when it becomes inactive, and a nymph when a 
 greater or less amount of activity remains. In a certain number 
 of days the pupa case bursts, or the nymph sheds its skin, and the 
 perfect insect emerges, partially developed at first, but in a few 
 minutes or hours, as the case may be, assuming its new functions 
 
DEVELOPMENT. 7 
 
 and form in their greatest perfection. Growth and development 
 now cease as far as the individual is concerned, to give place to 
 the generative function. 
 
 Every degree of metamorphosis exists amongst insects, from 
 that in which the larva, nymph, and imago, closely resemble each 
 other, where the successive changes are merely those of ordinary 
 development, as in the cockroach, to that in which the change 
 is so complete that it might almost be doubted whether the 
 larva and imago should be considered the same individual at 
 all, so closely does the process resemble an alternation of gene- 
 ration. This is the case in the fly. 
 
 Dr. Weismann, in Germany, as long ago as 1865* pointed out 
 that the development of the fly differs remarkably from that of 
 most insects, in an elaborate memoir upon the subject ; and two 
 years after, in another upon the development of a species of gnat, 
 Corethra plumicornis, proposed the division of insects into two 
 classes, according to the presence or absence of certain structures 
 in the larva which he calls Imaginal discs, f This distinguished 
 naturalist asserted in his former paper, that the head and thorax 
 of the fly does not depend for its development upon the corres- 
 ponding larval segments, but that these parts are developed 
 from a series of discs (Imagined discs) firmly adherent to the 
 nerves and tracheae of the anterior extremity of the larva ; and 
 he goes on to add, " I believe in all those insects in which the 
 anterior larval segments are unprovided with appendages (legs) 
 the head and thorax of the Imago are entirely redeveloped, 
 whilst in those in which the larva is furnished with legs, these 
 parts depend for their formation upon the anterior larval 
 segments." 
 
 Startling as the assertions of that writer appear, I have been 
 fortunate enough to confirm them myself, and not only to confirm 
 
 * K'jlliker aud Siebold Zeitschrift, band xiv-xvi. f Imaginal Scheibcn, 
 
8 ON THE ANATOMY OF THE FLY. 
 
 them, but I think to add to them as regards the development of the 
 proboscis. I believe that those segments of the fly immediately 
 connected with the functions of vegetative life, those surrounding 
 the anterior portion of the alimentary canal, that is, the 
 proboscis, as well as the abdominal segments, are immediately 
 dependent upon the corresponding larval segments for their 
 development and form ; whilst the head and thorax with all their 
 appendages, the chief centres of animal life, are developed from 
 Weismann's Imaginal discs, which do not coalesce with each 
 other until the third day of the pupa state, so that these re- 
 place the larval segments from the fourth to the eighth inclusive. 
 
 The brain seems to form a centre around which these changes 
 take place ; even this organ is entirely altered, but by a process 
 similar to ordinary development. The other structures are formed 
 at the expense of a granular fluid which is chiefly derived from 
 the so-called fatty bodies ; these extend along the entire length 
 of the larva on either side of the alimentary canal. 
 
 The skin of the larva of the fly is not shed when it 
 reaches maturity, but dies and becomes very hard, turning 
 yellow at the first, then red, and afterwards almost black. 
 It forms the pupa case. This is not the representative of 
 the skin of the pupa in most insects, for the corresponding 
 larval skin is usually shed, and the pupa case is either altoge- 
 ther wanting, as in many hawk-moths and butterflies, or it is 
 replaced by a cocoon of silk spun by the larva, as in the silk 
 moth, or by some other adventitious covering. 
 
 The respiratory function ceases, and all the organs proper to 
 the maggot undergo degeneration soon after the formation of 
 the pupa case. The nerve centres grow rapidly, and the Imaginal 
 discs unfold into delicate cellular expansions, which coalesce 
 with each other, and with the new layers of cells formed within 
 the three anterior and nine posterior larval segments, and ul- 
 timately form an exceedingly delicate membrane corresponding 
 to the pupa skin of butterflies. In both, this marks out the position 
 
THE INTEGUMENT. 
 
 of the new organs about to be developed, and in both, this is shed 
 when the insect emerges from the pupa. These changes complete 
 what Dr. Weismann calls the first stage of the development of 
 the fly the formation of the pupa. 
 
 It may be confidently asserted that not one structure exists 
 in the fly as it exists in the maggot. Every portion of the larva, 
 except the brain, and perhaps the basement membrane of the 
 alimentary canal, undergoes rapid degeneration, and the fly is 
 formed within the pupa skin by a process of redevelopment. 
 
 Section III. The Integument. 
 
 The integumental system of the perfect fly originates upon the 
 inner surface of the pupa skin, but as its parts become fully de- 
 veloped they shrink away from the pupa skin, so that it forms a 
 loose investing membrane around the new formed fly, and may be 
 looked upon as one of the integuments shed in the process of 
 development; for, although it is never thickened and indurated 
 in the fly, it is precisely analogous to the pupa case in the 
 lepidoptera, which is always more or less hardened. The reason 
 of its not being hardened in the fly seems to be due to the fact 
 that it is invested in the pupa shell or last larval integument. 
 
 The integuments of insects are usually said to .consist of three 
 layers, and these may be easily traced in the fly. The outer- 
 most is transparent and continuous over the whole surface of the 
 insect, investing all the appendages and processes of the skin, 
 even the hairs, and covering the surface of the eyes. It appears 
 to be continuous with the lining membrane of the tracheal system, 
 and to extend throughout the digestive cavity, although it is 
 somewhat modified in the latter. 
 
10 
 
 ON THE ANATOMY OF THE FLY. 
 
 This layer is usually called tlie cuticle or epidermis, but the 
 term is unfortunate as it is quite unlike the cuticle in vertebrates, 
 or, indeed, any structure found in them ; it is persistent and not 
 deciduous. The cells from which it is formed appear early in 
 the development of the insect and coalesce into an apparently 
 structureless membrane of extreme toughness, which has the 
 remarkable property of being quite insoluble in a hot solution 
 of caustic potash; I shall, therefore, substitute the term proto- 
 derm for this layer. 
 
 The continuity of the protoderm is a fact well worth con- 
 sideration at the commencement of the study of the anatomy of 
 insects, as the terms insect, articulate, and articulate animal, are 
 extremely liable to mislead, and to induce one to believe that the 
 various segments are distinct and separable. Such is not really 
 the case, but certain portions of the protoderm are thickened by 
 the induration of the epithelial cells beneath it, which become 
 inseparably united to it and to each other, and so produce the 
 hard parts of the integument. 
 
 The two layers beneath the protoderm consist of epithelial 
 cells in various stages of development ; these are best observed 
 in those parts of the integument which are transparent, as in 
 the lips of the proboscis, and between the different plates of the 
 pectus. The more superficial layer in these parts in the young 
 fly will be found to censist of a single layer of flattened angular 
 cells, containing well marked nuclei and beautifully coloured 
 with bright orange-coloured pigment. These cells are about 
 1- 1000th of an inch in diameter, and have a great tendency to 
 adhere at their edges. In the adult fly they form a continuous 
 membrane, but their nuclei become apparent on the addition 
 of a little acetic acid, I shall call this layer the mcsodcrm. 
 
 Beneath this, numerous spheroidal nucleated cells, slightly 
 angular by mutual pressure, are disposed in a layer of several 
 cells in thickness ; they seldom contain pigment, and correspond 
 closely to the rete-mucosum or deep layers of the cuticle in vcr- 
 
THE INTEGUMENT. 11 
 
 tcbratcs, but differ in being largely supplied with nerve filaments, 
 which ramify in a branching network amongst them. Following 
 the nomenclature I have already commenced, I think endoderm 
 an appropriate name for this innermost cuticular layer. 
 
 The development of the niesodemi and eudoderm is usually in 
 an inverse ratio, and this is no more than we might expect, 
 seeing that they are really different states of the same tissue. 
 In the fully developed fly the cellular structure of the nieso- 
 derm is almost or entirely lost, but it can be demonstrated in 
 many cases by carefully macerating the integument, drying it, 
 and mounting it in balsam, especially when viewed by the aid of 
 the polariscope. Its cellular structure is, however, beautifully 
 apparent in the insect soon after it emerges from the pupa. The 
 whole coloration, as well as the hardness of the integument, is due 
 to the indurated niesoderm.' As has been already observed it can 
 only be separated from the protoderm in those parts of the 
 insect in which it is but slightly developed. 
 
 The pigment with which the hard pails of the mesoderm are 
 chiefly coloured in the fly is deep blue or violet : whatever its 
 chemical nature may be, the colour depends upon oxidization. 
 It is first to be met with in the fat bodies of the larva. These are 
 perfectly white, but when cut from the larva and exposed to 
 the air they rapidly assume an inky blackness ; and the same is 
 true of the molecular fluid with which the pupa skin is filled. 
 When the perfect insect emerges from the pupa and respiration 
 again commences, the integument is nearly white, or a faint ashy 
 colour prevails. This soon gives place to the characteristic 
 blue or violet tint, first immediately around those portions 
 most largely supplied with air vessels, but soon over all those 
 parts of the insect which assume this colour; light seems to 
 have little to do with the process, but it is more rapid in warm 
 than in cold weather. The fat bodies of the larva are not per- 
 meated by air vessels, which will account for their retaining 
 their whiteness. The pigment is finely granular when examined 
 by high powers. 
 
12 ON THE ANATOMY OF THE FLY. 
 
 The chemical composition of the indurated mesodenn is but 
 partially understood. It is commonly said to be composed of 
 chitine, from its resemblance to a coat of mail. The induration 
 is partly due to animal tissue and partly to deposits of earthy 
 salts, especially phosphate of lime, which form beautiful double 
 refracting crystals in the wing cases of many beetles.* The animal 
 tissue is unlike the horny matter of vertebrates in being insoluble in 
 a hot solution of caustic potash, thus resembling vegetable cellulose 
 from which it is said to differ, however, in containing nitrogen. 
 
 The hairs with which certain parts of the body are clothed 
 are developed from the three layers of the integument. The bulb 
 is a single cell of the eiidoderm, and its cavity is continuous with 
 the hollow cavity of the hair. The hair itself is composed of 
 fibres which run through its entire length, giving it a channelled 
 appearance when sufficiently transparent. These are probably 
 developed from cells or nuclei of the mesodenn; whilst the 
 whole is invested by the protoderm. The hairs in the pupa at 
 the end of the first week consist only of hollow processes of 
 the protoderm filled with very transparent cells or granules. 
 No hairs arc developed on the pupa skin. 
 
 The base of each hair is surrounded by a thickened ring; this 
 is quite apparent from its earliest appearance, when the ring 
 exists only as a fold of the protoderm. 
 
 * I have used the term Crystal for these circular bodies , because I 
 believe they are similar to those remarkable bodies which were first pro- 
 duced artificially by the late Mr. Rainey by the crystalization of salts in 
 viscid fluids, and which are now known to abound in the animal kingdom. 
 Some naturalists think these sppts are only due to tension of the organic 
 substance of the elytra, but Mr. Hislop, who has paid the subject much 
 attention, tells me that their out-line is far more definite, when seen with 
 polarized light, than that of any spots he has seen produced by tension 
 alone, and that they disappear entirely when the wing-cases are treated 
 with chlorine or strong acids, whilst the lines of tension, which are distinctly 
 visible in the same elytra, are not effected by these reagents. 
 
THE NERVOUS SYSTEM. 13 
 
 Section IV. The Nervous System. 
 
 The nervous system, or rather, that part which is analogous 
 to the cerebro-spinal system of vertebrates, is next to the in- 
 tegumentary in point of importance as far as regards type. 
 Insects belong to that division of the animal kingdom which 
 Professor Owen terms Homogangliata, that is, animals in which 
 the nervous system is represented by a pair of nervous cords 
 which traverse the ventral portion of the body, and lie next to the 
 integument, uniting and forming a ganglion for each segment, 
 from which the proper nerves of the segment are given off. 
 No insect that has yet been examined departs from this type 
 so much as the fly and its allies ; for, even in the larva state, 
 the whole of the nervous system is collected in the anterior 
 segments, and the pair of ventral cords do not exist ; whilst, in 
 the imago, with the exception of two small ganglia in the pro- 
 boscis, there are but two nerve centres, one situated in the head, 
 and. one in the thorax. 
 
 The first of these, called the cephalic ganglion on account of 
 its position, is the homologue, that is, the anatomical represent- 
 ative of the oesophageal ganglion or ring of the lower forms of 
 life ; it surrounds the esophagus and gives off four pairs and one 
 single nerve, which are distributed to every part of the insect's 
 head. It is connected with the great thoracic nerve centre by 
 a thick cord which represents the double ventral cord, typical of 
 the insecta. 
 
 The nerves given off from this ganglion, or more correctly 
 speaking, collection of ganglia, for it is really composed of at least 
 six, may be properly divided into two classes, those which consist 
 entirely of nerve fibres having their origin in its substance, and 
 those of which some of the nerve fibres at least, probably the 
 greater portion, merely pass through it and have their real 
 origin in the thoracic ganglion. 
 
14 ON THE ANATOMY OF THE FLY. 
 
 The first class comprises five nerves,'-' those of the great 
 compound eyes and antennas, and the single nerve which 
 supplies the ocelli, or simple eyes. 
 
 The two remaining pairs of nerves are distributed to the 
 proboscis. The larger supply the greater part of that organ, but 
 chiefly the palpi and lips, and the smaller, which give off 
 filaments to the two small ganglia already mentioned, are 
 almost entirely lost in the muscles of the pharynx. Both pairs 
 are probably motor-sensory, or compound nerves. 
 
 The cephalic ganglion is in fact the centre of the special 
 senses in the fly. It is the analogue, that is the physiological 
 equivalent of the ganglia at the base of the cerebrum in ver- 
 tebrates ; and if, as I firmly believe, the antennoo are organs of 
 smell, it strictly represents the olfactory and optic lobes of 
 fishes. Next to bees and ants, that of the blow-fly is the largest 
 known in any insect proportionally to its size, being about 
 thirty times larger than the cephalic ganglia of the larger 
 beetles. But a more positive indication of a higher type of 
 organization than even the relative bulk of the sensory ganglia 
 is found in the fact that two very remarkable convoluted nerve 
 centres connected by a commissure, each about l-30th of an inch 
 in diameter, surmount the cephalic ganglion, and are connected 
 to it by a pair of distinct peduncles ; these are extremely like the 
 pedunculated convoluted nerve centres which occupy the same 
 position in bees and ants, first described by M. Felix Dujardin,f 
 and considered by him as analogous to the cerebral lobes of the 
 higher animals. That naturalist failed to distinguish these organs 
 
 * There is no nerve corresponding to the recurrent nerve of Lyonet 
 falsely called sympathetic by some authors, but a true sympathetic or 
 organic nerve system exists in the fly. 
 
 f F. Dujardin sur le Syste"me Nerveux des insectes., Ann. Sc. Nat. 
 Serie III. Tom. xiv. 195. and " Quelqucs Observations sur les Abeilles " 
 Ibid, xviii 231. 
 
THE NERVOUS SYSTEM. 15 
 
 in the fly, probably owing to their being imbedded in the sub- 
 stance of the cephalic ganglion. No one who has once seen them 
 can doubt their far-advanced type above the ordinary ganglia of 
 insects, and the superior intelligence manifested by the insects 
 possessing them is not difficult to perceive. In fact, long before 
 I detected these organs in the fly, I felt almost sure I should 
 eventually find them, from the fact that many of the acts of this 
 insect bear evidence of some memory at least ; for instance, the 
 manner in which flies will avoid a person who is pursuing them 
 is clearly more than an ordinary reflex act. 
 
 Nevertheless, it must be conceded that most of the acts of 
 insects are probably entirely reflex, or the result of impressions 
 from without, and this coincides with the comparative bulk of the 
 organs of sensation over that of the higher nerve centres. 
 
 M. Faivre, * however, thinks otherwise, from the fact that in 
 his experiments on the Dytiscus, that insect always gyrated to 
 the left when the right side of the supra-cesophageal ganglion 
 was removed, and to the right when the left was operated on, from 
 which he argues the existence of voluntary power. He sup- 
 poses the supra -CDSophageal ganglion to be analogous to the 
 brain of vertebrates; but M. Faivre clearly removed the great 
 organs of sensation, and the rotatory direction pursued by the 
 insect was probably entirely due to its seeking the direction in 
 which the light proceeded, and so turning away from the 
 injured side. 
 
 The thoracic ganglion gives off ten pairs of nerves, besides one 
 single one distributed to the abdomen, and the great trunk 
 which unites it to the cephalic ganglia. It is the homologue of 
 the thoracic and abdominal ganglia of the typical homogan- 
 gliata, and is, therefore, like the cephalic, a compound gan- 
 
 ' r M . Faivre Bur le Cerveau des Dytisques, Ann. Sc. Nat. Serie IT. 
 Tom viii. 245. 
 
16 ON THE ANATOMY OF THE FLY. 
 
 glion. This is the nerve centre of animal life ; its nerves are 
 both motor and sensitive ; it gives off likewise two pairs of nerves, 
 which are supplied to organs of special sense ; and its destruction 
 results in the instantaneous death of the insect. If a pair of 
 forceps are dexterously used so as to crush the whole thoracic 
 ganglion at once, not a single movement will follow its destruc- 
 tion, except it be the slightest possible quivering of the 
 muscles. 
 
 On the other hand, the removal of the insect's head, and with it 
 of course, the whole cephalic ganglion, does not destroy the life 
 of the trunk and limbs, but reflex movements can be produced 
 for hours after by touching any part of the integument, but 
 especially the integument of the abdomen. 
 
 When the head is removed frotm the fly there is nothing more 
 striking than the different character of the phenomena exhibited 
 by the parts. In the head a convulsive movementof the tongue 
 and antennae follows the division of the nervous cord which 
 unites the two nerve centres ; this continues at most a few 
 seconds, but no reflex act follows the application of external 
 stimulus afterwards. Of course we cannot tell to what extent the 
 functions of the cephalic ganglionremain, but their duration is pro- 
 bably short. Perhaps they cease as soon as the nerve of communi- 
 cation between the head and thorax is severed. On the 
 other hand, although the trunk continues to manifest signs 
 of life for many hours, no movements, except those which 
 result from the stimulus or shock produced by dividing the 
 great cephalo-thoracic nerve trunk at the moment of division, 
 take place, except the respiratory movement of the abdomen, 
 till some external stimulus is applied. A breath of air, or a 
 touch with a needle point in any part of the integument, will 
 produce vigorous movement of the legs and wings for hours 
 after decapitation, resulting either in an attempt to run or fly, 
 and a little water or dust placed upon the posterior tarsi or 
 
THE NERVOUS SYSTEM. 17 
 
 abdomen, will give rise to the well-known effort to clean it off, 
 by rubbing the tarsi together or against the part so irritated : 
 all these are clearly reflex acts. 
 
 These facts go far to prove that the same acts are always 
 produced by external stimulus in the living fly, in other words, 
 that they are reflex and not voluntary ; they also show how 
 extremely sensitive the integuments are to external impressions, 
 and that their hardness does not prevent their being the seat of 
 the sense of touch to a very great extent. Lastly, as there is 
 nothing irregular in any of these acts after decapitation, it seems 
 that the co-ordination of muscular movements is due to this nerve 
 centre, a function, which, according to the experiments of 
 M. Faivre,* is performed by the sub-cesophageal ganglion in 
 Dytiscus, and probably in all beetles : but I believe the homo- 
 logue of this ganglion is transferred in the fly to the thoracic 
 nerve centre, as the inferior portion of the posterior cephalic 
 segment enters into the composition of the thorax, and not of 
 the head. Thus the thoracic nerve centre appears to be the 
 analogue of the cerebellum, medulla oblongata, and spinal cord 
 of the vertebrata. 
 
 The nerves of the viscera are derived from a separate system 
 of ganglia, and differ in structure from those of animal life, like 
 the nerves of organic life, or the sympathetic system of vertebrates, 
 to which they may be justly compared. The chief ganglion of 
 this system is situated at the junction of the thorax and abdomen, 
 but numerous smaller ganglia varying in size are scattered over 
 the visceral organs, and the whole are connected by a complex 
 system of minute nerves, which form a close network over all 
 the viscera. Dr. Lockhart Clark tells me he has described a 
 similar system of nerves in the earthworm. 
 
 * Ann. Sc. Nat. IV. Tom riii 245. 
 
18 ON THE ANATOMY OF THE FLY. 
 
 Section V. The Wings and Legs. 
 
 The wings and legs are the most important thoracic append- 
 ages; their homologies have been already discussed.* 
 
 The wings consist of a double layer of the protoderm, which 
 may be demonstrated when the insect first emerges from the 
 pupa, as the upper and under layers are then easily separated. 
 The nervures are folds in one or both layers, which are 
 strengthened by the development of epithelial cells in their in- 
 terior. The larger ones contain the tracheal vessels and nerves, 
 the latter being chiefly distributed to the bulbs of the hairs on 
 the nervures. In the earlier stages of development the wings 
 are represented by mere crumpled sacs attached to the lateral 
 trachea) of the larva; afterwards they become sacs of the proto- 
 derm on the sides of the thorax in the pupa. 
 
 Several complicated folds of integument project into 
 the cavity of the thorax at the insertion of the wing, by 
 which the muscles are attached which regulate its position, but 
 the real muscles of flight are not connected with the wing at all. 
 They consist of a longitudinal mass which fills the greater portion 
 of the back, and the wings are so attached that the flank forms a 
 kind of fulcrum upon which they are elevated or depressed by 
 every alteration in the convexity of the back. The great longi- 
 tudinal dorsal muscles, by shortening the thorax, increase its con- 
 vexity and depress the wings, which are again elevated by the 
 flattening of the back and lengthening of the thorax, due partly 
 to its own elasticity and partly to the action of the lateral thoracic 
 muscles, which are vertical in their direction. So that in point of 
 fact the flight of insects is merely a modification of crawling, both 
 being effected by the alternate approximation and extension of 
 several segments. This may be confirmed by direct experiment ; 
 
 See Page 3, 
 
THE WINGS AND LEGS. 19 
 
 for, by taking a recently killed fly and using a pair of dissecting 
 forceps, placing one blade behind, and the other in front of the 
 upper part of the thorax, the movements of flight may be pro- 
 duced by'alternately compressing and relaxing the forceps ; the 
 wings should be extended in order that the full effect may be 
 seen. The amount of alteration in the convexity of the back 
 which is sufficient to depress the wings to their fullest extent, is 
 so slight that it is scarcely perceptible to the eye. 
 
 It is a remarkable fact, and one worthy of special attention, that 
 the great longitudinal thoracic muscles exhibit no striae, but con- 
 sist of muscular fibres similar to those of organic life in the higher 
 animals. This is an anomaly for which I can suggest no reason, 
 unless it be the immense rapidity of the vibrations into which 
 they throw the thoracic parietes.* 
 
 The legs consist of four parts : the coxa or hip, composed of two 
 pieces, which unites the limb to the thorax, the femur or thigh, the 
 tibia or leg, and the tarsus or foot ; the last three are tubular, and 
 each articulation contains the muscles which move the succeeding 
 one. The tarsus consists of five pieces, the first of which contains a 
 pair of muscles, which move the second upon it, but the remaining 
 four contain none ; the last bears a pair of pads upon which 
 the insect walks, and a pair of hooks above them. Both hooks and 
 pads are connected with a single apodeme which projects into 
 the last tarsal joint and is continued as a fine chitinous thread 
 or tendon through the tarsus and tibia, and terminates in a bi. 
 penniform muscle which arises from the interior of the 
 femur. 
 
 The foot pads are amongst the most interesting parts of the insect, 
 because they enable it to walk upon smooth surfaces in an inverted 
 position, apparently in defiance of the laws of gravity. Long ago 
 
 * The description of the histological elements of these muscles is given 
 at page 5. 
 
 C 2 
 
20 OK THE ANATOMY OF THE FLY. 
 
 this was first ascribed by Dr. Derham* 4 to the exhaustion of air 
 from the foot pads; recently, it has been supposed to be due to the 
 exhaustion of air from the extremities of the hairs with which 
 the pad is clothed ;f others have ascribed it to be the holdj which 
 these minute hairs take of trifling irregularities of surface, but none 
 of these explanations are correct, and one of the earliest notions upon 
 the subject is the nearest to the truth; that is, that the feet se- 
 crete a glutinous fluid which glues them to the surface on which the 
 insect walks. When the pads are carefully examinod it will be seen 
 that they have no cup-shaped cavity beneath them, but that they 
 are hollow with a nipple-like protuberance projecting into each. 
 This will be seen more plainly by pressing upon the tarsus which 
 forces it into the pad ; by cutting off the end of the pad first, it may 
 be exposed in this manner, and will be found to consist of a 
 closed sac. 
 
 This sac fills the whole of the last four tarsal joints and 
 is lined with pavement epithelium ; it secretes a perfectly clear 
 viscid fluid which exudes from it into the pad, and fills its cavity 
 as well as the hollow hairs with which its under surface is covered. 
 These hairs open by trumpet-shaped mouths, and the disc of each 
 mouth is kept full of the fluid. Sometimes, when the insect is cap- 
 tured and held between the finger and thumb, it exudes so rapidly 
 that the pads are soon covered with a little glistening drop of it, 
 which may be collected upon a glass slide where it rapidly solidi- 
 fies ; it is insoluble in water and solidifies under that fluid. The 
 whole contents of the tarsus becomes solid very rapidly as soon 
 as the insect is dead, or the part is removed. 
 
 * Kirby and Spence, Introduction to Entomology. The whole history 
 of the various opinions of different Naturalists upon this subject are given 
 at length. 
 
 t J. Hepworth, Koy. Micros. Soc. Journal, vol. iii. 314. T. West. Linn 
 Soc. Trans, vol. xxiii. 393. 
 
 t Blackwall, Linn. Soc. Trans, vol. xvi , on the pulvilli of insects. 
 
TJIC WINGS AND LEGS. 21 
 
 There is no essential difference in the pads of flies and the 
 pulvilli of beetles, moths, and other insects ; the same fluid is 
 secreted in all. The only difference is that the pads of flies are 
 membranous and transparent, instead of hard and opaque. 
 
 The feet of the smaller house-fly are the best to show the 
 manner in which the viscid fluid exudes from the extremities 
 of the trumpet-shaped hairs, as they are very large in this species, 
 and a glistening bead of fluid can be seen plainly at the extremity 
 of each hair by placing the living insect under the microscope. 
 The foot-prints left upon glass by flies consist of rows of dots 
 corresponding to these hairs ; this is best seen in those of the 
 lesser house-fly from their greater size. 
 
 The whole appears precisely analogous to the manner in which 
 caterpillars and spiders suspend themselves by silken threads. In 
 both cases the fluid is exuded from minute pores and bears the 
 weight of the insect, the only difference being in the nature and 
 quantity of the fluid exuded. Much discussion has arisen as to 
 the manner in which flies liberate their feet, and it has even been 
 objected that they would become so firmly adherent after a time 
 that the insect would be glued to the spot. Nothing can be simpler 
 than the arrangement by which the foot is liberated, and in the 
 healthy insect the secretion probably never becomes solid as long 
 as it remains in contact with the foot. It is sufficiently glutinous, 
 even in the fluid, or rather semi-fluid, state it assumes as it exudes, 
 to sustain the weight of the insect, when the strain is put equally 
 upon all the hairs, of which there are about 1200 on each pad; 
 but when the pad is removed obliquely, so that each row is 
 detached separately, the resistance amounts practically to nothing. 
 A neat experiment will demonstrate this even to the most sceptical. 
 If a piece of adhesive label be cut for convenience into a pear-shaped 
 disc, an inch in diameter, and caused to adhere to the hand by 
 slightly damping it, a force of many pounds applied to the narrow 
 extremity in the axis of the paper will not stir it, whilst it is im- 
 mediately removed, with very little resistance, when the force is 
 applied so as to lift it gradually up. 
 
22 ON THE ANATOMY OF THE FLY. 
 
 The direction and length of the hairs upon the pad are so 
 adapted to the oblique direction in which the strain is put upon 
 them when the tarsus is straight, that the insect has a perfectly 
 secure hold ; this is immediately released as soon as the tarsus is 
 curved, which is effected by the long slender tendon already men- 
 tioned. In the small house-fly the pads themselves are capable of 
 being curved, for the tarsal tendon branches, and is inserted into 
 the distal extremity of each pad. 
 
 The legs are developed from six pear-shaped corpuscles ad- 
 herent to the nerves andtracherc of the larva ; the changes these 
 undergo are seen in Plate VI., and are described in the second 
 part of this work. 
 
 Section VI. The Digestive System. 
 
 The viscera of the fly are remarkable for their symmetrical 
 arrangnnent, a fact well worthy of note, as it is at variance with 
 the usual disposition of the viscera of animals, although I 
 believe lateral symmetry will prove the rule amongst insects. 
 
 The organs of the mouth are extremely modified so that they 
 form a complicated proboscis admirably adapted for the purpose 
 of collecting food by suction ; and, although all the parts homo- 
 logous to the six lancets of the gadflies so terrible in those 
 insects, are present, they are all more or less intimately united to 
 the sheath of the proboscis, so that they are quite incapable of 
 inflicting a wound, but form by their union a long tubular mouth. 
 The extremity of the organ bears a pair of large fleshy lobes, 
 which form an oval sucker when open, by which the insect 
 collects its food, either receiving the nutriment so collected at 
 once into the opening of the mouth, which is placed between the 
 lobes, or straining the fluid from the solid part of the material on 
 \rhich it is feeding by means of a set of channels ( false trachea: ), 
 
THE CIRCULATION. 28 
 
 which open beneath by a very fine fissure and run from the margin 
 to the centre of the lobes, and then open into the mouth. This 
 is a provision highly necessary to the fly, as it feeds upon half- 
 rotten pulpy substances, which would otherwise entirely fill 
 its long tubular mouth and completely stop it up. A shuttle- 
 shaped chitinous organ supports the parts of the mouth, 
 and is enclosed within the proboscis. This acts as a pump, 
 drawing the fluid aliment from the mouth of the insect, and 
 injecting it into the sucking stomach or crop.* 
 
 The crop is a large reservoir for food, and is situated in the'ab- 
 domen, it is capable of holding a sufficient supply of nourishment 
 for several days. No solid food is taken by the fly, as it lives 
 entirely upon fluids or such substances as are dissolved by its 
 copious saliva, which is secreted by a pair of large tubular salivary 
 glands. The aliment is gradually regurgitated from the crop into 
 the proventriculus and chyle stomach,f which extends the whole 
 length of the thorax, and which is lined with cylinder epithe- 
 lium undistinguishable from that in the highest vertebrates. 
 
 The chyle stomach passes insensibly into the intestines, which 
 are about one inch in length. A pair of large bile tubes open 
 into them near their inferior extremity, and form a tubular liver 
 of considerable magnitude by continually dividing, and end in 
 blind extremities; they are very tortuous, and are filled with liver 
 cells, containing fat globules and pigment. 
 
 Section VII. The Circulation* 
 
 The circulation of the blood in insects is earned on with- 
 out the aid of blood-vessels. The circulating fluid is contained 
 
 * Sucken-Magen, Weismann, f Chyhis-Magen, Ibid. 
 
24 ON T.HE AX ATOMY OF THE FLY. 
 
 in the visceral cavity and bathes all the internal organs. A single 
 vessel runs the whole length of the back, commencing near the 
 apex of the abdomen and terminating in the head, called the 
 dorsal vessel. It is open at either extremity, and serves the 
 purpose of a heart pulsating rythmically, and pumping the 
 circulating fluid from the posterior extremity of the insect, 
 into the head, from which it returns amongst the visc- 
 era. The pulsation of this vessel may be observed in the fly 
 just after the insect emerges from the pupa case, before its 
 integuments become hardened and opaque ; it makes about 1 80 
 pulsations in a minute. 
 
 The dorsal vessel being small in comparison to the cavity oi 
 the body, the return of the blood is necessarily slow. This gives 
 rise to a peculiar distribution of the viscera as well as to a very 
 important modification of the respiratory organs and function, 
 adapting the insect to pass a life of great vital activity, with a 
 comparatively slow circulation. 
 
 Thus the salivary glands not only extend the whole length of 
 the thorax in complicated convolutions, but pass into the abdo- 
 men and only terminate at its posterior extremity. This brings 
 them into relation with the blood throughout a large part of the 
 circulation. So the ductless glands, called fat bodies or corps 
 graisseux, which seem to be analogous to the ductless glands of 
 animals, and to be concerned in the elaboration of the circulating 
 fluid itself, are found largely developed in the abdomen, and 
 also, though more sparingly, in the head. The analogues 
 of the kidneys, the rectal papillae, pulsating organs having a 
 central cavity alternately filled and emptied of blood by a 
 rhythmic muscular act, are placed near the posterior extremity 
 of the dorsal vessel, so that they excrete the efiete matter from 
 the circulation just before the return of the blood to the head 
 and thorax. The same may be said with regard to the dis- 
 position of the liver tubes, only to a less degree. 
 
THE HESI'UI.VTOKY SYSTEM. 
 
 Section VIII. The Itespiratofy tii/stem. 
 
 The cliief modification of structure dependent upon the 
 peculiarity of the circulation is in the respiratory organs, the 
 air being carried to all the tissues in a system of tubes, called 
 trachea?, from their cavity being kept open by a spiral fibre, 
 or more often by a series of rings which run into each other so 
 as to appear spiral. These tubes open externally by pores called 
 spiracles, and are merely an involution, so to speak, of the proto- 
 derrn or external layer of the skin ; they ramify in the interior 
 of the insect, dividing in an arborescent manner, until at last 
 they attain the minute diameter of something like 100,000th of 
 an inch, when they anastomoze freely with each other and form 
 a network of tubules over the surface, and in the substance, of 
 every organ. 
 
 The trachea! tubes consist of two coats. The internal, in the 
 substance of which the spiral fibre is developed, corresponds 
 to the protoderm, and like it, is formed from cells, which in the 
 case of the perfect fly originate in the pupa case, in the earliest 
 stages of the formation of the imago, Similar rings to those 
 in the tracheae are developed, either from cells or nuclei of the 
 protoderm, in the false tracheae of the proboscis, an important 
 fact, as it is confirmatory of the view I have taken of their 
 homologies.* The external layer is easily detected before 
 the fly attains its full development ; it is thick and struc- 
 tureless at first, containing nuclei, and thickest in their vicinity; 
 it seems to be a mere collection of protoplasm. As the 
 
 * The origin of the rings of the false trachcaj is even better seen in the 
 tongue of the Cricket, where the nuclear form is retained throughout life, 
 near the edges of the organ ; this was first pointed out to me by my 
 friend, Mr. C. Stewart. 
 
20 ON THE ANATOMY OF THE FLY. 
 
 insect reaches maturity, this layer becomes extremely attenuated, 
 and forms a structureless membrane very difficult to trace ; it is 
 best seen in the tracheae of the larva. 
 
 The spiral fibre has been described as a distinct coat between 
 the other coats, but it is really part of the internal. M. 
 Blanchard* even supposed that a circulation of the nutritive fluid 
 takes place in insects, in a channel between the coils of the spiral 
 and the two layers of membrane, which he supposed to be con- 
 nected with the dorsal vessel ; but no such channel exists. f 
 
 It is extremely difficult to say at what point in the division of the 
 tracheal vessels the spiral disappears. Most of the vessels which 
 are as small as the 10,000th part of an inch, exhibit no trace of 
 transverse marking, but a very distinct double contour is observ- 
 able in their walls ; whilst in others of the same magnitude dis- 
 tinct transverse markings are visible with high powers, at irreg- 
 ular intervals. The smallest divisions have probably 110 spiral, 
 and consist, I believe, entirely of the external coat, which 
 assumes the form of a structureless membrane. j 
 
 The larger tracheaB of the head and thorax of the perfect fly 
 differ very considerably in structure from those already described, 
 and form large reservoirs of air. Numerous sacculi which lie 
 between the muscles, mere dilatations of these tracheae, open into 
 them, and the main lateral trachea) terminate in a pair of large sacs 
 at the base of the abdomen, the pulmonary sacs c f the French 
 naturalists. The walls of these vessels and their sacculi are 
 composed entirely of a brittle membrane, corresponding to the 
 
 * Comp. Rend. xxiv. 
 
 t It is needless to repeat the arguments against such a circulation, which 
 even if the spiral were a distinct layer, are fully given in thexxviii. and 
 xxix. vols. of the Comp. Rend. 
 
 * In the stag beetle the irregularity of the spiral in the smaller tracheal 
 tubes is easily seen. 
 
THE RESPIRATORY SYSTEM. 27 
 
 protoderm ; it is extremely delicate, covered with reticulated 
 wrinkles, and marked with striae, transverse to the wrinkles, not 
 the 12,000th of an inch apart, which render it iridescent. This 
 membrane, as well as the inner coat of the other tracheae, assumes 
 a smoky tint in the adult insect, probably from the oxidization 
 of minute particles of pigment contained in its substance.* 
 
 The spiracles or external openings are all protected, either by 
 folds, the edges of which are armed with minute spiculae, or by 
 a ring of spiculae formed of chitine, and the admission of air is 
 controlled by a structure somewhat resembling a pair of 
 callipers, acted upon by a special muscle, which surrounds 
 the tracheae and is connected with their internal coat.f 
 
 The manner in which the circulation of air in the trachece is 
 carried on in insects is by no means certain ; that it is exceed- 
 ingly vigorous in the fly may be fairly assumed from the rapidity 
 with which the vapour of chloroform and other volatile sub- 
 stances act upon the insect ; and from the fact that a little turpen- 
 - tine, applied with a camel's hair brush to one of the anterior 
 thoracic spiracles, produces instantaneous insensibility. The 
 great vital activity of the fly indicates that the respiratory func- 
 tion is very perfect ; in bees again the high temperature of the 
 hive tends to the same conclusion. 
 
 I believe the pressure of the thoracic muscles acting unequally 
 upon the main membranous trachece and their sacculi, aided by 
 the valves at the spiracles, is the chief agent by which the air 
 is propelled through the smaller tubules. If such be the case, the 
 ordinary muscular acts of the insect would be sufficient to keep 
 up a very perfect circulation, as it is scarcely conceivable that 
 any muscle can contract, without giving rise to some movements 
 of the air in the large trachea). When a fly is held by the wings 
 
 * The disposition of the main tracheal vessels is shown wi Plate I. 
 f Landois, Kolliker Zeichrift, Band. 17, 105. 
 
28 OX THE ANATOMY OF THE FLY. 
 
 it will be observed to make vigorous exertions with all its legs 
 at regular intervals, varying from sixteen to thirty or more 
 in a minute. With each of these acts the valves of the anterior 
 spiracles are closed for a short space of time, during which the 
 air must necessarily be driven through the small tnbes, proving 
 that these are respiratory efforts. Their regularity and their 
 being accompanied by a contraction and dilatation of the 
 abdomen confirms this view. Again, if placed under the 
 exhausted receiver of an air pump, and removed before 
 death has taken place, the insect's abdomen immediately collapses 
 more or less from the exhaustion of the trachea! vessels, and 
 more especially of the abdominal pulmonary sacs; and the refilling 
 of the trachea! system is accompanied by violent movements of 
 the legs and wings. 
 
 Although, under ordinary circumstances, no proper respiratory 
 movement of the abdomen can be observed, if the head be re- 
 moved, a regular respiratory movement of the abdominial 
 parietes may be seen, and the same is observable when the in- 
 sect is at rest. There can be no doubt that every alteration 
 in the capacity of that cavity acts directly upon the pulmonaiy 
 sacs, and so upon the whole tracheal system. These ab- 
 dominal movements are scarcely sufficient to account for the rapid 
 change of air in the small trachea of the head and thorax, and 
 are probably only accessory to the ordinary muscular movements 
 in their effect on the air in the trachea?. The change in the volume 
 and shape of the thorax during the movements of the wings, 
 although slight in itself, is probably from its great rapidity also an. 
 important respiratory agent; and lastly, the pressure of the muscles 
 upon the smaller tracheae undoubtedly assists the movement of 
 air in these tubes by emptying them, the elasticity of their 
 internal coat causing them "to dilate and refill as soon as the 
 pressure is removed. 
 
 The humming sound made by the insect seems to depend upon 
 several causes, but the chief seems to be the rapid vibration of 
 
THE FAT BODIES AND DUCTLESS O LANDS. 29 
 
 the edges of the valves of the thoracic trachea?,* perhaps also of 
 a plate in the forehead, which covers a membranous cavity in that 
 part. The sound is attended with a violent vibration of the whole 
 thorax, probably chiefly produced by the rapid contraction of 
 the dorsal muscles. The vibration of the wings may, and probably 
 iloes, give rise to a peculiar note ; but owing to their movement 
 being always attended with vibrations of the thorax, it is difficult 
 to determine how much is due to them. Another note is pro- 
 duced by the rubbing of the head against the thorax. | 
 
 Section IX. The Fat Bodies and Ductless Glands. 
 
 Two dissimilar structures have been confounded under the 
 term fat bodies, the omenta of the larva, and the ductless glands 
 of the imago. 
 
 The former consist of plicated cellular expansions formed 
 of large flat cells of hexagonal outline, adhering to each other by 
 their edges, which are firmly attached to the lateral trachea? 
 of the larva, in which a store ef elaborated nutriment is laid up 
 for the development of the pupa. 
 
 The fat bodies or ductless glands of the imago consist of closed 
 follicles, arranged upon the tracheae in an arborescent manner ; 
 the follicles themselves are composed of structureless membrane, 
 and contain peculiar nucleated corpuscles, nuclei, and granules ; 
 in the mature insect they are often loaded with oil globules. From 
 
 * Dr. H. Landois, Die Ton und Stimmaparate der Insecten, Kolliker 
 Zeitchrift, Band. 17, 105. 
 
 t Dr. H. Landois and W. Thelen, Der Tracheenverschluss bei den 
 Jnsecten. ibid. Band 17, 187. 
 
30 ON THE ANATOMY OF THE FLY. 
 
 the great similarity of the contents of these follicles to those of 
 the spleen and other ductless glands in the vertebrata, I have 
 not the slightest doubt as to the similarity of their functions, 
 which is the elaboration of the circulating fluid. 
 
 Section X. The Organs of Special Sense. 
 
 The organs of special sensation in the fly are the compound 
 and simple eyes, the antenna, the halteres, and probably the 
 frontal sac, the cephalo-sternum, the lobes of the proboscis, 
 and maxillary palpi. 
 
 With the exception of the visual organs, which closely 
 resemble the essential parts of those of the vertebrata, great 
 difficulties occur in determining the functions of the sensory 
 organs, and the removal of any of these, even when this can be 
 done without injury to other parts, is no test of its function. Flies 
 seem completely helpless when the antennee are removed, and 
 quite unable to fly if the halteres are cut off, yet this is no proof 
 that the quickness of the insect is due to its antenna), or that 
 flight is a function of its halteres. All the senses seem so closely 
 bound together, or correlated to each other in insects, that we can 
 hardly expect any to be normally exercised when one^is damaged. 
 Even in man himself, we know that nothing but experience 
 enables him to judge correctly of objects with a single sense, and 
 we should hardly expect a dog to act normally, if one of its 
 senses were suddenly removed. Apart from the pain which may 
 accompany the removal of an organ, the new relations with 
 the external world might be expected to produce an abnormal 
 manifestation in a creature, almost all the acts of which are 
 probably purely reflex. So that structure, analogy, and the habits 
 of the insect, are really the only guides in such investigations. 
 
 The great compound eyes each consist of between four and 
 five thousand facets; each facet being a circular biconvex lens, the 
 thousandth of an inch in diameter, set in a flat hexagonal frame. 
 
THE ORGANS OF SPECIAL SENSE. 31 
 
 Behind every facet is a transparent cone, surrounded by dark 
 red pigment, which terminates posteriorly in a rod, connect- 
 ing it, through the medium of a layer of nerve cells, with u 
 clear distinct nerve filament. 
 
 The convexity of the facets of the cornea is a most important 
 point, as the vision of insects has been supposed to derive its 
 distinctness from the number of separate facets. It has been 
 supposed that distinctness of vision is due in them to the 
 absorption of all the rays of light which have not the same 
 direction as the axes of the cones and rods of which the eye is 
 composed. Such is not the case, however, in the fly ; the 
 lenticular structure of the cornea is most distinct, and the focus 
 of each lens appears to correspond with the apex of the 
 transparent cone beneath it. The necessity of any adaptive 
 power to distances, it may be, is entirely obviated by the small- 
 ness of the lenses, as the rays of light emanating from a point 
 an inch distant from the eye could not possibly make a greater 
 angle than a minute, so that virtually all rays would be parallel 
 when the distance of objects from the eye exceeded an inch. 
 Yet it is far from certain that even adaptation to distance is not 
 effected by a special arrangement of tracheal tubes between the 
 rods and cones.* Perhaps nearer objects are perceived only by 
 the ocelli, which are specially adapted for near vision, for which 
 the comparative flatness of the lenses of the compound eyes seems 
 but little fitted. 
 
 The great convexity and extent of the compound eyes is such 
 that it enables the insect to see all the objects in at least four- 
 fifths of the sphere which surround it; but as no two facets look 
 towards the same spot, no provision for stereoscopic vision 
 exists in these organs, even if the insect be able to perceive 
 near objects by them. These deficiencies seem to be supplied 
 by the ocelli, which, from the great convexity of their refracting 
 
 * Page 88. 
 
02 ON THE ANATOMY OF THE FLY. 
 
 medium and their position in a triangle upon the forehead, 
 seem not only eminently adapted for the perception of very 
 near objects, but also for the nicest possible appreciation of their 
 distances from the insect. 
 
 Beneath the cornea in the ocelli is a thick layer of nerve 
 cells, which probably receives the impression, and it is quite 
 likely that the adaption to distance depends upon the portion of 
 this layer upon which the picture falls. I see no reason why 
 the deeper cells should not receive an impression as well as the 
 more superficial, and the difference of the focus cannot be very 
 great. The whole internal surface of the ocellus is covered with 
 pigment of a bright orange red color, and the three ocelli are 
 connected with the cephalic nerve-centre by a single nerve. 
 
 The antennae have the third joint remarkably dilated, but 
 the others are comparatively little developed, except the last, 
 which is very slender and covered with hairs. The third joint is 
 covered all over with minute openings, which communicate 
 with little transparent sacs or groups of sacs in the interior. The 
 whole joint is lined with orange red pigment cells, and filled 
 with a pulpy mass, which consists of the antennal nerve, minutely 
 divided, and bearing multitudes of very small nerve cells, 
 arranged like bunches of grapes at the extremity of its divisions- 
 I believe myself that this is the organ of smell, although I by 
 no means consider the antemiee of all insects are necessarily 
 olfactory organs. I think in many instances they are merely 
 feelers. Perhaps the beautiful feather-like aiitennee of male 
 moths are sexual ornaments, although they may have special olfac- 
 tory organs connected with them ; and possibly the laminated an- 
 tennae of many beetles, which consist of thin chitinous lamellae, 
 may be hygrometric, indicating the state of the atmosphere to 
 the insect. I have littlg doubt, however, in other insects, 
 as in the fly, especially when they are thick and club-shaped, 
 that they are olfactory, or rather partly olfactory organs. 
 
 The halteres are most evidently modifications of the posterior 
 
THE OBGANS OF SPECIAL SENSE. 33 
 
 wings, they are also called balancers, from an idea that they are 
 connected with flight. A very large nerve, terminating in nerve 
 cells fills their cavity, and is connected with a number of 
 sin ill, highly refracting bodies, regularly arranged around the 
 base of the organ, which I am inclined to regard as oto- 
 conia, and if such is the case, the halteres are manifesty 
 organs of hearing. That flies do hear, may be inferred from 
 their possessing the power of emitting special sounds, and also by 
 the manner in which they disappear when one is captured and 
 allowed to make the peculiar plaintive note, which they in- 
 variably emit when distressed. In grasshoppers and other 
 orthoptera, the organs of hearing are situated on the anterior 
 tibiae, so that the position of the halteres cannot be urged as an 
 argument against such a function. A number of minute organs* 
 similar to those, which I believe to be otoconia in the halteres^ 
 are disposed in regular groups in the sub-costal nervure of the 
 wing. These are remarkably constant in insects, and are 
 especially interesting in relation to the halteres of flies, leading 
 us, as it were, gradually up to the complete modification of the 
 second pair of wings into organs of special sense. 
 
 The frontal sac may be connected with the sense of smell. The 
 cephalo-stemum is a most remarkable organ, hitherto undescribed : 
 and the lobes of the proboscis are probably the seat of taste, as 
 they are largely supplied with nerves and have peculiar nipple- 
 shaped organs on their surface. The palpi are probably also con- 
 cerned in taste when the food is regurgitated ; and perhaps they 
 assist the insect in its search for food as they are covered with 
 minute openings with transparent sacs beneath, like those of the 
 antenna?. Many of the hairs which clothe the various parts of the 
 insect may also be endowed with special sense, as their bulbs are 
 supplied largely with nerve filaments. I think it extremely pro- 
 
 * These organs were first pointed out by Dr. Braxton Hicks, but their 
 internal structure has not hitherto been described. 
 
34 ON THE ANATOMY OF THE FLY. 
 
 bable that their function is analogous that of the vibrissse, or 
 feelers of carnivorous animals. 
 
 Section XI. The Generative Organs. 
 
 The sexes are quite distinct in insects. The female fly is 
 distinguished from the male by the great eyes being more nearly 
 approximated and larger in the latter, as well as by the sexual 
 organs. These consist in the male of a pair of testes, a pair of 
 tubular glands, probably secreting an albuminous fluid, with 
 their ducts ; and of extremely complicated external organs, con- 
 sisting of the four last abdominal segments. 
 
 The generative organs in the female consist of a pair of ovaries 
 with their ducts. Three curious chitinous sacs which receive 
 the male fluid, a remarkable receptacle containing two peculiar 
 gelatinous masses after impregnation, beside a pair of tubular 
 glands, which probably secrete the glutinous fluid with which the 
 ovum is surrounded when it is deposited. The ova are fecundated 
 at the time of deposition, and the larva is hatched twenty four hours 
 after fecundation ; but sometimes a single ovum is fecundated 
 and retained within the ovipositor until the larva is perfected, 
 and the insect is thus occasionally viviparous, or the egg is laid 
 only a tew seconds before hatching. There are three or four sets 
 of ova contained in the ovaries, which are matured in succession. 
 
 The ovipositor consists of four joints, and is very long ; it is 
 introduced within the genital- fissure of the male, during copula- 
 tion, and it enables the insect to deposit its eggs in deep fissures, 
 or beneath decaying matter. It is furnished with a pair of 
 leaf-like appendages at its extremity, but has no hooks in the 
 blow-fly, although that of the house-fly is furnished with a set of 
 spines, pointed backwards around every ring. 
 
CONCLUDING REMARKS. 35 
 
 The ova undergo yolk segmentation, and just before impreg- 
 nation, possess a micropyle, germinal disc, and yolk canal, just 
 like the ova of the highest vertebrates. The skin of the larva is 
 formed from delicate cells, and is complete eight hours after im- 
 pregnation, but neither trachea3 nor viscera can be discerned so 
 early ; these are afterwards rapidly developed, and four and twenty 
 hours after impregnation the larva escapes, its alimentary canal is 
 filled with oil globules and molecular matter, and the trachea3 with 
 air, at the time. The fat bodies are at this period scarcely percep- 
 tible, a number of delicate cells only being attached to thetrachere. 
 The larva commences immediately to feed, and increases rapidly 
 in size, the fat bodies become largely developed, and at the end 
 of three or four weeks it assumes the pupa state. Its further 
 changes have been already described. 
 
 Section XII. Concluding Remarks. 
 
 The nervous system indicates that the Diptera, Hymenoptera 
 Xeuroptera and Orthoptera, are the most highly organized of all 
 insects ; the presence of convoluted nerve centres supported by 
 peduncles, and united by a cominisure, which do not give off 
 nerves, is an approach to the nervous system of the Vertebrata, 
 and such nerve centres are perhaps the most important charac- 
 ters of the above mentioned orders. The elaborate modifications 
 of the mouth in these orders for special purposes, the high 
 development of the sense organs, and in the Diptera the remark- 
 able modification of the posterior wings, and the articulation 
 of the head with the thorax by a pair of condyles in some, point 
 to a high type of organization. But at present it is extremely 
 difficult to decide which of these orders is most highly organized : 
 they will probably eventually form a distinct sub-class of the 
 insecta, characterized by its highly organized nervous system ; 
 and its divisions may perhaps be founded upon the various 
 
 D 2 
 
36 ON THE ANATOMY OF THE FLY. 
 
 degrees of metamorphosis, especially the presence or absence of 
 imaginal discs in the larva, as suggested by Dr. "Weismann ; or 
 upon the manner in which the head is articulated with the thorax, 
 But until the types of other families have received the same 
 attention as the fly, nothing certain can be said upon the subject. 
 
 In conclusion, it may not be out of place here to make a few 
 observations on the position of the insectain the animal kingdom. 
 The highest insects should probably be regarded as the highest 
 members of the annulose type ; which is apparently so distinct 
 from the vertebrate type that they cannot be compared with each 
 other ; for my own part I confess that I see no evidence of in- 
 feriority of type in either case, the structure of insects is as com- 
 plex and their organs are as perfectly differentiated as those of the 
 vertebrata. Professor Huxley has very ingeniously compaired 
 the embryonic forms of the vertebrate and invertebrate types* and 
 supposes the nervous system of the insecta to be homologous with 
 the sympathetic system of vertebrates, and that the cerebro -spinal 
 nervous system and neural canal are added in the latter ; such 
 can hardly be the case if my observations on the nervous system 
 of insects are correct, as they have also two distinct nervous 
 systems. Even if such an homology should be substantiated, which 
 appears unlikely, the supposition that the alimentary canal ter- 
 minates anteriorly by a neural instead of a ha3mal flexure would 
 not explain the different development of the embryo from the 
 entire surface of the yolk, and the fact that all the organs of man- 
 ducation, prehension, locomotion, and the external organs of gen- 
 eration, originate from the lateral appendages of the various 
 segments. No homologies can be traced for any of these organs 
 between the insecta and vertebrata. 
 
 Introduction to the classification of animal?, 
 
PART. II . 
 
 DETAILS OF THE ANATOMY OF THE FLY 
 
 Section I. The Integument of the Head. 
 PLATE II, FIGS. 1 and 2. 
 
 The head presents several regions which have been variously 
 ivainei by entomologists ; the most important are the occipital or 
 posterior, the frontal or forehead, the cheeks, and the facial, or 
 that between the antennce and the niouth. A correct apprecia- 
 tion of these will materially facilitate the[description of the head, 
 which, with its appendages, consists, as has already been stated, 
 of five segments ; of these however three belong to the 
 proboscis, and three-fourths, that is the ventral and lateral plates 
 of the fifth segment, are incorporated with the thorax ; so that 
 the integument or external skeleton of the head itself consists 
 entirely of the fourth, united with the dorsal plate of the fifth 
 segment. The hard parts of the head are best examined by 
 soaking or boiling it in liquor potassa?, drying it in hot turpentine 
 and mounting it whole in a deep cell in Canada balsam, 
 without pressure: this will also be found the best method of 
 examining most of the integumentary organs of insects with low 
 powers ; it is troublesome at first, but becomes easy by practice 
 
 Some little difficulty may be experienced in realising the 
 relations of the segments of the head, but by considering the 
 
38 ON THE ANATOMY OF THE FLY. 
 
 DESCRIPTION OF PLATE II. 
 
 This plate represents the hard parts of the head and proboscis, with the 
 three anterior larval segments for comparison with those of the proboscis 
 
 Fig. 1. An anterior view of the head of the male, with the part 
 belonging to the fifth segment colored ; the hairs are removed and are 
 represented only by their scars. X 10 diam 
 
 Fig. 2. Posterior view of the head of the female showing the greater 
 width of the frontal plate ; colored like the last, x 10 diam. 
 
 Fig. 3 and 3<t. Ventral and side views of the hard par tr. of the mouth 
 of the adult larva, consisting of a, the pharynx to which the appen- 
 dages bb are attached ; it is probably composed of the third segment 
 united to the true pharynx, as in the perfect insect. The other parts are 
 represented on a larger scale in figs. 4 and 4, x 25 diam. 
 
 Figs. 4 and 4a. Ventral and side view of the first and second segment 
 of the same, c, second segment, rf, larval hooks or mandibles, e, labrum? 
 / g* parts of the first segment, constituting the margin of the mouth, x 7 5 
 diam. 
 
 Figs. 5 and 50. The triangular opening of the mouth of the perfect fly, 
 seen from the front and side ; a a, rudimentary mandibles; &, apodemes of 
 the same; c, labium ; c?, labial appendages, x 75 diam. 
 
 Fig. 6. The labium and tongue. X 25 diam. 
 Fig. 6a. The apex of the tongue x 75 diam. 
 
 Fig. 66. Section of the mouth, through the lower part of the oper- 
 culum . 
 
 Fig. Qc. Section of the second joint, through "the upper part of the 
 operculum. 
 
 Fig. 7. The operculum composed of the labium and terminal lobes 
 of the maxillae, with their apodemes. x 25 diam. 
 
 Figs. 8 and 9. The mentum. x 25 diam. 
 
 Fig. 10. Scale-like basal lobes of maxillae, with the basal bands bearing 
 the maxillary palpi, x 12 diam. 
 
 Fig. 11. The fulcrum, consisting of the dorsal plate of the third segment 
 the epistoma and the chitinious pharynx united, dorsal view, x 25 diam. 
 Fig. 12. The same with its muscles attached. 
 
 Fig, 13, Side view of the same, to show the attachment of the muscles; 
 a, exsertor muscles ; b, pharyngeal musdes; c, c/, muscles to apodemes 
 of the operculum ; e, muscles of salivary valve. 
 
 Fig 14. Section through the lower part of the pharynx, showing the 
 central cavity and the disposition of the muscles by which it is opened. 
 
 . Figs. 15 and 15a, Small scale for protecting the bend of the pharyngeal 
 tube, developed in the wall of the same. 
 
Plate 11 
 
r- 
 
THE INTEGUMENT OF THE HEAD. 39 
 
 (esophagus as the axis around which they are arranged, it will be 
 readily seen that the more anterior is horizontal and surrounds the 
 base of the proboscis ; its anterior extremity forming the boundary 
 of the oral cleft, whilst the posterior is vertical, and is arranged 
 around the cephalo-thoracic opening. The fourth segment bears 
 the antennae, and the fifth the ocelli ; the great eyes also seem to 
 " 3long to this segment, although externally they appear between 
 tne two ; hence I shall call the fourth the antennal, and the 
 fifth the optic segment, naming their dorsal plates respectively, 
 from the positions they occupy, the facial and occipito- frontal 
 plates. 
 
 Each of the plates which form the head will need a separate description, 
 and as the dorsal plate of the fifth segment, the occipito frontal, is the 
 centre around which the others are all arranged, it presents the fittest 
 starting point : it is colored in figs. 1 and 2, Plate II. It commences at 
 the superior and lateral margins of the occipital foramen, forms the 
 central portion of the occipital region, and then bends forwards and down- 
 wards, forming the forehead, and divides into two lateral processes, which 
 bound the lower margins of the great eyes and terminate by becoming 
 intimately blended with the cheeks. The line of division is indicated 
 only by a more or less perceptible ridge, but it is well marked in several 
 of the Syrphidae. I do not think, however, that the blending of the two 
 segments in this region would afford the slightest argument against their 
 distinctness, as it is in the dorsal region that we first find the segments 
 become blended with each other, and as already stated, the two correspond 
 ing segments 'in the larva ere united into one above. It is true that the 
 anterior edge of this plate is only loosely attached to the dorsal plate of 
 the fourth segment ( the facial plate ) by a membrane, but then this 
 membrane, as will hereafter appear, serves a special purpose, and the 
 union of the dorsal plate of the fifth with the lateral plates of the fourth 
 segment, gives the chief strength to the anterior portion of the head. 
 On each side of the occipital foramen the occipito-frontal plate is hollowed 
 out into a cnp-shaped depression to receive the apex of the conical condyle, 
 which steadies the head, and which consists of the lateral plate of this 
 segment. 
 
 Below the occipital foramen is the basilar plate, the ventral or inferior 
 portion of the fourth segment ; it bears a pair of internal processes, which 
 support the cephalic nerve-centre and the main tracheae of the head. 
 
 The lateral plates of the fourth segment are very large, as they compose 
 the greater part of the occipital region, and bend forwards to form the 
 cheeks, where they bear a number of strong hairs. 
 
40 GN THE ANATOMY OF THE i'L^ . 
 
 The dorsal plate of the fourth segment, the facial plate, bears the antennas 
 at its superior or curved margin, which corresponds with a concavity in the 
 anterior margin of the occipito-frontal plate ; inferiorly it is continuous 
 with the proboscis ; its inferior angles are united to the cheeks, and the 
 remainder of its margin is continuous with the membrane already referred 
 to, which attaches it loosely to the occipito-frontal plate. 
 
 When the perfect fly emerges from the pupa, a large pro- 
 tuberance occupies the front of the head immediately above the an- 
 tennse. It consists of a membrane, by the dilatation of which the 
 insect bursts off the anterior extremity of the pupa case ; when 
 fully dilated it is larger than the head, which is as yet retained 
 within the pupa ; but as soon as the latter escapes, this protuber- 
 ance rapidly collapses and is withdrawn into it. Until the integu- 
 ments are thoroughly hardened it may be made to protrude 
 from the head by gentle pressure, but is not afterwards 
 visible externally ; by pressing back the facial plate, however, 
 a fissure is opened between it and the occipito-frontal plate and 
 cheeks, by which the cavity formed by the retraction of the 
 membrane communicates freely with the external air. The 
 membrane is continuous with the integument at the margins of 
 the fissure ; it assumes a bright yellow color in the mature insect, 
 and is thrown into numerous plications ; its connections are best 
 seen from the interior of the head. The extent of its surface is 
 very great, and is much increased by being closely covered with 
 papillse. It has probably three distinct functions : first, it serves 
 to break off the anterior extremity of the pupa case ; secondly, it is 
 undoubtedly connected with the humming of the insect, the 
 facial plate being thrown into rapid vibrations during the 
 emission of sound; and lastly, it seems to receive numerous nerve 
 filaments from the antennal nerves, and is probably an accessory 
 organ of smell. From its position I propose the term frontal 
 sac to designate it. 
 
THE 
 
 Section II. The Proboscis. 
 PLATE II. Figs. 315. PLATE III. 
 
 The proboscis is one of the most remarkable and complex 
 structures found in the insect world ; it consists of three joints, 
 each of which represents a distinct segment. Its hard parts are 
 the homologues of those usually found in the mouths of insects, 
 and although they are greatly modified to serve special purposes ? 
 yet, owing to the elongated nature of the organ, the several 
 segments are perhaps more decidedly distinct in the fly than in 
 any other insect. 
 
 The integument of the basal joint consists chiefly of a funnel- 
 shaped membrane continuous with the antennal segment around 
 the oral cleft ; a semi-transparent trapezoidal plate of chitiue, 
 the epistoma, which forms a hinge with the facial plate, is devel- 
 oped upon its dorsal surface, * (it is represented in outline in 
 Plate II. Figs. 11 and 12). Connected with the inferior edge of 
 this and almost entirely enclosed within the funnel-shaped 
 integument, a very complicated organ is found ; it is the largest 
 piece in the proboscis, and forms a fulcrum upon which all the 
 other parts move. It gives attachment to five pairs of muscles 
 and guides all the movements of the basal joint, which is other- 
 wise membranous. The walls of this organ are double, and the two 
 plates are kept in apposition by their elasticity, but are capable 
 of being separated and of forming a cavity by the action of a 
 pair of large muscles, the cavity so produced being continuous 
 
 * In order to avoid repetitions, it is to be understood that the words 
 dorral and ventral, as applied to the proboscis, relate to the surfaces which 
 assume those positions when the organ is extended and drawn forward ; 
 anterior and posterior likewise refer to the same position of the organ, 
 that which it would assume if the oesophagus were straight instead of ci;rvcd 
 in the head. 
 
42 OX THE ANATOMY OF THE FLY. 
 
 anteriorly with the mouth, and posteriorly with the oesophagus. 
 The dilatation of this cavity draws fluid from the mouth, and its 
 obliteration which is entirely effected by the elasticity of its walls, 
 injects it into the oesophagus and sucking stomach or crop, the 
 absence of valves being made up for by the peculiar rythmic 
 action of its muscles, the anterior fibres contracting and again 
 relaxing before the posterior, so that an undulating motion is 
 given to the superior wall of the cavity. Although I believe this 
 organ is homologous to the fulcrum in bees, I prefer the term 
 pharynx, from its double connection with the mouth and oesopha- 
 gus, as well as its peculiar function. ( It is represented in Plate II, 
 Figs, 11, 12, 13, 14 and Plate III. a.) 
 
 The pharynx or fulcrum consists of a body and four processes, and the 
 body of a dorsal and ventral arch ; the dorsal arch is continuous with the 
 integument, and is united above with the trapezoidal plate, which I have 
 called the epistoma, of which it probably forms a part ; the ventral arch 
 is more properly pharyngeal, being probably developed in the walls of the 
 alimentary canal and not belonging to the integument ; it consists of two 
 layers, between which the cavity connecting the mouth and the oesophagus 
 is situated. The superior, or dorsal layer, is thickened in the mesial line, 
 forming a rod of chitine, which extends beyond the inferior plate at its 
 anterior extremity. On either side of this a row of fine seta) project into the 
 interior of the cavity, their points being turned towards the oesophagus ; these 
 are probably sensory organs by which the act of swallowing is controlled. 
 The inferior layer is thicker and stronger than the superior, and gives 
 strength to the whole organ. The hollow between the ventral and dorsal 
 arches is occupied by a pair of sets of muscles, which arise from the inferior 
 part of the dorsal arch, and from the dorsal edges of the ventral arch, and 
 are inserted into the mesial line of the superior layer of the pharyngeal 
 tube, which they separate from the inferior by their contraction. 
 They are remarkable in containing a large number of branched mus- 
 cular fibres. 
 
 The posterior processes give insertion to the exsertor muscles of the 
 proboscis, (Plate II, Fig. 13 a) which arise from the interior of the 
 cheeks. In order to understand their action, we must consider the position 
 of the pharynx when the proboscis is retracted. It is then withdrawn into 
 the cavity of the head and lies with its posterior processes against the upper 
 part of the facial plate, and its anterior or distal end against the inferior 
 (posterior) margin of the oral cleft, so that its direction is almost 
 horizontal, the epistoma being folded back on the facial plate. The 
 contraction of the exsertor muscles draws down the pharynx and causes it 
 to revolve upon the junction of the epistoma and facial plate. 
 
THE 1'KOBOSCIS. 43 
 
 At the anterior part of the basal joint a pair of bands partially 
 surround the proboscis (Plate II. Fig. 10, and Plate III. e.) 
 These bear six or eight stiff set, which act as vibrissae or 
 feelers. The great maxillary palpi are attached to the upper part 
 of the posterior edge of the bands, the anterior edge of which is 
 continuous with a thin scale, covered with curved hairs. The 
 scales and bands are undoubtedly homologous to a part of the 
 maxilla? of other insects. The remainder of the maxilla? seems to be 
 united with the labrum and to be supported by a pair of lever-like 
 apodemes or internal processes, which lie one on either side of the 
 pharynx, between it and the bands in question. (Plate II. Fig. 7.) 
 
 I believe the basal joint of the proboscis, \vith the epistoma and 
 maxillse, form the third segment in both the larval and perfect forms of the 
 fly. In grasshoppers, the three segments, fcnned by the facial plate, 
 the epistoma and the labrum, are quite distinct in front and correspond 
 exactly to the parts I have so named. Moreover, the maxillse clearly 
 belong to a segment posterior to that which bears the mandibles, not only 
 in the fly but in other insects. This is quite plain in the cockroach. The 
 position of the labiuin, which consists most undoubtedly in great part of 
 a pair of modified limbs, indicates that it forms the most anterior segment 
 of all. 
 
 The homology of the maxillae needs careful study and comparison with 
 other insects. In bees, especially Xylophaga, these organs consist of a basal 
 sheath, which partially surrounds the pharynx and bears a comb of stiff 
 teeth, as well as the palpi. Two lobes arise from this sheath, a small 
 superior one covered with hairs, and a large sharp-pointed spathulate lobe 
 with a straight superior edge, fitting against that of the opposite side, and 
 forming the upper lip. The band in the fly represents the ba*>al sheath, and 
 the comb of bees is replaced by the setas. The thin membranous scales are 
 homologous to the small superior lobes, and the long pointed scales united 
 in the fly to the labrum are clearly the homologues of the terminal lobes in 
 bees, although they have usually been described as the mandibles. My 
 friend .Mr. C. Stewart pointed out to me that the thin scales homologous 
 to the superior lobes are united with these organs in Rhingia, an arrange- 
 ment which seems common in the Syrphida3, and which puts the matter, I 
 think, beyond a doubt. 
 
 The pharynx and maxilla? are both represented by corresponding 
 organs in the larva, (Plate II. Figs. 3 and 3, a and Z.) The 
 sucking organ , both in function and form, is too like the same 
 
44 ON THli ANATOMY OF THE FLY. 
 
 part iii tlie perfect insect to doubt its homology. The small 
 straight appendages II, appear to belong to this organ, which 
 supports them, and to represent the maxilla). I am even inclined 
 to believe the superior ventral processes of the sucking organ 
 are the homologues of their apodernes, No connection can, 
 however, be absolutely traced between these parts in the larva 
 and perfect insect, as all parts of the mouth of the former remain 
 attached to the pupa case, but I am inclined to believe this is a 
 mere shedding of integument, arid that the new organs of the 
 mouth are directly dependent for their formation upon the old, 
 like the abdominal segments and the whole digestive system.* 
 
 The second segment of the proboscis ( Plate III ) folds upon 
 the basal one with an elbow-like joint, so that their dorsal sur- 
 faces arc in contact when the organ is at rest. The maxillary 
 palpi then lie against the edges of the oral cleft, and are directed 
 
 DESCRIPTION OF PLATE III. 
 
 This plate represents the right half of a vertical section of the proboscis 
 a little to the left of the mesial line, a, The pharynx, cut through above but 
 not divided vertically. 6, The operculum, with one of its apod^mep, which 
 may be traced behind the pharynx, c, The labium, which forms the floor 
 of the mouth and is continuous above with the salivary tube, rf, Half 
 the mentum. e, Band supporting the scale-like lobe of the maxilla, 
 several setae, and the maxillary palpus. /, Valve in the salivary duct. 
 g, Great tracheal sac giving off the main tracheal vessel, which supplies 
 the muscles of the canula and lobes. ^, The lobes, with some of the hex- 
 agonal pigment cells shown as well as the spherical cells, to which the 
 greater portion of the main nerve which accompanies the principal 
 tracheal vessel is distributed, z, Chitinous band surrounding the trian- 
 gular opening of the mouth, which supports the false trachea? and the 
 teeth ; its anterior extremities bear a pair of curved hooks, probably the 
 homologues of the mandibles. A group of spherical cells, immediately 
 connected with a set of recurved hairs, is seen above this piece, with a 
 branch of the main nerve distributed to them. 
 
 * Wcismann asserts ( Koll. Zeitschrift, Band. XIV. ) that the salivary 
 glands of the perfect fly are new formations, and that he once found the 
 salivary gland of the larva by the side of that of the perfect insect, in the 
 same imago ; I think, however, the malformation must be otherwise 
 accounted for. Sec Section III. page 51. 
 
ru-m 
 
 / 
 
THE PROBOSCIS. -15 
 
 forwards, occupying exactly the same position as in the Lep- 
 doptera. Whatever., their function may be, it is probably 
 called into activity when the proboscis is retracted, for they 
 are turned back when it is exserted ; they are probably connected 
 with the sense of taste or smell.* 
 
 The second joint consists of two portions, one dorsal and the 
 other ventral, the former being the roof, and the latter the floor 
 of the cavity of the month ; so that they might well be named 
 lips, if the term had not been unfortunately applied to the ter- 
 minal lobes of the proboscis ; especially as the upper consists of 
 the labriun and terminal lobes of the maxilla?, and the inferior of 
 he united labium and inenturn. I shall call them respectively 
 the operculum, and the canula. 
 
 The canula is thick and fleshy, being filled with muscles ; it 
 is terminated by the lobes of the proboscis, and is deeply grooved 
 on its dorsal surface, the groove forming the mouth, and 
 concealing the tongue. The edges of this groove overlap the 
 operculum, and completely cover its anterior extremity, meeting 
 in front of it, so that it is held firmly in its place, but by a 
 little violence it can be made t o start upwards and assume a 
 position at an angle with the canula, without any rupture of 
 membrane. It never leaves the groove, however, during life, 
 although in those diptera in which the maxillso, mandibles, and 
 tongue form lancets, the canula is capable of being folded 
 back, to bring the lancet into play. 
 
 The operculum, Plate II, Fig. 7, consists of a central and two lateral 
 portions. 
 
 The labrum or central part is tubular behind, and is united by a mem- 
 branous tube to the anterior extremity of the cavity of the pharynx, a 
 small semicircular piece of chitine protecting its bend ; the food passes from 
 the mouth by this means into the pharynx. The anterior two- thirds of 
 the labrum is open below and forms the roof of the mouth. 
 
 The homologies of the lateral portions of the operculum have already been 
 discussed.f They are firmly attached to the back of the labrum and form, 
 
 See Section IV. page 55, and XV. page 93. f Page 43. 
 
46 ON THE ANATOMY OF THE FLY. 
 
 joints with their apodSmes, which give insertion to two pairs of muscles 
 both arising from the pharynx. Plate II, Fig. 13, shows the points of origin 
 at c and d ; both are inserted into the free extremity of the apodeme ; they 
 are probably flexors of the second upon the first joint of the proboscis. 
 They act by drawing the head of the apodeme towards the dorsum 
 and pressing it forward ; as the apode'mes act upon the operculum 
 slightly behind the line of flexion, they press its superior extremity down- 
 wards and forwards, causing its anterior extremity to revolve in the 
 opposite direction. The whole of the interior of the operculum is lined 
 with epithelial cells containing bright red pigment. 
 
 The canula consists below of the rnentum, a shield-like plate bearing a 
 pair of processes at its anterior extremity, which keep the lobes of the 
 proboscis closed by their elasticity ; a small triangular plate unites 
 them inferiorly. The whole superior surface of the mentum gives 
 origin to muscles. The largest pair being inserted into the terminal pro- 
 cesses serve to open the lips. A pair of thin muscular slips arising from 
 the margin of the occipital foramen is inserted into the posterior margin 
 of the mentum. The latter are retractors of the whole proboscis as well as 
 extensors of the second joint. The edges of the mentum are everywhere 
 united to the membranous sheath of the proboscis. 
 
 Tracing the membranous covering of the canula from the edges 
 of the mentum, it is found to become chitinous in the groove on 
 its upper surface and round the anterior extremity of the mouth, 
 forming the base of the great lobes ; these parts represent the 
 labium or lower lip of which the last named are lateral appen- 
 dages. Perhaps the labium is the most difficult organ to under- 
 stand in the mouth of insects. It consists of five distinct parts, if 
 the tongue be considered as a portion of it, and these seem to be 
 very constant. They are 1 , the floor of the mouth united to the 
 upper side of the mentum ; 2, the tongue united to the posterior 
 extremity of the floor of the mouth, but more or less free in front ; 
 3, a plate attached to the anterior border of the floor of the mouth 
 in front of the mentum ; 4, a pair of palpiborne by the base of the 
 last part on its inferior surface which are turned downwards ; 5, a 
 pair of lobes turned upwards and borne by the extremity of the same 
 plate : these are the modified limbs which belong to the most an- 
 terior segment. All these parts may be distinguished in the labium 
 of the fly, except the palpi. Much confusion exists as to their 
 nomenclature especially between the last-named and the tongue, 
 both having been called the ligula I shall call them respectively 
 
THE PROBOSCIS. 47 
 
 the posterior labial plate, the tongue, the anterior labial plate, 
 and the labial appendages or lobes. 
 
 The posterior plate of the labium, ( Plate II, Fig, 6, c, and Plate III, c,) 
 replaces the membrane at the bottom of the groove in the canula. It forms 
 the sides and floor of the mouth. Two lateral rods are interposed between 
 it and the operculum anteriorly ; but behind, it tits against the oper- 
 culum, as far back as that part is incomplete. Where the operculum becomes 
 tubular the labium also becomes closed above, and is continued backward 
 united with the tongue, as a tube which joins the salivary duct. The 
 tongue, ( Plate II, Figs. G and 6a,) is tubular behind and united to the 
 Lujium, but becomes thin and lancet-shaped in its anterior third, before 
 the opening of the salivary tube. Its free extremity is covered with fine 
 setae. 
 
 The posterior plate of the labium is probably part of the mental or 
 second segment. Its inferior surface has a number of muscular fibres in- 
 serted into it, which arise from the mesial line of the mentum. On either 
 side of the posterior plate of the labium is a rod, (Plate II, Fig. 6 b,) united 
 to it in the mature fly, but quite distinct when the insect first 
 emerges from the pupa. These rods support a pair of hooks, (Plate II, Figs. 
 5 and 5a,) which, on comparison with the larval hooks (shown in Figs. 4 
 and 4a in the same plate,) will be found to correspond very nearly to 
 those organs. I am inclined to believe they are the homologues of the 
 mandibles, and therefore the lateral appendages of this segment. The 
 larval hooks are supported by anH-shaped organ, (Fig, 4 c.) which seem 
 to represent the second segment, and I have found the same rods and 
 hooks in all the Diptera which I have examined, in which the larva ai e 
 furnished with hooks. Like the maxillze and labial lobes they are placed 
 on the dorsal aspect, and in front of the segment to which they belong. 
 There is no other representative of the mandibles. 
 
 The third segment of the proboscis (Plate III, A. .,) consists of 
 the margin of the anterior opening of the mouth, which is trian- 
 gular, and of the large fleshy lobes at its extremity. These when at 
 rest are folded against each other, but when they are opened, 
 about two-thirds of their anterior surface forms an oval sucker 
 divided into two parts by the fissure between them. Behind this 
 fissure the remainder of the lobes forms a cavity bounded pos- 
 teriorly by the triangular opening of the mouth. 
 
 The anterior surface of the lobes is channelled by a series of 
 canals, kept open by incomplete rings, called false tracheae, 
 which open internally into the cavity between the lobes and so 
 into the mouth ; these form a fine strainer through which the in- 
 
48 OX THE ANATOMY OF THE FLY. 
 
 sect is enabled to filter the fluid from the solid portion of the 
 substances on which it feeds. The lobes are, however, capable of 
 further separation, exposing the triangular opening which is sur- 
 rounded by from fifty to sixty bidentate rods or teeth, which are 
 usually concealed between the posterior portions of the lobes but 
 are used, when exposed, for grinding hard substances, such as 
 sugar, so assisting the salivary secretion to dissolve them- 4 . Fluids 
 are often drawn directly into the mouth, either by the fissure 
 between the lobes when it is partially closed, or even when it is 
 fully opened to expose the teeth, but they more usually pass 
 through the false trachere. This is best demonstrated by giving 
 the fly a little blood and afterwards examining the proboscis, 
 when all these channels will be found filled with it. 
 
 The parts forming the anterior triangular opening of the mouth are 
 represented in Plate II, Figs. 5 and 5a ; they are a dark solid mass of 
 chitine and two lateral plates, which probably represent the anterior 
 labial plate and the basal portion of its lateral appendages ; they bear the 
 lobes and bideutate rods or teeth, already mentioned, 
 
 The false tracheas, (Plate IV, Figs. 2 and 3,) are channels on the anterior 
 surface of the lobes, kept open by semi-rings, which are developed from 
 cells of the protoderm. These rings are bifurcated at one extremity, and 
 the bifurcated extremity is alternate in the several ringsf. The channels 
 become gradually smaller towards the edge of the disc and terminate in 
 closed extremities. Their number is variable and not always the same of 
 the two sides of the proboscis ; there are generally twenty-nine or thirty 
 on each side. They usually open on the inner edge of the lobes in the fol- 
 lowing order. The five anterior channels unite and form a large trunk, 
 which runs along the upper border of each lobe and opens close to the upper 
 margin of the triangular opening ; the next seven or eight open separately, 
 each externally to a set of teeth, and the remainder run into a pair of 
 parallel channels in the middle of the disc below the opening of the 
 mouth, which open at the lower part of the triangular opening. Between 
 the margin of the triangular opening and the orifices of the false tracheal 
 channels, a number of curved chitinous pillars are found radiating from the 
 gides of the triangular opening to these orifices ; they arc distinctly seen 
 to represent the semi-rings of 4he channels themselves flattened out. 
 
 *f Some very interesting information on this subject with good figures of 
 the details of the false trachere, is given by Mr. W. T. Suffolk, in a paper 
 read before the Eoyal Micros. Soc. Ap, 14. 1869. 
 
 * Described by Mr. Hunt, Microscopical Quarterly Journal, 1856, p. 238. 
 
THE PROBOSCIS. 49 
 
 Between each pair of false tracheae the surface of the lobes is marked by 
 a pair of wavy lines ; these are folds of the integument, and disappear 
 when the membrane between the false tracheal channels is stretched. Four 
 or five nipple-shaped papillae project from the surface, between each pair 
 of wavy lines, and nerve filaments terminate in them ; they are probably 
 organs of taste. The disposition of these parts is shown in section, in 
 Plate IV, Fig. 3. 
 
 A row of cells around the margin of the lobes are hardened into little 
 square plates ; these are developed in the membranous protoderm ; several 
 other, but less developed, rows of a similar nature extend towards the 
 back of the lobes. 
 
 The mesoderm consists of beautiful nucleated pigment cells, filled with 
 a bright orange-colored pigment ; in the mature insect these are united into 
 a continuous membrane ; their nuclei become apparent on the addition 
 of a little acetic acid. 
 
 The back of the lobes is covered with hairs, which form a fringe aroun^ 
 its disc. The endoderm consists of numerous spherical cells, containing 
 granular nuclei, and the bulbs of these hairs, with a plexus of nerve loop g 
 distributed to them, from the terminal branches of the great nerve of th e 
 proboscis. A group of recurved hairs with a similar plexus of nerve loops 
 is situated on each side of the second joint, near its anterior extremity. 
 
 The false tracheal channels are partially surrounded at their orifices by 
 the curved extremities of the chitinous pillars which support them upon 
 the anterior border of the mouth. Closely connected with the points of 
 these, and immediately beneath the integument, is a tendinous looking 
 band of dense tissue, which extends the whole length of the fissure between 
 the lobes of the proboscis. A round cord of similar tissue arises partly 
 from this band, and partly from the point of the chitinous pillar, on either 
 side of each false tracheal channel. The pair which arise from each pair 
 of points unite and form a thicker round cord, which runs outwards towards 
 the margin of the lobe, near which it is inserted into a thickened ridge of 
 chitine, connected with the anterior extremity of the anterior process 
 of the mentum, at right angles to it. The purpose of these cords 
 seems to be to connect the extremity of the anterior processes of the 
 mentum with the margins of the fissure between the lobes, which is pro- 
 ably kept closed by the elasticity of the pillars which support the orifices 
 of the false tracheae, and which is opened by the muscles attached to the 
 anterior processes of the mentum, through the agency of these tendinous 
 bands. If these processes of the mentum acted upon the margin of the 
 fissure by drawing upon the integument of the lobe, the margin of the 
 sucker, formed by the union of the lobes, would have been disturbed in 
 order to open the fissure; whilst, by means of these bands, the anterior 
 process of the mentum not reaching the margin of the lobe, acts directly 
 
50 ON THE ANATOMY OF THE FLY. 
 
 upon the fissure, without impairing the action of the sucker or altering 
 the position of its edge. These bands are dissolved by liquor potassse, 
 they exhibit a dark outline ; do not swell up when treated with acetic acid 
 and are elastic. They are developed from spindle-shaped nucleated cellsj 
 which may be found in great numbers in the lobes of the proboscis of the 
 immature fly. 
 
 There is no better method of examining the proboscis, in 
 order to obtain a general idea of its structure and functions, than 
 by observing that of a living fly whilst in action. This may be 
 conveniently done by enclosing the insect in a live-box, with a 
 few spots of syrup on the thin glass cover, and illuminating it 
 from above. This method was first suggested by the late 
 Richard Beck. Another very good way of examining the entire 
 organ, is to take a recently killed fly ; exsert the proboscis by 
 pressing on the thorax, and tie its base to keep it distended. 
 It may then be examined dry, or mounted in glycerine, in which 
 it becomes a very beautiful object, when it has soaked in that 
 medium a sufficient time to become transparent, especially if 
 illuminated by a black spot lens, as recommended by Mr. 
 Suffolk. This method is subject to the grave objection that 
 specimens so obtained are unnaturally tumid, and assume an 
 appearance quite unlike that natural to the organ. Sections of 
 the proboscis made with a lancet, or the entire organ cut off 
 and mounted in glycerine, are the most instructive specimens, 
 and are the only means of making out the minute anatomy of 
 the part. Those required to show the distribution of nerves 
 should first be hardened in spirit, or chromic acid. Numerous 
 recent sections should be examined as well as mounted specimens. 
 Mr. Topping's well-known and beautiful preparation only 
 exhibits the chitiiious parts, and the rneiitum is cut away in 
 order that the lobes may lie flat : although most useful as a 
 means of examining the structure of the false tracheal channels 
 and the teeth, the relations of the parts are so altered by 
 pressure, that some cannot be made out at all. All the hard 
 parts may be seen in situ by treating the whole head, with the 
 extended proboscis, in the manner already recommended fo* 
 examining the integument of the head ; but it is very difficult 
 
THE SALIVARY GLANDS AND DUCTS. 51 
 
 to mount a perfect specimen, as the lobes of the proboscis 
 usually collapse. 
 
 Section III. The Salivary Glands and Ducts. 
 PLATE III. /. PLATE IV. Figs. 1 and 6. 
 
 The salivary glands consist of a pair of long convoluted tubes 
 which lie one on either side of the chyle stomach in the thorax 
 and extend into the abdomen, terminating near its apex in blind 
 extremities. (They are shown in situ in Plate I, and in their 
 relation to the alimentary canal in Plate IV, Fig. 1.) Each gland 
 tube opens near the, posterior portion of the thorax into an 
 elongated sac, which its convolutions closely surround. This sac 
 is lined with delicate pavement epithelium, and terminates 
 anteriorly in the duct. The salivary ducts converge and unite 
 in the thorax just behind the cephalo-thoracic opening, forming 
 the common salivary duct. This enters the head beneath the 
 oesophagus, which it immediately leaves, and passes between the 
 pharynx and membranous integument of the basal joint of the 
 proboscis, finally opening on the upper surface of the tongue. 
 
 The salivary glands of the larva consist of elongated sacs, lined 
 with coarse pavement epithelium. I think there can be little doubt 
 but that the sacs of the glands in the perfect insect are dependent for 
 their formation on the salivary glands of the larva, the epithelial coat 
 having alone under-gone change. Dr. Weismann asserts that he once 
 found a fly in which he detected the larval salivary gland by the side 
 of that of the imago. I can only understand his assertion by sup- 
 posing the malformation in question was due to the salivary apparatus 
 of the larva having been double in that specimen, only one of the 
 glands having undergone the usual changes. I once found the sacs 
 of the salivary glands of the adult fly lined with the same epithelium 
 as the larval salivary gland, but only once. 
 
 The tubular portion of the gland in the fly is tilled with gland cells ; 
 and its external coat is covered with a plexus of tracheal vessels, (a por- 
 tion is represented in Plate IV Fig. 6 ). The whole length of each tube 
 if unwound would exceed two inches, or nearly four times the length 
 
 E 2 
 
52 OX THE ANATOMY OF THE FLY. 
 
 of the entire insect. The gland cells are 1- 1000th of an inch in 
 diameter. 
 
 The salivary ducts and common salivary duct are kept open 
 by a spiral fibre, developed in their internal coat. The external 
 coat is yellowish, thick, and thrown into rugrc, except when the 
 ducts are stretched. The ducts are exceedingly elastic and exten- 
 sile. Just above the labium the common duct changes its 
 character, losing its spiral fibre and extensibility. It becomes 
 perfectly transparent and dilates into a cavity somewhat resem- 
 bling the human glottis. This is a most perfect valve ; for its 
 superior* wall is elastic and pressed down at its anterior portion 
 against the inferior wall, so entirely closing the duct at this 
 point. A paii* of long slender muscles, arising at the posterior 
 extremity of the pharynx, run beneath that organ, and are in- 
 serted into the anterior part of this valve, opening it by their 
 contraction. 
 
 It will be easily understood that the sucker-like action of the 
 lobes as well as the opening of the phaiyiigeal cavity, would tend 
 to exhaust the mouth of ah', and so to cause a flow of salivary fluid 
 into its cavity, which would continue as long as the proboscis 
 remained in action, were it not for this valvular arrangement in the 
 salivary duct. The opening and shutting of the valve is controlled 
 by a ganglion placed in front of and beneath it. Nerves from 
 this ganglion are distributed to the mouth, as well as to the long 
 muscles which act upon the valve. Any dryness of the epithelial 
 lining of the mouth, in which these nerves ramify, would un- 
 doubtedly communicate an impression, and cause a reflex act by 
 which the valve muscles would contract, and so cause a flow of 
 saliva by opening the valve. 
 
 * See note to page 41. These relations may be observed in Plate III 
 but the lower portion of the plate, it must be remembered, represents 
 the anterior extremity of the proboscis, and the left side its dorsal or 
 upper surface. 
 
THE ALIMENTARY CANAL AND ITS APPENDAGES. 53 
 
 Section IV. The Alimentary Canal and its Appendages. 
 PLJLTE IV. Figs 18. 
 
 The o?>ophagus commences at the posterior extremity of the 
 pharynx, and curving sharply backwards, pierces the cephalic 
 ganglion and passes through the cephalo-thoracic opening, above 
 the great nerve trunk, and beneath the dorsal vessel. As soon 
 as it enters the thorax it divides into two branches of equal 
 diameter. One passes upwards and immediately opens into the 
 proventriculus ; the other runs the whole length of the thorax, 
 beneath the chyle stomach, and passes into the abdomen under 
 the intestine, above the great nerve, and between the salivary 
 glands, to all of which it is closely bound by a plexus of 
 tracheiil vessels, in the meshes of which the nerve cells of the 
 great sympathetic ganglion lie. As soon as it enters the abdo- 
 men it opens into the sucking stomach or crop. 
 
 The oesophagus may be said to have four coats ; the most external of 
 circular and the second of longitudinal muscular fibres ; the third con- 
 sists of membrane insoluble in liquor potassse; and the fourth of 
 pavement epithelium. The membranous coat exists throughout the 
 whole alimentary canal, and bears the epithelial lining ; like the 
 protoderm it is continuous, and insoluble in hot solutions of caustic 
 potash. 1 have no doubt that a new layer is formed in the pupa on the 
 exterior of the old, which is shed like the integument of the larva, but 
 this has not been demonstrated; it differs from the protoderm in develop- 
 ing epithelial cells upon its free surface, but I believe it is continuous 
 with it. 
 
 The sucking stomach consists of a pair of almost hemis- 
 pherical sacs, united anteriorly in the mesial line ; it serves as a 
 reservoir for the food when first taken. There is no organ 
 homologous to this stomach in the larva, but it is developed 
 from a bud which appears in the pupa on the under side of the 
 oesophagus. The organ in the maggot which performs the same 
 function is a large flask-shaped sac, which opens directly into 
 the pharynx, and the oesophagus commences from the under 
 side of its neck. The neck of this sac is not surrounded by the 
 great ganglia, but is anterior to them ; and the sac is dorsal to 
 all the other organs in the larva. When the maggot is feeding 
 
54 ON THE ANATOMY OF THE FLY. 
 
 it appears upon the back as a dark egg-shaped mass ; it shrivels 
 gradually after the larva ceases to feed, and disappears entirely 
 before the third day of the pupa state ; I shall call it the 
 pharyngeal pouch. 
 
 In the sucking stomach the external or muscular coat consists of 
 interlacing muscular bands, which terminate in pointed, and often in 
 
 DESCRIPTION OF PLATE IV. 
 
 Fig. 1. The alimentary canal, as it appear after the insect has been 
 fed on food mixed with cochineal. The oesophagus and crop are filled 
 with fluid, coloured with the pigment. In the chyle stomach the 
 food assumes a violet colour ; the salivary glands remain uncoloured 
 and present a transparent blueish tint. The bile tubes are only 
 partly shown, with their ducts filled with yellow bile pigment, and 
 the rectal papilla are deeply stained with cochineal. X 12 diam. 
 
 Fig. 2. A portion of one of the false tracheal channels of the pro- 
 boscis ; the protoderm is torn away from the edge of the channel at its 
 right extremity, to show the manner in which it is folded, x 500 diam. 
 
 Fig. 3. A section through a portion of the surface of one of the lobes 
 of the proboscis ; showing a false tracheal channel cut through trans- 
 versely, and the folds of membrane between it and the next channel 
 with one of the nipple-shaped papillae between them, x 500 diam. 
 
 Fig. 4. Side view of the anterior extremity of the chyle stomach ; 
 showing the proventriculus and the sacculated walls of the chyb 
 stomach, x 36 diam. 
 
 Fig. 5. Cylinder epithelium from the chyle stomach, x 250 diam. 
 
 Fig. 6. A portion of the salivary glands, to show the ramification of 
 the tracheal vessels and the disposition of the gland cells, x 250 diam 
 
 Fig. 7. A portion of one of the Malphigian or liver tubes, x 250 
 diam. 
 
 ^Fig. 8. The valve between the large intestine and reclnra. x 36 
 diam. 
 
 Fig. 9. A vertical section of one of the rectal papillce or renal 
 organs, x 60 diam. 
 
 Mg. 10. Gland cells from the same, x 250 diam. 
 
 Fig. 11. The apex of the same, x 250 diam. 
 
 Fig. 12. A transverse section of the same, x 60 diam. 
 

THE ALIMENTARY CANAL AND ITS APPENDAGES. 55 
 
 branched extremities, This coat is thickest near the opening of the 
 oesophagus, and the fibres of that portion of the organ which lie in the 
 mesial line are mostly circular. Numerous nerve cells belonging to the 
 sympathetic system are scattered over the external coat, and form a 
 pair of large patches, near the entrance of the oesophagus. The cavity 
 of the sucking stomach is lined with pavement epithelium, and 
 frequently contains a beautiful mould, allied to Botrytis, which often 
 causes the death of the insect by its rapid development. 
 
 The proventriculus is a small globular cavity at the commence- 
 ment of the chyle stomach ( Plate IV, Fig. 4,) ; it has thick 
 glandular walls, and secretes the gastric fluid ; it opens 
 directly into the chyle stomach, where absorption commences. 
 The epithelium of the proventriculus consists of several layers 
 of clear nucleated cells, those upon the surface being angular 
 and closely adherent to each other ; the deeper layers are oval, 
 and the most external float freely in fluid. They difier remark- 
 ably from the epithelium of the chyle stomach, which consists of 
 a single layer of conical cells. This is especially noticeable in the 
 larva, where the epithelium of the chyle stomach is opaque with 
 granular matter : whilst the gland structure of the proventricu- 
 lus is perfectly transparent. The proventriculus of the larva is 
 larger and more oval than that of the imago. 
 
 The proventriculus may be said to consist of an intussusception of 
 the oesophagus, the muscular coat disappearing in the middle of the 
 organ, and the membranous coat being reflected back upon it and 
 swelling out to make room for the thick glandular lining. It is again 
 constricted at the commencement of the chyle stomach by a muscular 
 sphincter. The tracheal vessels, which ramify in a close net work over 
 the membranous coat, dip down into the interior with the oesophagus ; 
 and then take a radiating direction. It has a larger supply of tracheas 
 than any other part of the alimentary canal. 
 
 The analogy between the functions of the crop proventriculus 
 and chyle stomach, and the crop rennet and paunch of Ruminants 
 is too obvious to need enlarging upon. But the resemblance of 
 the digestive act in the two cases is even closer, for the fly 
 frequently brings its food back into the mouth, probably for the 
 purpose of mingling it with the saliva. I have repeatedly 
 observed flies exude a large drop of fluid from the oral cleft 
 
56 ON THE ANATOMY OF THE FLY. 
 
 without exserting the proboscis, and gradually reimbibe it. As 
 such a drop entirely surrounds the palpi, I think it not unlikely 
 they are the organs of sensation concerned in this act, and per- 
 haps the seat of the special sense of taste. In the larva the food 
 passes directly from the pharyngeal pouch into the proventri- 
 culus. 
 
 The chyle stomach extends the whole length of the thorax ; 
 it is divided by reticulations of circular and longitudinal bands 
 of muscular fibres, between which the membranous coat bulges, 
 so that its inner surface is covered with little square pits. It is 
 lined with conical cylinder epithelium. It is much longer in the 
 larva than in the perfect insect, and four ca)cal glands, about half 
 its length, open into its anterior extremity. These disappear in 
 the pupa,<and are not redeveloped in the imago. 
 
 The small intestine extends from the stomach to the opening 
 of the bile tubes, about three quarters of an inch ; its upper 
 end is like the chyle stomach, which passes insensibly into it, 
 but it soon looses its reticulated structure, and its muscular coat 
 becomes uniform and thin. Both circular and longitudinal 
 muscular fibres may be traced in it.* In the larva the bile tubes 
 open at the posterior extremity of the stomach, and the small 
 intestine first appears in the pupa. 
 
 The bile tubes ( Plate IV, Figs. 1 and 7 ) commence in blind 
 extremities, and unite in pairs, until they open into the in- 
 testine, by two large ducts. They are very numerous, and occupy 
 the dorsal portion of the abdomen ; they have a beaded outline 
 and contain irregular nucleated cells, yellow pigment, and fat 
 granules. I have found occasionaly dumb bell crystals in the 
 bile tubes, but have failed to detect uric acid in them, either in 
 the larva or imago. 
 
 * Its peristaltic action is very vigorous and may be observed plainly 
 in the smaller house fly, the sides of the abdomen of which are very 
 transparent. It may be seen also in the immature Blow-fly. 
 
THE BECTAL PAPILLAE. 5? 
 
 The large intestine is about four lines in length in the perfect 
 insect, but exceeds an inch in the larva ; it extends from the 
 openings of the bile tubes to a double muscular sphincter or 
 valve, which separates it from the rectum. It differs considerably 
 in structure from the small intestine. Its muscular coat, especially 
 the longitudinal fibres, is much more largely developed and its 
 middle or membranous coat is covered on its internal surface with 
 a number of sharp spines, with broad bases, which project down- 
 wards, and somewhat resemble narrow shark's teeth. These 
 processes become more blunt at the valve, between the large 
 intestine and the rectum, and extend into the upper portion of 
 that intestine; they afford an additional argument in favor of the 
 membranous coat being continous with the protoderni. Both the 
 large intestine and the rectum are lined with ordinary mucous 
 tesselated epithelium. 
 
 The rectum dilates into a pyriform cavity, just below the 
 recto-colic valve, which contains the rectal papillae. Below this 
 dilatation it becomes again constricted, and a well developed 
 sphincter of circular fibres may be observed close to its outlet. 
 It opens at the upper part of the last joint of the ovipositor of 
 the female at its extremity, and in the male it terminates above 
 and behind the penis. 
 
 The viscera may be carefully examined in situ, by partially 
 imbedding the insect in wax, and removing those parts of the 
 integument which are not imbedded, with a pair of fine scissors. 
 
 Section V. The Rectal Papilla. 
 PLATE IV. Figs 9 to 12. 
 
 The organs for which I have retained the name given to them 
 by Weismann, are four in number, situated near the termination 
 
58 ON THE ANATOMY OF THE FLY. 
 
 of the alimentary canal. They are hollow, conical, glandular or- 
 gans, about l-30th of an inch in length, enclosed in the dilated 
 portion of the rectum, and having their bases only external to its 
 cavity. Their function is probably the excretion of a urinary 
 fluid. 
 
 Each papilla consists of three parts ( Plate IV, Figs. 9 and 12 ) : an 
 internal central cavity, surrounded by a transparent structureless mem- 
 brane ; around this a hollow cone of gland cells, the secreting portion 
 of the organ, is disposed; and external to this again, a tough transparent 
 cone of membrane, Avhich 3 shall call the calyx of the papilla, perforated 
 by numerous minute pores, 'surrounds the whole of that portion of the 
 organ -which is internal to the rectum. By a little dexterous manipula- 
 tion these parts may be separated completely from each other, or, by 
 hardening the papillae in chromic acid, sections may be made showing 
 the relations of the parts in situ. Fig. 9 represents such a section. 
 
 The calyx is perforated by about 300 minute pores, each pore being 
 surrounded by a nipple-like projection, which is surmounted by from 
 three to eight minute setse, ( Plate IV, Fig. 11 ). The whole membrane 
 of the calyx becomes thickened towards its apex, where it has a faint, 
 yellow tint ; it is near the apex that the nipple-like projections and their 
 setas are best seen. Indeed a casual observer might overlook their 
 presence entirely at the upper portion of the calyx. 
 
 The calyx itself is marked by faint reticulations, and its margin is 
 deeply crenated. These are the only indications of structure which it 
 presents, and these seem to point to its being a fibrous membrane, 
 especially as two sets of muscular fibres arise from the crenations of its 
 margin. 
 
 These muscular fibres are, first, a set from the muscular coat of the 
 rectum, and, secondly, a layer of converging fibres which cover the whole 
 base of the papilla to within a very short distance of its centre, and 
 which apparently end in the edge of the membrane forming the 
 boundary of the central cavity, although from the extreme transparency 
 of this membrane it is nearly impossible to be certain of their insertion. 
 
 Each papilla is supplied with air by a large tracheal vessel from the 
 fifth abdominal spiracle, which divides into several generally five or six 
 large trunks before entering the papilla. The tracheae of the papilla 
 may be divided into two sets as soon as they enter the base of the organ. 
 First, from twenty to thirty radiating lateral branches run to the 
 edge of the base, and then pass over the outer surface of the glandular 
 structure to the apex of the cone, giving off numerous branches, which 
 anastomose freely, and form a fine reticulation around the gland cells } 
 the larger branches running directly towards the central cavity, and 
 forming loops by anastomosing with other similar vessels. 
 
59 THE RECTAL PAPILLAE. 
 
 The second set are the terminations of the main tracheee. After giving 
 off the lateral branches these run directly into the central cavity, where 
 they become tortuous, and anastomose with each other, giving off com- 
 paratively few small vessels in proportion to their size, and forming a 
 network which fills the central cavity, but none of their branches pierce 
 its investing membrane anywhere. Fig. 9 represents the central cavity 
 with its trachece, with a small portion of the lateral branches and their 
 terminations amongst the glandular structure. 
 
 The gland cells are large about l-300th of an inch in diameter and 
 slightly angular by mutual pressure. Each contains a granular nucleus 
 about one-third the diameter of the cell. 
 
 That these organs secrete uric acid seems probable both from 
 ^ their structure, and from the fact that uric acid is found in abun- 
 dance below the recto-colic valve, and not above it. 
 
 I believe the central cavity is continuous with the visceral 
 cavity, and that it affords a means of bringing the circulating 
 fluid into almost immediate contact with the secreting cells, a 
 fine structure -less membrane only being interposed and still 
 further I believe I am justified in asserting that the circulating 
 fluid is expelled from, and a fresh supply is drawn into, the cen- 
 tral cavity by a rhythmic muscular act. 
 
 In the female fly, the rectal papillae lie fortunately between the 
 second and third rings of the ovipositor when that organ is ex- 
 serted, where it is sufficiently transparent to allow of the papillae 
 being seen during life. I have repeatedly observed a movement 
 of the kind I have described. I believe this is the explanation 
 of the radiating bands of muscle at the base of the papillae. They 
 probably open the central cavity, at the same time pressing 
 upon the contents of the papilla, and, assisted by the muscular 
 wall of the rectum, which contracts at the same instant, not 
 only expel the contents of the central cavity, but also press the 
 secreted fluid through the minute pores in the calyx into the 
 cavity of the rectum. The central cavity is probably refilled by 
 the elasticity of the calyx itself. 
 
 With regard to the urinary secretion, that which is passed by 
 the insect when it first emerges from the pupa case, is a semi-solid 
 
60 ON THE ANATOMY OF THE FLY. 
 
 mass of nearly pure uric acid. That passed afterwards is a turbid 
 fluid, sometimes almost clear, and very irritating when applied to 
 any tender part of the skin. This fluid when acidulated with 
 hydrochloric or nitric acid deposits an abundance of crystals of 
 uric acid. I believe this acid is held in solution by ammonia. 
 The excrement of the fly when heated over a lamp gives off a 
 strong urinary smell. 
 
 No trace of these organs exists in the larva, which may be ac- 
 counted for by its inactive mode of life. The bile tubes appear 
 to be its only excretory organs. 
 
 Section VI. The Integument of the Thorax. 
 PLATE V. 
 
 The integument of the thorax consists of no less than fifty-one 
 pieces, some of which are extremely complex in form. The 
 number is, however, much reduced if the smaller pieces con- 
 nected with the articulation of the wings and halteres, the 
 orifices of the respiratory organs, and the twelve forming the 
 cox8G, are omitted. As these will be described in their respect- 
 ive places, and as five of the seventeen which remain are pairs, 
 but twelve will need description here. These belong to four 
 distinct segments, the first being the ventral and lateral plates of 
 the nfth or last cephalic segment. 
 
 The ventral plate or sternum of the fifth segment forms a very 
 remarkable organ, which I have named the cephalo-sternum (Plate V, 
 Fig. 9). It consists of a saddle-shaped body with a pair of thin lobes 
 directed forwards. The body is opaque and strengthened by ridges which 
 suppoit the lobes. These are thin and almost transparent ; they are 
 about l-200th of an inch in diameter and are covered by a multitude of 
 setas, about l-600th of an inch in length, each having its base sur- 
 rounded by a thickened ring, which gives the lobes a honeycombed 
 appearance. In front of each lobe is a small leaf-like plate of opaque 
 integument. The upper surface of the lobes is in apposition with a tuft 
 of nerve filaments, which covers it, and which may be distinctly seen to 
 terminate in nerve cells. The use of this organ is at present unknown, 
 but it is probably an organ of special sense. 
 
THE INTEGUMENT OF THE THORAX. 61 
 
 The lateral plates of this segment support the head, I hare named 
 them condyles (Plate V, Fig. 9). The condyle is very hard and opaque. 
 When seen from beneath it is almost triangular, with the anterior angle 
 prolonged into a rounded protuberance, which fits into tho cup-shaped 
 depression at the side of the occipital foramen, Seen from within it has 
 an irregular form, due to the turning in of its edges. A process projects 
 from its posterior external angle. Numerous muscles are inserted into it, 
 which regulate the movements of the head. 
 
 The remainder of the thorax is composed of three segments, 
 called the Pro-thorax, the Meso-thorax, and the Meta-thorax, 
 They are all visible externally beneath, but above the whole 
 pro-thorax and meta-thorax are concealed by the meso-thorax ; 
 laterally a small part of the pro-thorax and meta-thorax only is 
 visible. Each consists of four plates, except the meso-thorax, 
 in which the lateral plate is divided vertically into two. The 
 inferior plates are called sterna, the lateral plates episterna, and 
 the dorsal plates the terga, of their respective segments. The 
 edges of each plate are turned in and form ridges and processes 
 in the inside of the thorax, and the terga of both the meso- and 
 meta-thorax are divided by similar ridges into four parts called 
 respectively from before backward, the pree-scutum, the scutum, 
 the scutellum, and the post-scutellum. 
 
 It is a fact worthy of remark that, although the tendency of 
 the coalesence of segments is, that they unite first in the dorsal 
 region, so in the dorsal region we find the greatest tendency for 
 a single segment to become split up into several distinct parts. 
 This is doubtless, in both cases, due to the fact that development 
 commences in the embryo at the ventral surface, and the number 
 of segments is there first determined by the number of distinct 
 spots from which development proceeds. As the segments grow 
 up towards the dorsum, they either split or coalesce with each 
 other. 
 
 The back of the thorax is almost entirely formed by the tergum 
 of the meso-thorax. That of the pro-thorax is reduced to little 
 more than a mere rim around the superior part of the anterior 
 
62 ON THE ANATOMY OF THE FLY. 
 
 DESCRIPTION OP PLATE V. 
 
 Fig. 1. An anterior view of the thorax with the head removed, show- 
 ing the cephalo-sternum and condyles in situ, x 8 diam. 
 
 Fig. 2. A vertical section of the thorax behind the wings, showing the 
 manner in which the wings are united to the thorax, the meso -sternal 
 ento-thorax, and para-pleura, together with the arrangement of the 
 thoracic muscles. The non-striated longitudinal muscles are distinguished 
 from the striated muscles by a difference in their tint, and the small 
 space occupied by the viscera is well seen; it is marked by a blue tint. 
 X 8 diam, 
 
 Fig. 3. The meso-thorax seen from above, showing its division into 
 four distinct portions, as well as the insertion of the wings and winglets. 
 X 6 diam. 
 
 Fig. 4. A posterior view of the thorax, the abdomen being removed 
 showing the meta-thoracic tergum a, divided like the meso-thoracic 
 tergum into four parts ; i, is the ventral portion of the first abdominal 
 segment ; and e, c, represent the lateral portions of the meta-sternum. 
 The winglets, halteres, posterior thoracic spiracles, and the posterior 
 thoracic opening are shown, x 6 diam. 
 
 Fig. 5. A lateral view of the thorax, in which a is the meso-sternum, 
 b, the anterior lateral plate of the meso-thorax, and c the side of the 
 meta-sternum. The other portions will be easily understood by reference 
 to the text, X 6 diam. 
 
 Fig. 6, A longitudinal vertical section of the thorax, showing the dis- 
 position of the main trachea, the ento-thorax of the meso and meta- 
 sternum, and the great muscles of the back, x 6 diam. 
 
 Fig. 7, The under surface of the thorax, showing the four sterna, and 
 the openings with which the coxse are articulated, x 8 diam. 
 
 Fig. 8. The lower half of the thorax, seen from within, showing the 
 four sterna with the ento-thoracic processes, a Eepres^nts the meso- 
 sternum ; 6, 6, the lateral plates of the meso-thorax ; and c, c, portions of 
 the meta'-sternum. x 8 diam. 
 
 Fig. 9. The cephalo-sternum, condyles, and a portion of the pro-sternum, 
 seen from within, X 15 diam. 
 
 Fig. 10. The pro-tergum, seen from within, x 15 diam. 
 
 Fig. 11. A portion of the flank, "seen from within, showing the 
 insertion of the wing and para-pleura, a, Part of the meso-sternum ; 6, 
 episternum of meso-thorax ; c, episternum of meta-thorax ; c?, side of 
 scutellum ; e, part of the dorsum of scutellum, x 8 diam. 
 
 Fig. 12. The insertion of the wing, from without, showing the wing, 
 pivot and socket, a, The ampulla ; b, support of the winglets ; c, the 
 sub-costal nervure, seen in section, x 25 diam, 
 
Plate 
 
 fi.TL. Sculp? 
 
THE INTEGUMENT OF THE THORAX. 03 
 
 thoracic opening, and that of the meta-thorax is invisible exter- 
 nally, except when the abdomen is very much depressed, being 
 hidden under the post-scutellum of the meso-thorax. 
 
 The pro-sternum or sternum of the pro-thorax (Plate V, Figs. 1, 7, and 
 8,) is situated immediately behind the cephalo-sternum. It consists of a 
 central portion and two cornua, which partially surround the anterior 
 coxae. The central portion is a long narrow plate, widest anteriorly ; it is 
 grooved along the mesial line externally, and presents a slight ridge 
 internally ; posteriorly, it sends a narrow plate along the edge of the 
 meso-sternum, and between it and the posterior edges of the coxae on 
 either side. This plate becomes broader externally to the coxa, and ex- 
 tends along the outer edge of its articulation, reaching the lower anterior 
 margin of the anterior spiracle, where it unites with the lateral plate of 
 the pro-thorax and terminates in a curved point in front of the articula- 
 tion of the coxa, near its outer anterior angle, behind the condyle. A 
 stout tooth springs from the outer margin of the aperture, near the an- 
 terior part of the opening, with which the coxa articulates ; it is hard 
 and polished, fits into a hollow on the outer side of the coxa, and forms a 
 point d'appui, upon which the leg moves. 
 
 The posterior extremity of the sternum internally bears a pair of 
 processes, which are directed outwards upwards and forwards, and reach 
 the lower margin of the anterior thoracic spiracles, where they bifurcate 
 to surround those openings. I think it extremely probable that these 
 processes are part of an undeveloped segment intermediate between the 
 pro-thorax and meso-thorax. 
 
 The meso-sternum (Plate V, Figs. 2, 5 a, 7, and 8 a,) forms the greater 
 part of the under surface of the thorax as well as part of the flank. It is 
 marked interiorly by a groove in the mesial line, and is deeply notched 
 posteriorly for the intermediate coxse. It bears very important internal 
 processes; one, the ento-thorax of the meso-sternum, extends its whole 
 length in the mesial line, and supports the great thoracic ganglion. 
 This is Y-shaped when seen either in front or behind, but presents 
 a large surface above, which lodges the ganglion ; its lateral 
 margins give insertion to muscles. Numerous muscles arise from nnd 
 
 Fig. 13, The rod borne by the ampulla, the free extremity of which 
 fits into a concavity in front of the wing pivot. X 50 diam. 
 
 Fig. 14, The wing socket. X 50 diam. 
 
 Fig. 15, The valve, or epitreme, and peritreme of the meta-thoracic 
 spiracle. X 25 diam. 
 
 Fig. 16. The valve, or epitreme, of the anterior thoracic spiracle, 
 X 15 diam, 
 
 Fig. 17. The t>eritreme of the anterior thoracic spiracle, x 15 A' am. 
 
64 ON THE ANATOMY OF THE FLY. 
 
 are inserted into the ento-thorax. Another process on each side is 
 formed by the turning in of the posterior superior angle of the sternum ; 
 it gives origin to a number of muscles connected with the wings, and 
 a broad muscular slip unites its inner edge with the external edge of the 
 superior surface of the ento-thorax. I propose the term para-pleuron 
 for this process. 
 
 The meta-sternum is very narrow beneath, the posterior being 
 nearly approximated to the intermediate coxse. It forms, however, a 
 more considerable portion of the flank. Internally it bears a process in 
 the mesial line, the meta-thoracic ento-thorax, which supports the great 
 nerve trunk, the osscphagus, intestine, dorsal vessel, and salivary glands 
 in their passage to the abdomen. This process is very similar to the 
 ento-thorax of the meso-sternum, but is much narrower ; it almost meets 
 that process anteriorly. A pair of strong ridges extend from it laterally 
 along the anterior edges of the meta-sternum. It gives origin to several 
 pairs of muscles. 
 
 The lateral plates of the pro-thorax are one on each side ( Plate V 
 Figs. 1, 5, and 8,) ; they are best seen internally, where their limits are 
 defined by ridges. They are almost square, but the superior anterior 
 angles unite them to the dorsal plate of this segment. Each is bounded 
 inferiorly by the coxa, posteriorly by the anterior thoracic spiracle, and 
 anteriorly by the condyle and anterior thoracic opening. 
 
 The lateral plates of the meso-thorax are two on each side, the larger 
 is called the episternum, but the smaller, which is likewise posterior, does 
 not appear to have received a name. I shall call it the posterior lateral 
 plate of the meso-thorax, 
 
 The episternum ( Plate V. Figs. 5 and 8 6) is almost square^and very 
 solid. Its superior posterior angle is prolonged into the thorax as a narrow 
 process, from which the anterior muscles of the wing arise. 
 
 The posterior lateral plate ( Plate V, Fig, 5 ) is deeply notched above 
 for the insertion of the wing, but is otherwise nearly square ; a small 
 knob-like projection marks its anterior superior angle and is immedi- 
 ately in front of the insertion of the wing. 
 
 The lateral plates of the meta-thorax are one on each side. They are 
 smaller than those of the meso-thorax, and are notched posteriorly for 
 the insertion of the halteres ; each is bounded below by the posterior 
 thoracic spiracle, anteriorly by the- posterior lateral plate of the meso- 
 thorax, above by the ra eta-thoracic tergum, and behind by the insertion 
 of the halter and the first abdominal segment. 
 
 The tergum or dorsal plate of the pro-thorax (Plate V, Fig. 10 ) is 
 reduced to a narrow ridge, which thickens the edge of the anterior thoracic 
 opening and which gives off a pair of diverging processes posteriorly, for 
 the origin of muscles, These are connected near their base by a thin trian- 
 
THOBACIC APPENDAGES. 65 
 
 gular membrane, in the free edge of which is a leaf -like thickening. Three 
 curved ridges extend from the base of these processes upon the inner 
 side of the prse-scutum of the meso-thorax, and give origin to muscles, 
 
 The meso- thoracic tergum is marked externally by four transverse 
 lines, which divide it into four nearly equal portions. These are separated 
 internally by ridges, which give origin to muscles as well as strengthen 
 the whole structure ; that between the scutum and scutellum sends 
 inwards a small pointed process near the origin of the wings, from which 
 muscles arise for their movement, The lateral edges of the scutellum 
 form a pair of thickened rods for the support of the articulation of the 
 wing, and a loose fold of integument unites the scutellum to the lateral 
 and dorsal plates of the meta-thorax and gives the chief freedom of 
 motion to the whole dorsal region, as the scutum, praescutum, episternum 
 and sternum of the meso-thorax are firmly soldered to each other, but are 
 ufficiently elastic to allow of some movement in the folds between them. 
 The cavil y of the post-scutellum is entirely filled by a pair of air vessels 
 ( Plate V, Fig. 6). 
 
 The tergum of the meta-thorax is almost vertical in its direction. It 
 forms the posterior wall of the thorax, and it is distinctly divided into four 
 parts by ridges corresponding to the same parts of the meso-thoracic 
 tergum ; two of these, however, the scutellum and post-scutellum, only 
 appear internally, the first abdominal segment arising apparently from 
 the junction of the scutum and scutellum. 
 
 Section VJLL The Thoracic Appendages. 
 PLATE I. ; PLATE V., Fias 1214 ; PLATE VL 
 
 The wings consist, as has been stated above, of a double 
 membrane.* This is kept extended by a system of nervnres, 
 which chiefly arise from two primary nervnres. An anterior 
 or costal, which runs along the anterior margin of the wing to 
 its extremity, and a posterior or sub-costal, which arises in 
 common with the former, and runs for a short distance 
 parallel to it, and then splits up into four trunks, which all 
 terminate by joining the costal nervure. Another set of 
 
 * Page 18. 
 
66 ON THB ANATOMY OP THB FLY. 
 
 nervures arise in a fold of the wing posterior to the sub- costal 
 nervure, which all become very fine, and are ultimately lost 
 'near the posterior edge of the wing, which is not guarded by 
 any nervure. These are not connected with the cavity of 
 the thorax by any open canal, like the costal and sub-costal 
 nervures ; but the two systems communicate by a short trans- 
 verse trunk from the posterior branch of the sub-costal near 
 the middle of the wing. 
 
 . The costal nervure is protected at its origin by a small scale or epaulet 
 called the pterogodium, fringed with hairs ; beneath this is a broad scale- 
 like fold, which forms the basal joint of the costal nervure, and which is 
 
 DESCRIPTION OF PLATS VI. 
 The Development and Appendages of the Thorax, 
 
 Fig. 1. The nerve centres of the larva, with the imaginal discs of the 
 optic and antennal segment-, and of the four anterior legs firmly attached 
 to them, x 30 diam. 
 
 Fig. 2. The lateral tracheal vessel of the larva, showing its inner and 
 outer coat, "with the imaginal discs of the wing, halter, and posterior leg 
 firmly attached to it. x 30 diam. 
 
 Fig. 3. The first stage of unfolding of the imaginal discs of the ante- 
 rior pair of legs, showing their capsules, x 15 diam. 
 
 Fig. 4. One of the anterior leg discs unfolded and considerably de- 
 veloped ; about the third day of the pupa state. X 30 diam. 
 
 Fig. 5. The same, at the end of the first week, with the leg of the fly 
 seen within in process of development, x 20 diam. 
 
 Fig. 6. The intermediate leg of the fly. x 10 diam. 
 
 Fig. 7. The tibia of the same, with muscles attached, x 10 diam. 
 
 Fig. 8. The first tarsal joint, with its muscles, x 10 diam 
 
 Fig. 9. The last three tarsal joints, x 20 diam. 
 
 Fig. 10. The last tarsal joint, with its long apodeme and muscle, 
 arising from the femur. X 10 diam. 
 
 Fig. 11. The pulvilli and claws, x 75 diam. 
 
ft*?, VI 
 
THOEACIC APPENDAGES. 67 
 
 connected closely by a membrane with the scutellum above and loosely 
 with the flank below ; it is covered with fine hairs. The second joint of 
 the costal nervure is united to the first by membrane. It is an elongated 
 scale, with its anterior edge curved downwards, so that the anterior edge 
 of the wing appears to possess considerable thickness. The cavity of this 
 scale is very narrow, and, I believe, becomes obliterated entirely in the 
 adult insect. This scale bears a large number of short thick hairs. The 
 costal nervure is continued directly backwards along the margin of the wing 
 from the posterior extremity of this scale. There is, however, the indica- 
 cation of a joint between them. The uervure is set with strong spine-like 
 hairs, and its canal is entirely obliterated by the fibrillation of the 
 mesoderm. 
 
 The sub-costal nervure, unlike the costal, remains tubular during the 
 whole life of the insect, and contains a nerve and tracheal vessel in their 
 passage to certain structures, which may be observed on the nervure. 
 
 The sub-costal is almost cylindrical, except at its origin, where it 
 spreads out into a broad scale, which is closely united to the basal scale 
 of the costal nervure. A strong tooth projects from its underside near to 
 its origin, which I shall call the wing-pivot. This consists of a clear 
 amber- coloured, excessively hard process of chitine, exactly similar to 
 that which forms the extremity of the mandibles in some insects. Its 
 extremity fits into a hollow, in a similar process, which is supported by 
 the flank, and which is shaped something like an inverted molar tooth 
 (Plate V., figs. 12 and 14). 
 
 The organs on the sub-costal nervure, to which the nerve is distributed, 
 are several groups of transparent spots, exactly similar to those on the 
 base of the halteres. These were first pointed out by Dr. Braxton Hicks,* 
 and appear very couhtant in all insects except the Orthoptera and the 
 Coir opt era, where their presence needs confirmation, as they are very 
 different from the double refracting circular bodies so common in their 
 elytra, with which they appear, however, to have been confounded. These 
 organs will be described at length in Section XVIII. I have also traced 
 filaments of the nerve to the bulbs of the hairs, which are situated on an 
 enlargement of the posterior branch of the sub-costal. 
 
 Behind the origin of the sub-costal nervure is a thickened chitinous 
 fold, which bears the winglets (plate V., fig. 126). It is closely connected 
 above with the sub-costal nervure, and beneath with the rod which forms 
 the interior margin of the scutellum. Its external process bears the 
 winglets which are attached to its margin. When the wing is at rest, this 
 
 * Linn. Soc. Trans., 1856-57. 
 
 F 2 
 
68 ON THE ANATOMY OF THE FLY. 
 
 process is directed forwards, as is represented in the plate, but it is capable 
 of revolving like a hinge, so that when the wing is open this process is 
 directed outwards. 
 
 Beneath the last is an irregular-shaped piece of chitine, completing the 
 outer circle of the wing joint, and uniting with the anterior superior angle 
 of the posterior lateral plate of the meso-thorax. This latter bears a pro- 
 tuberance (Plate V., fig. 12 a), filled internally with muscle, which I shall 
 call the ampulla. 
 
 The ampulla hangs over the anterior portion of the wing joint. Its 
 posterior edge bears an irregular ring of chitine beneath the integument* 
 which supports three very remarkable processes ; the centre of this ring is 
 occupied by a loose puckered membrane. 
 
 The most important of these processes is shaped like an inverted molar- 
 tooth (Plate V., figs. 12 & 14) ; it supports the wing pivot in its cavity^ 
 and is probably the joint upon which the wing swings, when rapidly 
 vibrating. It is excessively hard and semi-transparent, like the wing pivot, 
 I shall call it the wing-s ocket. 
 
 In front of the wing-socket is a flattened rod, broadest at its free ex- 
 tremity, and slightly curved (Plate V. figs. 12 & 13). Its free extremity 
 is covered with a soft elastic substance, which easily separates from the 
 rod itself, which is as hard as the wing-socket. Its free extremity fits a 
 small cavity in the anterior part of the base of the tooth beneath the sub- 
 costal nervure, or wing-pivot. Its function may be to keep the wing back 
 whilst vibrating, or even to stop its vibrations, a view indicated by its 
 eoft elastic extremity. The third and last process appears to be rudimen- 
 tary, it consists of a single hardened tooth directed backwards ; I think 
 it may support the wing-pivot in some positions, but have not been able 
 to make out its use. 
 
 The muscles attached to these parts are very numerous, and are most of 
 them furnished with fine tendons ; each muscle consists of a few fibres. They 
 arise from the several internal processes already indicated, but at present the 
 exact origin and insertion of each has not been made out. Shortly, it may 
 be said that every separate piece is furnished with at least two muscles* 
 an extensor and flexor, or an elevator and depressor as the case may be. 
 
 The legs consist of several tubular articulations, the names 
 of which have already been given,* but the three pairs differ 
 somewhat, especially as to the coxas. 
 
 * Page 19. 
 
THORACIC APPENDAGES. 69 
 
 The anterior coxae are the largest ; each consists of a flattened, almost 
 quadrilateral, tubular joint, loosely united by membrane to the pro-sternum. 
 A depression on its outer side lodges a strong tooth, borne by the edge of 
 the pro-sternum, upon which the whole limb moves, so that the joint 
 possesses a freedom of motion almost equal to that of a ball and 
 socket. Another smaller triangular plate is interposed between the 
 coxa and the anterior part of the sternum on the inner side of the limb. 
 This is probably the representative of the basal joint of the coxa, whilst 
 the larger part of the joint corresponds to the second article of the other 
 coxae. 
 
 The femur is articulated with the coxa by a hinge allowing of move- 
 ment in only one direction ; all lateral motion being prevented by the edges 
 of the coxa overlapping the femur, and forming a groove, in which it moves. 
 A small accessory plate, closely united to the femur, protects the under side 
 of the joint. 
 
 The intermediate coxae each consist of a basal segment, which is some- 
 what boat-shaped, but open both above and below. It is flattened 
 anteriorly, and pointed externally. Its anterior edge fits against the 
 posterior edge of the meso-sternum. A stout tooth, but not so large as 
 that which supports the anterior coxa, fits against its inner anterior 
 angle. It projects from the posterior edge of the meso-sternum, near the 
 mesial line. The joint between this part of the coxa and the meso- 
 sternum is a hinge joint, moving on its posterior edge, and giving the 
 limb a motion in an antero-posterior direction. 
 
 The second joint of the intermediate coxa is short, and almost conical. 
 It is loosely connected with the last, and has a hinge-like motion upon its 
 posterior and inner edge, which gives the limb a considerable lateral play. 
 
 The posterior coxae differ from the intermediate, in the first joint being 
 more nearly round, and in the absence of any well-marked tooth at the 
 inner and posterior part of the nieta-sternum. A very small tooth exists, 
 however, externally to the limb at the edge of the meta-sternum. These 
 teeth, both in the intermediate and posterior limbs, seem to act rather as 
 checks to the movement of the limb, preventing the thoracic muscles 
 from drawing the coxa into the interior of the thorax, than as pivots, upon 
 which the coxae move, like those of the anterior limbs. 
 
 The joints between the coxae and femora in the posterior and intermediate 
 limbs, resemble the same in the anterior legs. The articulations between 
 the femora and tibiae allow of no lateral motion, the femur being grooved 
 vertically for the reception of the tibia. The tarsal joints allow of slight 
 lateral motion, the distal extremity of each being hollowed beneath into 
 a cone-like cavity, which receives the narrow proximal end of the next. 
 A similar joint exists between the proximal tarsal articulus and the tibia, 
 
70 ON THE ANATOMY OF THE FLY. 
 
 only very little lateral movement is allowed, from the manner in which 
 the muscles are inserted into it. 
 
 The muscles which move the coxae and femora arise from the back, and 
 form the vertical groups of muscles, which occupy the lateral regions of 
 the thorax. A pair are inserted by fleshy insertions into each portion of 
 the coxa, and another pair terminate in tendons, and are inserted into the 
 proximal extremity of each femur, one into its upper or outer, and one 
 into its lower or inner margin, the former being a flexor and the latter 
 an extensor of the limb. The tendons of the anterior femoral muscles 
 receive fleshy slips, which arise within the anterior coxae, and are 
 inserted into them in a bipenniform manner. 
 
 Each of the other joints, except those of the tarsus, gives rise to a pair 
 of bipenniform muscles, a flexor and extensor of the succeeding joint. The 
 muscle, which moves the terminal joint of the tarsus, arises in the upper part 
 of the femur, and is connected with the last tarsal joint by a long thread- 
 like apode'me. The apode'me is connected with the proximal extremity of 
 each tarsal joint by a membranous slip, and terminates in a triangular 
 plate, which bears the pulvilli and claws. This plate is grooved below by 
 a number of parallel grooves on either side of the mesial line. These 
 are seen very frequently in the feet of insects. It terminates in a long 
 blunt spine, and supports the pulvilli beneath, as well as the claws above. 
 
 The last four tarsal joints are occupied, as has been already stated,* by a 
 sac, which secretes a viscid fluid. As I have only succeeded in seeing 
 this sac in the manner already indicated, I cannot say whether it is single 
 or double ; neither can I tell whether the fluid secreted by it is conveyed 
 into the pulvilli by ducts or by an opening in the neck of the sac, or 
 whether it passes into it simply by transudation. The great delicacy of 
 the secreting sac and the opacity of the tarsus renders it exceedingly 
 difficult, if not impossible, to make out these details in the fly. 
 
 Section VIII. The Abdominal Segment*. 
 
 The abdominal segments are far less complicated than those 
 of the thorax. They are nine in number, but the last four are 
 modified to form the sexual organs, these will be described in 
 their proper places.f 
 
 * Page 19. f Page 104. 
 
THE BESPIBATORY OBGANS. 71 
 
 The five anterior abdominal rings each consists of a narrow 
 ventral plate loosely united by membrane to a dorsal arch, 
 which encircles more than four-fifths of the abdomen, and 
 which is perforated beneath by a pair of small round spiracles. 
 
 The first abdominal segment is reduced above and laterally to an 
 exceedingly narrow ring; its ventral plate is well marked but very 
 short, although it is wider than the other ventral plates. The next 
 four abdominal segments enclose the greater portion of the abdominal 
 cavity. The first ring is contracted in front, the second is equally 
 contracted both in front and behind, whilst the two posterior rings 
 form the conical apex of the abdomen. The ventral plates of these 
 segments are longest from before backward ; and the last in the male is 
 deeply divided behind. The membrane between the ventral and dorsal 
 plates is sufficiently loose to allow of considerable extension of the abdomen 
 by the separation of the plates, but they are united by a series of flat 
 muscular bands which cover the whole inner surface of the abdomen. 
 These may be divided into three sets ventral, dorsal, and transverse. 
 The ventral muscles are the largest and most conspicuous. They form 
 pyramidal bundles on each side of the mesial line. They arise from the 
 anterior part of each ventral plate near the mesial line, and are inserted 
 into the anterior border of the next ventral plate, by an insertion rather 
 broader than their origin. A few fibres pass occasionally from one 
 muscle to join the insertion of the next. The transverse muscles are narrow 
 muscular bands, which arise from the sides of the ventral plates at 
 the junction of their middle and outer thirds, and are inserted into the 
 adjacent edges of the dorsal arches. The dorsal muscles arise from the 
 anterior part of each dorsal arch below, but near their posterior edges 
 above, and are inserted into the anterior edge of the next dorsal arch. 
 All the muscles are best seen by hardening the part in spirit or chromic 
 acid. This renders them opaque, so that they may be easily observed by 
 viewing the abdominal integument by reflected light. 
 
 Section IX. The Respiratory Organ*. 
 PLATE I., PLATE V., Fias 6 AND 15 17, AND PLATE IX., Fia. 4. 
 
 The spiracles or external openings of the respiratory organs 
 are ten in number on each side of the male, and eight in the 
 female. 
 
72 ON THE ANATOMY OP THE FLY. 
 
 The anterior thoracic spiracle (plate V., fig. 17) is oval and 
 narrowest above. Its margin consists of a thickened ring 
 chiefly visible internally, and formed by the extremities of 
 the process already described as extending from the posterior 
 extremity of the central portion of the pro-sternum to the 
 Bpiracle. 
 
 The anterior spiracle is situated between the pro-thorax and 
 meso-thorax, but the meso-thoracic tergum reaches over its 
 superior margin, and joins the pro-thorax in front of the 
 spiracle ; beneath it is entirely bounded by the pro-thorax. 
 The edge of the spiracle is continued as a thin semi-lunar 
 fold, which reduces the opening to a mere slit. This is 
 strengthened by chitinous arborescent nervures, which project 
 beyond its free edge, so as to fringe it with a delicate fringe of 
 sharp spicules. I shall call it the peritreme, although in 
 the posterior thoracic spiracle the same structure forms a pair 
 of leaf -like valves. 
 
 Close behind the peritreme is the true valve by which the 
 aperture is closed This resembles a pair of callipers jointed 
 below, but the two sides shut against each other like the leaves 
 of a book. I shall call this structure the epitreme (plate Y., 
 fig. 16). It is closely united to the wall of the tracheal vessel, 
 and is probably developed in its inner layer or that which 
 corresponds to the protoderm. By breaking away a small 
 portion of the peritreme of the anterior spiracle of the living 
 fly, the epitreme may be observed opening and closing regularly 
 with each act of respiration. As has been already observed, 
 when a fly is held by the wings it makes regular efforts with 
 its legs ; during these the peritreme remains closed, but may 
 be observed to open as soon as the effort ceases. 
 
 The peritreme of the anterior spiracle is opened and closed 
 
THB NEBVOUS SYSTEM. 73 
 
 by a small muscle, which arises from the margin of the 
 spiracle below, and is inserted into a pair of thin plates, close 
 to the hinge of the peritreme. The structure of the posterior 
 thoracic spiracles differs only in detail from that of the anterior. 
 
 The posterior thoracic spiracle may be said to be triangular, with 
 rounded angles ; its superior border is longest and straight, the other sides 
 are slightly curved, the posterior edge is shortest. It is surrounded 
 internally by a deep ridge, derived from the posterior margin of the 
 meta-sternum, which is especially marked at the posterior inferior angle 
 of the spiracle. 
 
 The pseudo- valves, or peritreme, consist of a pair of valve-like plates, a 
 small posterior, and a large anterior valve. These are attached beneath to 
 the meta-sternum, which forms the inferior boundaries of the spiracle ; 
 the upper margin of which is formed of the lateral plate of the meta- 
 thorax, and this is so deeply notched behind, that the halter arises 
 quite close to the posterior superior angle of the spiracle. The pseudo- 
 valves are strengthened by arborescent nervures similar to those of the 
 peritreme of the anterior spiracle. 
 
 The valve or epitreme arises from the posterior inferior border of the 
 spiracle : its hinge resembles that of the epitreme of the anterior spiracle, 
 except that the plates into which the muscle is inserted are beneath 
 the hinge. Its action has not been observed during life, owing to its 
 depth from the surface being greater than that of the epitreme in the 
 anterior spiracle. It undoubtedly entirely closes the opening, but it is 
 uncertain whether its opening and closing is synchronous with that of 
 the anterior epitreme. 
 
 The main trachea of the thorax extends on each side from the anterior 
 to the posterior thoracic spiracle, lying close to the thoracic viscera. It sends 
 a large branch which curves inwards and forwards into the head, and another 
 smaller branch into the abdomen. This latter swells out on each side into 
 the great abdominal pulmonary sac. More properly, perhaps, the anterior 
 and posterior portions should be called continuations of the lateral tracheae 
 and the portions uniting it with the spiracles should be looked upon aa 
 branches. A large trunk from the posterior thoracic spiracle follows the 
 wall of the thorax upwards and inwards ; it sends a large branch forwards 
 in the mesial line between the muscles and the back. This gradually be- 
 comes narrower and supplies the dorsal muscles, and sends a prolongation 
 backwards which fills the whole cavity of the post-scutellum. (Plate I. & 
 Plate V. fig. 6). 
 
 A number of small sacculi, connected with each other in a chain, extend 
 
74 ON THE ANATOMY OF THE FLY. 
 
 on each side from the main trachea near the anterior thoracic spiracle, 
 to the same vessel near the posterior thoracic spiracle. These lie between 
 the lateral thoracic muscles which they partially supply they also send 
 branches to the legs. 
 
 The cephalic branches of the lateral tracheae pass into the head and 
 expand into sacs which envelope the posterior and lateral portions of the 
 cephalic ganglion. Each of these sends a large branch forward to the 
 insertion of the antennas, which is continued along the side of the pharynx 
 to the proboscis ; and numerous offsets to the compound eyes, which they 
 almost surround. 
 
 Another and smaller portion of the cephalic branch of the lateral 
 thoracic trunk passes over the cephalic ganglion and unites with its fellow 
 in the mesial line just behind the nerve to the ocelli. It sends off several 
 lateral offsets to the lateral portions of the cephalic ganglia, and supplies 
 the ocelli. 
 
 The abdominal pulmonary sacs are connected by a narrow vessel with 
 the posterior extremity of the lateral thoracic trunk. They are very large 
 and spherical, occupying almost the whole base of the abdomen. 
 Each gives off several sets of vessels, kept open by a spiral fibre to the 
 abdominal viscera and dorsal vessel. With this exception, all the tracheae 
 already described are membranous in the adult insect. 
 
 The abdominal spiracles are small round pores, about l-200th 
 of an inch in diameter although they vary considerably in 
 size in the underside of the dorsal arches. There are none 
 to the first segment, the dorsal arch being rudimentary. The 
 next four have each a pair ; but the sixth abdominal or first 
 segment of the generative organs has two pairs in the 
 female, one above the other, and only one in the male. The 
 remaining segments have none in the female, but the seventh 
 has two pairs, and the eighth a single pair, in the male. 
 
 The peritreme of the abdominal spiracles is a mere thick- 
 ened rim, with a number of minute spiculas projecting from 
 its margin. The epitreme is situated about 1-1 5 Oth of an inch 
 within the spiracle, and is connected with the peritreme by a 
 short unbranched tracheal tube, with a few thread-like rings 
 
THE DOB8AL VESSEL. 75 
 
 in its wall. The epitreme consists of a ring more or less 
 hardened on one side, from which a kind of handle projects. 
 This handle is connected with the ring differently on its two 
 sides, the junction being vertical on one side and horizontal 
 on the other (Plate IX., fig. 4). A short muscle arises from 
 the peritreme, and is inserted into the handle of the epitreme. 
 The spiracle is closed by depressing the handle, which 
 folds the vertical against the horizontal half of the ring. 
 Immediately within the epitreme the trachea swells out, and 
 gives off numerous large branches, each one as large or larger 
 than the trunk between the epitreme and the peritreme. 
 
 Section X.The Dorsal Vessel 
 PLATE IX., FIG. 1. 
 
 The dorsal vessel of the fly is perhaps the most difficult 
 organ in the insect to make out thoroughly. Its walls consist of 
 three coats at least; an external muscular, an intermediate 
 membranous, and an internal elastic coat. The external and 
 middle coats seem to form only a loose investment around the 
 elastic tunic in the abdomen, and after death when the vessel 
 is empty they appear as a semi-transparent fusiform flattened 
 band beneath the integument. They seem to be stretched 
 laterally by the tracheal vessels which supply them, and pro- 
 bably also by a thin membranous exparsion which connects 
 them with the abdominal parietes. The muscular fibres are 
 striated ; and the majority, if not all, are longitudinal. I am 
 inclined to regard this portion of the organ as homologous to 
 the lateral wings of the dorsal vessel ef the cockchafer. 
 
76 ON T11E ANATOMY OP THE PLY. 
 
 The internal or elastic coat is constricted at intervals, 
 forming five chambers, with a valvular fold between each. 
 The three posterior chambers are very narrow, slightly enlarged 
 at their anterior extremities ; but the two anterior form a 
 considerable dilatation. Each has a pair of small lateral openings 
 protected internally by a valvular fold. The elasticity of this 
 coat causes the vessel to curl up when detached from the 
 abdominal wall ; it probably keeps the chambers open* during 
 life, whilst they are emptied successively by the contraction of 
 the muscles which surround them. So that the action of the 
 vessel is analogous to that of Dr. Richardson's sether spray 
 apparatus. The contraction of the whole length of the mus- 
 cular coat is not simultaneous, but takes place by a series of 
 undulations commencing at its posterior extremity. 
 
 The anterior abdominal chamber is prolonged anteriorly as 
 a tubular vessel, which lies above the chyle stomach and 
 beneath the great dorsal thoracic muscles. This vessel enters 
 the head above the oesophagus, and ends by branching into six 
 or eight minute tubes, which terminate in open mouths in tho 
 cephalic cavity. The muscular wall of the dorsal vessel in the 
 abdomen is continued in the thoracic portion of the vessel and 
 its terminal branches ; but the elastic coat is either absent or 
 very little developed. 
 
 The development of the dorsal vessel has not hitherto been 
 worked out. The difficulties of the necessary observations are 
 very great, from its great transparency and elasticity. It is 
 
 * This is the only theory I can suggest to account for the dilatation of 
 the dorsal vessel, since the influx of fluid can have no effect in dilating 
 the chambers, as in animals having a closed circulation. I think 
 it probable that the wings of the dorsal vessel of the cockchafer enclose 
 an elastic tube exactly as the muscular tubs encloses the elastic one in the 
 fly, except that their direction is transverse instead of longitudinal. 
 
THK NERVOUS SYSTEM. 77 
 
 most easily observed in the living insect when it first 
 emerges from the pupa. When removed from the abdomen 
 the folds of the elastic coat are easily obliterated, unless it is 
 first hardened in a solution of chromic acid. 
 
 Section XI. The Nervous System. 
 PLATE VI., Fio. 1, AND PLATE VII., Fias. 1 1 AND 10. 
 
 The parts of which the nervous system is composed have 
 already been indicated.* The cephalic nerve centre will be best 
 described by describing its component ganglia. These are three 
 pairs, which I shall name the central or antennal, the anterior 
 or cerebroid, and the lateral or optic ganglia. 
 
 The great thoracic nerve cord forms the lowest portion of 
 the cephalic nerve centre, and when seen from beneath appears 
 as a somewhat lozenge-shaped flat surface, from the anterior 
 portion of which the great nerves of the probocis arise. These 
 appear to be chiefly composed of fibres from this cord. A pair 
 of crura arise from the upper side of the thoracic cord, and 
 diverge to allow the oesophagus to pass between them ; these 
 enter the cephalic nerve centre where their fibres spread out, 
 and connect the different ganglia of which it is composed. 
 The tracing of these is exceedingly difficult, but I believe 
 I have clearly made out, that a large bundle extends to each 
 optic ganglion, that another set chiefly run up the back of the 
 central ganglia, whilst others accompany the oesophagus, and 
 appear anteriorly as a pair of small nerves, upon which the 
 
 * Page 13. 
 
78 ON THE ANATOMY OP THE PLY. 
 
 ganglia of the proboscis are situated. I shall call these the 
 cssophageal nerves. 
 
 e^ central ganglia form an ovate mass pierced by the 
 oesophagus, and supported below by the expanded extremity 
 and crura of the thoracic nerve trunk. A slight groove above 
 
 DESCRIPTION OF PLATE VII. 
 
 Fig. 1 . The entire ventral cord and the cephalic and thoracic ganglia. 
 The great splachnic ganglion and a few sympathetic nerves are repre- 
 sented as they would appear if it were possible to detach them from the 
 viscera, to show their connection with the moto-sensory nerves, x 15 diam. 
 
 Fig. 2, A vertical section through the middle of the cephalic ganglion, 
 showing the oesophagus, the nerves of the antennas and proboscis at their 
 origin, and a portion of the cephalo-thoracic ventral cord, and the cerebroid 
 ganglion of the right side with its cms : to show the relations of these 
 parts. X 30 diam. 
 
 Fig, 3. The anterior termination of the great cephalo-thoracic ventral 
 cord, showing the crura of the optic ganglia and nerves to the proboscis ; 
 the pharyngeal and oral ganglia are also represented, x 30 diam. 
 
 Fig. 4. The cerebroid ganglia, isolated. X 100 diam. 
 
 Fig. 6. A portion of the interior of the great eyes, showing from 
 above downwards, the following parts : a portion of the compound corm a, 
 the cones and rods, with nuclei imbedded in their inner < xtremities, the 
 common retina, the clear optic nerve, and the optic zone with its tracht al 
 vessels. X 150 diam. 
 
 Fig. 6. Inner extremities of rods with their nuclei ; a plexus of 
 tracheal vessels and the choroid pigment is seen between the rods, x 500 
 diam. 
 
 Fig. 7. A cone detached, with the outer extremity of its rod and a 
 single facet of the cornea, showing the structure of the rod, disposition of 
 pigment, and nuclei in the pigment layer of the cone, x 600 diam. 
 
 Fig. 8. One of the ocelli, showing the cornea, choroid, and plexus of 
 tracheal vessels, x 200 diam. 
 
 Fig. 9. The cornea of one of the ocelli, seen in profile, x 200 diam. 
 
 Fig, 10. Ultimate nerve loops of the sympathetic nerves of the ovary. 
 X 600 diam. 
 
K.T.L. Sculp* 
 
THE NEBVOU8 8YBTEM. 79 
 
 indicates the division of the two ganglia ; the nerve to the 
 ocelli appears to arise from this groove, but has more pro- 
 bably a deep origin from the optic ganglion. Anteriorly, the 
 central ganglia are hollowed out for the reception of the 
 anterior ganglia ; beneath this hollow they give off the great 
 antennal nerves. I have observed slight convolutions situated 
 in the interior of the central ganglia, on making a section 
 through them. 
 
 The anterior or cerebroid ganglia (Plate VII., figs. 2 and 4) 
 are very remarkable nerve centres, supported on a pair of 
 round peduncles, deeply convoluted and united by a well 
 marked commissure in the mesial line. These give off no nerves, 
 and their peduncles seem to arise partly in the substance of the 
 central ganglia, and partly from the thoracic nerve trunk. 
 They are best seen by making vertical sections of the whole 
 cephalic nerve centre after hardening in spirit; but may be 
 seen very frequently by carefully compressing the hardened 
 nerve centre between the slide and thin glass. Their demon- 
 stration is not very easy without great practice, owing to their 
 being imbedded in the central ganglia. 
 
 The lateral ganglia form an ovate mass on either side of the 
 central ganglia. They differ in no way apparently, as far as 
 structure is concerned, from the other ganglia ; but they are 
 covered externally by a thick layer of nerve cells, very largely 
 supplied with trachea 1 vessels, which form long loops parallel 
 to each other. They give origin externally to the nerves of the 
 compound eyes. 
 
 The cephalic nerve centre is supplied by very minute tracheal 
 vessels. These arise from the great tracheal sacs, which almost 
 entirely invest it. 
 
 The oosophageal nerves appear close to the under side of the 
 
80 ON THE ANATOMY OP THE PLY. 
 
 oesophagus, where it emerges from the cephalic nerve centre. 
 Near the posterior extremity of the pharynx, each gives 
 off a branch, and these uniting, bear a small ganglion between 
 them, which I shall name the pharyngeal ganglion. Although 
 I believe it may possibly correspond to the frontal ganglion of 
 Lyonet, as I am not sure, it seems better to name it after 
 the organ against which it is situated. The remainder of the 
 cesophageal nerves is distributed to the muscles of the pharynx. 
 
 A pair of small nerves arise from the pharyngeal ganglion, 
 and run beneath the muscles of the salivary valve, which they 
 supply ; a little anterior to the valve they again unite, and 
 bear a ganglion, which sends filaments to the mouth, I shall 
 call it the oral ganglion. 
 
 The great nerves to the proboscis arise from beneath the 
 central portion of the cephalic nerve centre. Near the posterior 
 part of the first joint of the proboscis they divide into three 
 large trunks, which supply the palpi, the muscles of the pro- 
 boscis, except those of the pharynx, and the great lobes of the 
 proboscis, ending in loops upon the hair bulbs, and supplying 
 the nipple-shaped processes upon the surface of the lobes. 
 
 The nerves to the antennaa, ocelli, and great eyes, will be 
 described with those organs. 
 
 The thoracic nerve centre is supported by the meso-sternal 
 ento-thorax, and consists of four pairs of large ganglia, of 
 which the second are the largest, and the fourth the smallest. 
 
 Externally, the thoracic nerve centre appears to consist 
 chiefly of nerve fibres, having a longitudinal direction. These 
 form a kind of envelope around the several ganglia, which they 
 connect with each other. The ganglia consist chiefly of serve 
 cells, and are largely supplied with tracheal vessels. 
 
THE NERVOUS SYSTEM. 81 
 
 The nerves which are given off from the thoracic nerve centre, appear 
 to arise chiefly from its external or fibrous layer, hence it is difficult to 
 determine precisely to which ganglion they belong. Three large and two 
 small pairs arise from the upper, and an equal number from the under 
 surface of the nerve centre. 
 
 The first large pair arise from the upper surface of the first pair of 
 ganglia ; they are distributed to the muscles of the pro-thorax, and pro- 
 bably also, to the integument.* A branch of this nerve ends in nerve 
 cells upon the thin lobes of the cephalo-sternum. 
 
 The second pair arise with the last, and apparently supply the muscles 
 and integument of the pro-thorax. 
 
 The third pair arise beneath the nerve centre, apparently between the 
 first and second ganglia, chiefly from the first ; they are especially dis- 
 tributed to the anterior pair of legs. They supply the muscles ; a large 
 terminal branch bifurcates in the tarsus, and supplies the integument of 
 the tarsup, the pulvilli and claws. 
 
 The fourth pair are very large ; they arise from the upper surface 
 between the first and second ganglia. They chiefly supply the muscles of 
 the meso-thorax; they probably also supply the integument. A small 
 branch enters the subcostal nervure of the wing, and is distributed to 
 certain special structures described in Section XIII. 
 
 The fifth pair arise from the under surface of the second pair of ganglia ; 
 they are distributed to the intermediate legs. Their distribution is similar 
 to that of the third pair. 
 
 The sixth are a small pair from the upper surface of the second ganglia ; 
 they are probably chiefly nioior, and accessory to the fourth. 
 
 The seventh pair are distributed to the halteres ; they arise from above 
 the junction of the second and third ganglia. 
 
 The eighth are probably merely au oifaet of the ninth ; they are a small 
 pair like the second and sixth. I am not certain, whether these small 
 
 * In, giving the distribution of these nerve*, 1 have attempted to give an 
 approximate idea; 1 ha/e only traced ike larger branches, and even that 
 LS a most minute and dinlcult dissection. Many which appear to end in 
 mutcles may pierce them, whence it is almost impossible to be quite certain 
 of their distribution. 
 
 O 
 
82 ON THE ANATOMY OP THB FLY. 
 
 nerves are sensory, or motor, or both ; they seem to be lost amongst the 
 muscles. 
 
 The ninth arise beneath, between the third and fourth ginglia, chiefly 
 from the third; tl>3y supply the posterior legs, and are distributed 
 similarly to the third and fifth nerves. 
 
 The tenth pair are small cords, which arise from the fourth pair of 
 ranglia behind ; they enter the abdomen, and are apparently chiefly lost 
 amongst the cells and fibres of the sympathetic plexus. I have distinctly 
 traced a connection between the two systems at this point. 
 
 The ventral cord is continued from the posterior extremity 
 of the thoracic nerve centre ; its fibres may be traced passing 
 backwards from the surface of the latter. It is the great 
 abdominal nerve, and gives off a pair of nerves for each segment ; 
 it terminates in numerous branches to the generative organs. 
 No ganglionic swellings exist in this nerve. 
 
 The sympathetic or visceral system of nerves has three 
 large and a multitude of small nerve centres. 
 
 The largest, which I shall call the great splanchnic, should 
 probably be considered as a pair of ganglia. It is situated in 
 the posterior opening of the thorax, and is covered beneath by 
 a layer of orange brown hexagonal pigment cells. It is largely 
 supplied with minute tracheal vessels, which form a loose 
 network, between the meshes of which, a vast number of nerve 
 cells and fibres lie. The great splanchnic seems to be connected 
 with all the visceral ganglia, and to send branches over the 
 surface of all the viscera. 
 
 A pair of large ganglia are spread out upon the surface of 
 the crura, which connect the oesophagus and lateral lobes of 
 the sucking stomach, one on each side of that organ. These 
 are closely connected with the great splanchnic ganglion. 
 
 Numerous small ganglia abound upon the viscera, and 
 
THE NERVOUS SYSTEM. 83 
 
 the whole are connected by meshes of fibres, which appear 
 like strings of clear gelatinous looking nerve substance, 
 flattened and spread out in places, apparently without any 
 neurilerama (Plate VII., fig. 10). These nerves are especially 
 abundant around the ovaries and testes. 
 
 In the larva, all the nerve centres are concentrated in the 
 anterior segments. They consist of a pair of hemispherical 
 ganglia above the oesophagus, supra-cesophageal ; and a large 
 flattened nerve centre, composed of a series, probably of 
 twelve pairs of ganglia, beneath the oesophagus, sub-cesophageal 
 (Plate VI., fig. 1). 
 
 The supra-oasophageal ganglia are connected by a pair of 
 thick crura with the sub-oesophageal nerve centre. These crura 
 become considerably lengthened at the commencement of the 
 pupa state, and ultimately unite posteriorly, and form the 
 great cephalo-thoracic trunk, uniting the cephalic and thoracic 
 ganglia. 
 
 The supra-cesophageal ganglia of the larva seem to cor- 
 respond to the lateral or optic ganglia of the imago. They give 
 off a pair of nerves to the rudimentary sensory organs of the 
 maggot; but I have been able to detect no commissure between 
 these ganglia, above the oesophagus. About the third day of 
 the pupa state a pair of small oval ganglia appear, uniting 
 the supra-cesophageal ganglia ; these are the rudimentary an- 
 tennal or central ganglia of the imago. They seem to bo 
 developed in the crura of the supra-cesophageal ganglia, and 
 they may exist in an extremely rudimentary condition in the 
 larva, although I have entirely failed to detect them. The sub- 
 cesophageal nerve centre consists of a series of ganglia, closely 
 united to each other, which give off twelve pair of nerves to 
 the twelve posterior segments of the larva. In the pupa it IH 
 
 o 2 
 
84 ON THE ANATOMY OF THE FLY. 
 
 gradually separated from the snpra-cesophngeal ganglia, by the 
 development of the cephalo- thoracic nerve trunk from the 
 crura of the latter. It still supplies the twelve posterior 
 segments of the imago, although the number of ganglia is 
 reduced to four pairs, either by the suppression, or more pro- 
 bably by the coalescence, of several. 
 
 The nerve centres of the larva are surrounded by the imaginal 
 discs, from which the head and part of the thorax of the 
 imago is developed. 
 
 Section XII. The Compound Eyes. 
 PLATE VII., FIGS. 57, AND PLATE VIII., Fios. 1 & 2, 
 
 The compound eyes consist of a compound cornea, beneath 
 which are the rods and cones. These are connected with the 
 nerve centre through the medium of nerve cells and fibres. 
 
 The compound cornea is covered by a layer of protoderm, 
 which retains the form of the lenses beneath, when these 
 are dissolved out with a solution of caustic potash. The 
 lenses (Plate VIII., fig. 1), are bi-convex, consist of a trans- 
 parent substance, and are united to each other by a flat 
 frame-work of the same material. In fact, the compound 
 cornea ought properly to be considered as a single piece, sculp- 
 tured, as it were, into numerous lenses The diameter of each 
 lens is about l-1000th of an inch. The inter-lenticular spaces 
 are generally hexagonal frames, but some near the edge of the 
 eye are almost square. 
 
THE COMPOUND ETES. 85 
 
 Beneath the cornea are the cones (Plate VII., figs. 5, & 7) 
 These are clear conical transparent bodies, about l-700th of an 
 inch in length, which are surrounded by a membrane coloured 
 with bright red pigment, and containing nuclei. A clear basement 
 membrane exists between the substance of the cones and the 
 pigment membrane, and the latter is probably formed entirely 
 by the coalescence of pigment cells. The inner surface of the 
 lenses of the cornea and the bases of the cones correspond, each 
 lens having its own cone. 
 
 The rods (Plate VII., figs 5 and 6), by which the cones are 
 continued to the nerve centre, are cylindrical, and connect the 
 apices of the cones with a layer of nerve cells beneath. They 
 are separated from each other by considerable spaces, traversed 
 by tracheal vessels, and filled with red pigment. Each rod is 
 nearly 1-1 00th of an inch in length. 
 
 The rods consist of an external or cortical substance, which 
 forms the greater part of their thickness. A clear axis 
 cylinder, continuous with the axis of the cone, may be observed. 
 A third layer of transparent material appears to be inter- 
 posed between the axis cylinder and the cortical part of the 
 rods, and to extend as a thin layer into the cones between the pig- 
 mentary and transparent portions of those bodies. At the apex of 
 each cone and around the outer extremity of the corresponding 
 rod is a thickened ring of pigment, the use of which is unknown. 
 A nucleus-like body, refracting light strongly and coloured like 
 the pigment, is imbedded in the inner extremity of each rod. 
 It is best Been by hardening the part in a solution of chromic 
 acid. The inner extremities of the rods are imbedded in a 
 thin layer of nerve cells, which I shall call the common retina. 
 
 Little can be said concerning the uses of the parts above 
 described. I think it not improbable, that the hollow cylinders 
 
86 ON THE ANATOMY OF THE FLY. 
 
 intermediate between the cortical and axis cylinders of the 
 rods are nerve tissue, expanding into retinae between the 
 pigmentary and transparent portions of the cones, and con- 
 necting these partial retinee with the common retina beneath. 
 On the other hand it is, perhaps, the common retina which 
 receives the picture, the whole of the rods and cones being 
 
 DESCRIPTION OF PLATE VIII. 
 
 Fig. 1. A small portion of the compound cornea, x 300 diam. 
 
 Fig. 2. The epineuron of tho right eye, seen from within, with part of 
 its muscle, x 25 diam. 
 
 Fig. 3. The haltere, seen from abov- . x 100 diam. 
 
 Fig. 4. A portion of one of the cylinders, from the base of the haltere. 
 X 600 diam. 
 
 Fig. 5. A portion of one of the dome segments, from the base of the 
 halteres. x 500 diam. 
 
 Fig. 6. Section of one of the cylinders, showing the otoconia and nerve 
 c -11s and fibres beneath its ridges, x 500 diam. 
 
 Fig. 7. A small portion of the subcostal neivure, showing pait of a 
 row of wing organs, x 300 diam. 
 
 Fig. 8. The antenna of the light side, x 50 diarn. 
 Fig. 9. First joint of the same, x 75 diam. 
 Fig. 10. Second joint of the same. X 75. 
 
 Fig. 11. Diagramatic i epresentation of a componnd sacculus, from 
 the third joint of the antenna, seen fiom above. 
 
 Fig. 12. Diagramatic section of the wall of the third joint of the 
 antenna, showing one of the compound and two simple sacculi- 
 
 Fig. 13. Terminal branches of antennal nerve, from the third joint of 
 the antenna, x about 500 diam. 
 
 Fig. 14. A small portion of the frontal sac, showing its papillae 
 X 750 diam. 
 
 Fig. 15. A small portion of the maxillary palpus, showing its saccu- 
 lar organs, x 200 diam, 
 
 Fig, 16, One blade of the cephalo-steinum. x 300 diam. 
 
J3.T.L. Sculp? 
 
THB COMPOUND EYES. 87 
 
 concerned in its formation. At present nothing certain can 
 be said on the subject, although I incline entirely to the 
 former view. 
 
 The cones and rods form a layer rather more than 1-1 00th of 
 an inch in thickness beneath the cornea, and are so arranged 
 that all the rods stand at right angles to their respective 
 facets. The layer of nerve cells behind the rods, is connected 
 with the cephalic nerve centre by a number of very clear nerve 
 filaments, these are about l-2500th of an inch in diameter. 
 I shall call them collectively the optic nerves (Plate VII., figs. 
 1 and 5). 
 
 A thick chitinous ridge, probably the anterior margin of the 
 dorsal plate of the fifth segment, surrounds the anterior edge 
 of the compound cornea within, and bears an oval ring of 
 chitine connected with it below by a narrow pedicle. This 
 ring is very similar to the valves of the spiracles, and is 
 jointed like them, both above and below. A thin loose mem- 
 brane unites its outer edge with the edge of the cornea all 
 round. A narrow tendon is connected with its posterior half 
 about midway between its upper and lower joints ; by means of 
 which a small round muscle, which arises from the margin of 
 the occipital foramen, is inserted into it ; I shall call it the 
 epineuron* (Plate VIII., fig. 2). 
 
 A large tracheal sac, as has been already stated, spreads out 
 behind the cephalic nerve centre, and almost surrounds the 
 optic nerve. This sac sends a vast number of small tracheal vessels 
 between the nerves and the epineuron, which dip into the spaces 
 between the rods, forming a dense plexus around their inner ex- 
 tremities. A long slightly tortuous branch is given off from this 
 
 * In giving this name I have been guided by its similarity to the 
 epitrema, to which I believe it is homologous. 
 
88 ON THE ANATOMY OF THE FLY. 
 
 plexus, between each pair of rods, which runs directly towards 
 the cones, tapering gradually to a fine blind point between 
 them, but giving off no branches. These vessels are pro- 
 portionally of large size, the diameter of the largest near its 
 origin being quite 1 -2000th of an inch. The use of these 
 vessels is unknown ; but the quantity of air supplied to them is 
 manifestly controlled by the epineuron. They may be concerned 
 in altering the shape of the cones and rods, so adapting the eye 
 to distances or to changes of light ; they certainly have a 
 special function distinct from the supply of air to the structures 
 of the eye. 
 
 The optic nerves connect the common retina with a layer of 
 nerve cells, nearly l-100th of an inch in thickness, which 
 surrounds the outer portion of the optic or external ganglion 
 I shall call it the optic zone. This is largely supplied with 
 long loops of trachea3, which commence in a fine plexus between 
 the zone and optic ganglion. The plexus is formed of vessels 
 which dip in between the zone and ganglion, and give off 
 very fine almost straight tracheae, which form long narrow 
 loops by again uniting on the surface of the optic zone, after 
 having traversed its entire thickness. 
 
 The development of the compound eyes commences very 
 early in the formation of the imago. When the pupa is 
 removed from the pupa case on the fourth day, they appear as 
 narrow transparent oval spaces externally. The optic ganglion, 
 even at that early period, appears separated into two parts by a 
 groove. The more external of these eventually forms the optic 
 zone, and the rods and cones are represented by several layers 
 of cells. On the fifth day, these parts are sketched out, as it 
 were, in the cellular layers. Pigment appears on the sixth day 
 of a faint rust colour, and all the parts of the eye are recog- 
 nizable. The compound cornea consists at this period of n. 
 
THB OCELLI. 89 
 
 compact layer of cells, and the protodennic layer is distinctly 
 divided into hexagonal facets. 
 
 Sometimes all these changes apparently take place a day 
 earlier ; but this is partly due to the change of colour in the 
 larval skin taking place much earlier in some cases than in 
 others. The imaginal discs are frequently unfolded to a 
 considerable extent before the pupa case is formed. All the 
 changes in the pupa are much affected by temperature, and I 
 believe they are accelerated by heat. 
 
 Section XIII. The Ocelli. 
 PLATE VIIL, Fio. 8 & 9. 
 
 The Ocelli are three in number ; they are situated in the 
 frontal region, close to the bend in the occipito-frontal plate. 
 The three cornese have not their axes in the same direction ; 
 but are situated on the sides of a small eminence, so that 
 their axes are divergent. Nevertheless they can all be seen 
 from above, so that a simultaneous picture is probably formed 
 on the three retina3, in certain positions, although not upon 
 corresponding portions in all the ocelli. 
 
 The cornese of the ocelli vary in diameter, from 1 -200th to 
 1 -400th of an inch. They are smaller in the males than in the 
 females, differing in this respect from the compound eyes. 
 They are very convex, the inner curve being slightly greater than 
 the outer, and their thickness varies, from 1 -400th to l-600th 
 of an inch. In the adult insect the protoderm is not easily 
 
90 ON THE ANATOMY OF THE FLY. 
 
 demonstrated over the ocelli, but from the manner in which 
 they are developed, their structure is probably similar to that 
 of the compound corneas (Plate VII., fig 9). 
 
 I think it quite probable that, in both the compound and 
 simple eyes of insects, the cornea consists of the protoderm and 
 mesoderm ; and the choroid or pigment layer of the modified 
 endoderm. If so, they are essentially tegumentary organs, 
 with the nervous structures piercing the endoderm, and ex- 
 panding between it and the mesodermic layer. 
 
 Immediately behind the cornea are several layers of nerve 
 cells ; at least, I have not been able to distinguish any inter- 
 mediate structure, although a fluid probably intervenes during 
 life. 
 
 The nerve cells form a cone, about l-600th of an inch in 
 depth, from the back of the cornea to the entrance of the optic 
 nerve. This is surrounded by a layer of hexagonal pigment cells, 
 filled with bright rose-coloured pigment, and containing well 
 marked nuclei. This choroid is surrounded by a plexus of 
 tracheal vessels (Plate VII., fig. 8). 
 
 The ocelli first appear about the fourth day of the pupa 
 state, and the pigment on their posterior surface is seen in 
 beautiful hexagonal nucleated cells about two days later. 
 During the whole pupa stage they are nearly twice as large as 
 they are in the adult insect. In the same way, the legs and 
 several other organs are larger when first formed, for condensa- 
 tion and contraction frequently accompany the development of 
 organs in the pupa ; but as the ocelli are flat at this period, 
 something must be allowed to the subsequent convexity of the 
 cornea. 
 
THE ANTENNJB. 91 
 
 Section XIV. The Antenna. 
 PLATE VIII., Fios. 8 TO 13. 
 
 The antenna consists of six joints, of which the third forms 
 the largest and most important part of the organ. 
 
 The first joint (Plate VII., fig. 9) is an irregular ring with four stiff 
 setse in front at its distal margin. 
 
 The second joint fits into the first and third. It consists of two parts. 
 The posterior (upper) is sub-conical, notched in front, and covered with 
 hairs, one of which on its outer side, is much longer than the rest. The 
 membrane uniting it to the first joint is rough, with exceedingly minute 
 hairs. The upper part of this piece fits into the cavity of the first joint 
 which partially conceals it. The anterior (lower) portion is an inverted 
 cone; its cavity opening posteriorly by a narrow ovate slit into the 
 cavity of the third joint. The integument is reflected back from the 
 margin of this slit to the base of the cone, and this reflection forms part of 
 the wall of the third joint. The external surface of the cone, and that of 
 the integument reflected back upon it, is covered with numerous minute 
 crenated plicae, directed backwards, not more than 1 -3000th of an inch in 
 breadth, and 1 -6000th of an inch in length their use is unknown. 
 
 The third joint (Plate VIII., fig. 8) is elongated and ovate. 
 Its surface is covered with minute hairs, between which are 
 a vast number of pellucid dots, about 17,000 or 18,000 on each 
 antenna, with about 80 large irregular spots, of a similar 
 character, along the inner edge of the joint.* 
 
 The smaller dots (Plate VIII., tig. 12) are about l-5000th 
 of an inch in diameter, or less. They consist of minute sacculi 
 opening on the surface, by an extremely small pore, not more 
 than l-8000th or l-10,000th of an inch in diameter. 
 
 * First described by Dr. Braxton Hicks, Linn. soc. trans., Vol. 22. My 
 calculations confirm their number according to the same author. 
 
92 ON THE ANATOMY OF THE FLY. 
 
 Dr. Braxton Hicks describes them as closed by a thin membrane, 
 which he calls the outer layer of the antenna, evidently my protoderm. 
 It appears to me however, from the most careful investigation, that the 
 protoderm is reflected into them, but forms a thin perforated collar 
 around the opening. The early period at which they appear in the pupa, 
 seems to indicate that they are mere sacculi of the surface ; the larger or 
 compound sacculi are moreover, in all probability similar organs, and un- 
 doubtedly open upon the surface. 
 
 The larger spots (Plate VIII., figs. 11 and 12) are compound, 
 consisting of several sacculi, with a common opening ; the 
 opening in these is often as large as l-1500th of an inch. The 
 several chambers are arranged in a radiating manner. 
 
 The fourth and fifth joints are mere rings, on the outer side of the third 
 joint, near its proximal extremity ; and the sixth forms a flattened plumose 
 tapering bristle, which is solid through the greater part of its length in 
 the mature insect. 
 
 The cavity of the first and second joints is chiefly occupied by cells, 
 probably of endodermic origin ; they are traversed by the antennal nerve. 
 
 The third joint is filled with pulp, which consists of cells 
 and large nuclei, which are especially abundant, together 
 with the terminal branches of the antennal nerve. This pulp 
 is surrounded by a layer of very delicate brown pigment cells, 
 which are immediately beneath the integument. 
 
 The antennal nerve splits up into numerous fibrilla3 on 
 entering this pulp, some of which terminate in ordinary nerve 
 cells ; but by far the greater number are covered with minute 
 corpuscles, clustered round their terminations, like bunches of 
 grapes (Plate VIII., fig. 13). A large tracheal vessel ac- 
 companies the nerve and ramifies in the pulp. 
 
 I believe the sacculi are minute expansions of surface, open 
 freely to the atmosphere, and that the entire third joint is an 
 olfactory organ. This seems probable, as the extensive ter- 
 
THE MAXILLARY PALPI. 93 
 
 minations of the nerve are brought into almost immediate 
 contact with the air, being only separated by the extremely 
 thin membrane of which the sacculi are composed ; and likewise 
 from the fact that the antennal nerves are given off from 
 ganglia, next in magnitude and importance to the optic ganglia. 
 Moreover, it appears more than probable that the olfac- 
 tory sense or some modification of it is, if not the dominant 
 sense, certainly only less important than vision to the insect. 
 The third autenual joint is equally developed in both sexes ; 
 but the habits of the male are not sufficiently understood to 
 justify any conclusions which might be drawn from this fact. 
 
 Section XV. The Maxillary Palpi. 
 PLATE III. e., AND PLATE VIII., FIG. 15. 
 
 The minute structure of the maxillary palpi is so like that 
 of the third joint of the antenna that its description naturally 
 follows the description of that organ. The integument of the 
 palpus is covered with short broad curved scales, about l-2000th 
 of an inch in length, as well as by the large hairs with which it 
 bristles. Besides these there are a number of sacculi, about 1,000 
 on each palpus, each opening by a pore in the centre of a collar, 
 like transparent ring, formed by the protoderm. Some of 
 these sacculi are as large as l-3000th of an inch in diameter, 
 with an external opening about I-5000th of an inch in 
 diameter, but others are as small as those of the antenna. 
 They are very closely set about the extremity of the palpus, 
 but are less numerous at its proximal end. The cavity of the 
 palpus is filled with cells, chiefly belonging to the endoderm. 
 A large nerve is distributed to it from the great nerre trunk 
 
94 ON THB ANATOMY OF THE PLT. 
 
 of the proboscis. Its filaments chiefly end in nerve cells in 
 the extremity of the organ, where the sacculi are most 
 numerous ; a tracheal vessel accompanies the nerve in its 
 distribution. It may be both an olfactory and gustatory 
 organ, directing the insect to its food. However this may be, 
 it can scarcely fail to be connected with the sense of taste, 
 seeing that it is often enveloped in a drop of fluid regurgitated 
 from the sucking stomach (see Section IV., page 55). 
 
 Section XVI. The Frontal Sac 
 PLATE I., AND PLATE VIII., FIG. 14. 
 
 The evidence that induces me to include this organ under 
 those of special sense is, that its extensive surface, which is 
 thrown into numerous plications, is still further increased 
 by being covered with blunt papillee (Plate VIII., fig. 14). Its 
 internal surface is covered with yellow pigment, and with a thin 
 layer of nerve cells connected with filaments from the anten- 
 nal nerves. As has been already stated, its cavity is in free 
 communication with the atmosphere, and the air contained 
 must be frequently changed by the vibrations of the antennal 
 plate. 
 
 With regard to the respective share which the antenna, 
 palpi, and frontal sac, probably have in the olfactory sense, we 
 have little to guide us ; but it appears to me likely, that the 
 olfactory organs of the antenna enable the insect to perceive 
 odors, which are quite inappreciable to the vertebrata. If, 
 as has been lately asserted, these are the organs which guide 
 
THE CEPHALO-8TEBNUM. 95 
 
 such insects as the male Kentish Glory moth to its female, 
 which may be miles away, their power of appreciating odors 
 would be little adapted for an insect when in the near vicinity 
 of putrid carrion. The palpi, as has been already remarked, 
 are probably gustatory organs, whilst I think it very probable 
 that the frontal sac, the nerve supply of which is far more 
 limited, and the structure of which is far coarser, is 
 especially adapted for the appreciation of more powerful 
 odors. 
 
 Section XVII. The Gephalo-Stcrnum. 
 PLATE V., FIGS. 89, AND PLATE VIII., Fio. 16. 
 
 The relations and position of this organ have been suffi- 
 ciently described.* Its blades are covered with numerous 
 sharp hairs or seta3. I can suggest no function which it is 
 likely this remarkable organ subserves. Its blades are not 
 more than l-200th of an inch in diameter, so that it can hardly 
 be concerned in the sound emitted by the insect when the head 
 is rubbed against the thorax, nor does its position seem adapted 
 to such a purpose. The only evidence that it is an organ of 
 special sense is that its inner surface is covered with nerve 
 cells. 
 
 * Section VI., page 60. 
 
96 ON THE ANATOMY OF THE PLY. 
 
 Section XVIII. The Halteres and Wing Organs. 
 PLATE VIII., FIGS. 3 TO 7. 
 
 The halteres are evidently modifications of the posterior 
 wings. The halter consists of three portions : the base, which 
 is covered with numerous special organs arranged in rows, 
 the pedicle, and the globe. 
 
 The base is connected with the thoracic integument. Three 
 hardened folds surround it, and have several slender muscles 
 inserted into them. These move the whole organ, which is 
 very movable. 
 
 The special organs at its base are arranged in four sets, two 
 on the upper and two on the under surface. Each set consists 
 of a series of curved ridges, under which these organs are 
 placed. Three of the sets are represented in Plate VIII., 
 fig. 3. The fourth is similar to the lower set, and is placed 
 opposite to it. 
 
 The proximal sets form segments of a hemisphere, and the 
 distal ones are arranged on two half-cylinders. 
 
 On making a section of the base, there appears to be but 
 one cavity. If the two half-cylinders are completed internally, 
 it is only by a thin membrane ; but as each seems to open 
 into the pedicle of the halter by a separate opening, I think 
 this is extremely probable. The ridges under which the 
 special organs are arranged are convex transversely, and each 
 of the grooves between them bears a single row of curved 
 hairs. These are close and long in the proximal sets, but 
 short, strongly curved, and distant in the distal groups. 
 
THE HALTEBES AND WING ORGANS. 97 
 
 The organs themselves have been described as openings in 
 the integument, closed by a thin convex membrane ; but the 
 appearance is deceptive, as far as I can tell, and I have 
 examined these parts with great care. There are no openings 
 in the integument, and the bright spots are entirely due to the 
 presence of lenticular corpuscles of high refractive power 
 beneath. By crushing the base of the halter in fluid between 
 the thin glass and slide, I have succeeded in seeing these 
 organs in profile, in situ (Plate VIII., fig. 6). I have been 
 unable to discover their exact connection with the integument, 
 but from the difficulty of displacing them, I suspect they are 
 contained in sacculi. When set free, which may be effected 
 by crushing the halter by a side motion of the thin glass, the 
 bright spots disappear from the ridges, which could not 
 happen if these were due to perforations ; and the corpuscles 
 appear free in the fluid. They are chiefly elongated, but a 
 few are spherical. They vary in size from l-3000th to l-5000th 
 of an inch in diameter. There are from 250 to 300 in each 
 of the basal, and about 200 in each of the proximal sets, 
 making nearly 1,000 in each halter. 
 
 The basal part of the halter becomes contracted, and forms 
 the pedicle, which is connected with each of the two half- 
 cylinders by a hollow ridge. The pedicle is clothed with 
 fine hairs, especially at its distal extremity. It terminates in 
 the globe, which is a hollow dilatation about l-80ibh of an inch 
 in diameter. The integument of the globe is extremely delicate 
 and transparent. It exhibits rounded pigment spots of orange 
 pigment beneath it, and has several large scattered hairs upon 
 its surface. 
 
 The interior of the globe is occupied by a number of large 
 vesicles about 1 -300th of an inch in diameter, filled with and 
 surrounded by fluid, which also fills the base and pedicle of 
 the organ. 
 
98 ON THE ANATOMY OF THE FLY. 
 
 The nerve of the halter, on entering the base, divides into 
 several branches, the greater part of which end in nerve cells 
 immediately below the corpuscles above described ; but a 
 large branch traverses the pedicle and ends in loops and nerve 
 cells in the globe. 
 
 Before entering into any speculation as to the function of 
 the halteres, I shall describe some closely allied structures, 
 which I have named the wing organs. These are corpuscles 
 exactly like those in the halter, arranged in rows and groups 
 on the sub-costal nervure of the wing (Plate VIII., fig. 7). 
 These corpuscles usually have a smoky tint, but I have no 
 doubt as to their identity with the structures I have described 
 in the halteres. They disappear in the same manner when 
 the nervure is crushed. I have not, however, been fortunate 
 enough to see them in profile. According to Dr. Braxton 
 Hicks, they are very constant in insects in the same part, as 
 well as on the sub-costal nervure of the posterior wings. They 
 are supplied with nerve filaments, like the organs at the back 
 of the halter. 
 
 These corpuscles, as well as those of the halteres, are opaque 
 when they first appear in the pupa, and I have frequently 
 observed those of the wing quite black and opaque in the 
 perfect fly, as if their development had been arrested. Each 
 of the wing organs seems to be situated under an oval convex 
 portion of the sub-costal nervure, corresponding to the ridges 
 of the halteres. Owing to their minute size, however, it is 
 almost impossible to be certain whether these are convexities 
 or depressions. There does not appear to be the slightest 
 thinning of the integument ai these points. 
 
 Looking only to the structure of these organs, I am inclined 
 most strongly to regard them as auditory. The corpuscles 
 are, I believe, otoconia. It is true, they are not acted upon by 
 
THE DALTEBES AND WING OBQANS. 99 
 
 acetic acid, but the otolitea of molluscs are not affected by 
 prolonged treatment with it.* I further think it extremely 
 probable that the globe of the halter receives the undulations 
 of sound, which are propagated through the fluid to these 
 corpuscles. 
 
 Perhaps in insects not provided with halteres, the wings 
 may receive the undulations of sound upon their whole 
 surface, which are so propagated, probably less perfectly to 
 the organs on the sub-costal nervure, which are very numerous 
 in most genera. 
 
 With regard to the position of the organs, bearing in mind 
 the rapid vibrations to which the integuments of the head 
 and thorax are subject during flight, and more especially during 
 the emission of sound, I conceive that the wings and legs are 
 probably the least agitated parts, and are hence fittest for the de- 
 velopment of auditory organs. During flight, the expansion of 
 the wing must evidently be the steadiest portion of the insect, as 
 it forms the point d'appui upen which it moves. The halteres 
 are certainly not much affected by the vibrations of the body ; 
 they are beautifully balanced by muscles, and are connected 
 only by membranous integument with the flank. 
 
 To conclude this subject : we know the auditory organs of 
 some saltorial Orthoptera are developed on their anterior, 
 femora, and, according to Dr. B. Hicks, the wing organs are 
 absent in the Orthoptera. Much careful work is needed 
 however, in examining these structures in various genera. 
 
 The halteres are developed, like the wings, from a distinct 
 imaginal disc. 
 
 * W. Burnett's Translation of Siebold, Sec. 211, Note ii. 
 
 H 2 
 
100 ON THE ANATOMY OP THE PLY. 
 
 Section XIX. The Folliculate or Ductless Glands. 
 PLATE IX., Fms. 2 AND 3. 
 
 The ductless glands are most abundant in the abdomen on 
 the dorsal surface of the viscera, near the dorsal vessel, and 
 
 DESCRIPTION OF PLATE IX. 
 
 Fig. 1. The dorsal vessel, x 12 diam. 
 
 Fig. 2. A portion of one of the folliculate glands from the abdomen 
 of a young fly, x 25 diam. 2 a, a portion of the same, x 100 diam. 
 2 6, a portion of the same, to show the ultimate distribution of the 
 tracheae and the cellular contents of the follicle, x 600 diam. 
 
 Fig. 3. A portion of one of the folliculate glands from the side of the 
 dorsal vessel of the larva. X 40 diam. 
 
 Fig. 4. The epitreme of one of the abdominal spiracles, x 100 diam. 
 Fig. 5. The internal generative organs of the male, x 15 diam. 
 
 Fig. 6. The four posterior abdominal segments of the male, showing 
 the loop and pouch in the position it occupies when the four segments are 
 partially exserted. The dotted line indicates the position of this part 
 when the segments are fully exserted. x 10 diam. 
 
 Fig. 7. A diagrammatic section of the last two segmeuts of the abdo- 
 men in the male, in which a represents the penis, b the ventral plate of 
 the penultimate segment, and c the dorsal portion of the last segment. 
 X 10 diam. 
 
 Fig. 8. The last two segments of the abdomen of the male, with the 
 ejaculatory organ and duct seen laterally, x 45 diam. 
 
 Fig. 9. The same seen from below, x 36 diam. 
 
 Fig. 10. The penis, with its palpi. X 60 diam. 
 
 Fig. 11. The apod&ne of the penis, viewed from above. X 50 diam. 
 
 Fig. 12. The extremity of the penis, x 100 diam. 
 
 Fig. 13. The ejaculatory organ, x 100 diam. 
 
 Fig, 14. Spermatozoids. X 100 diam. 
 
FlatclX 
 
THE MALE ORNEBATIVE OBGANS. 101 
 
 also at the posterior extremity of the abdomen in the vicinity 
 of the ovipositor in the female, and around the testes in the 
 male. A small pair exist behind the optic nerves, and a second 
 but still smaller pair may be observed adhering to the posterior 
 cornua of the pharynx. 
 
 They consist of arborescent follicles (Plate IX., figs. 2, 2a, & 
 2b), which adhere to the tracheal trunks which supply them. 
 The follicles are very transparent and structureless, but are 
 largely supplied with minute tracheal vessels. In the young 
 fly, they only contain nucleated cells, l-3000th of an inch in 
 diameter, free nuclei, and molecular matter ; but later in the 
 life of the insect they become loaded also with oil globules and 
 granular cells, especially in the female, sometimes to such an 
 extent that the abdomen becomes distended as if with mature 
 ova ; this fat disappears as the ova are developed. 
 
 A pair o folliculate glands* (Plate IX., fig. 3) are found in 
 the larva, one on either side of the dorsal vessel ; each follicle 
 contains one or two nuclei or nucleated cells. When the larva 
 arrives at maturity they become filled with granular matter, 
 and undergo disintegration before the formation of the pupa 
 
 Section XX. The Male Generative Organs. 
 PLATE IX. 
 
 The male organs (Plate IX., fig. 5) are a pair of testes 
 with their ducts, and a pair of albumen glands, which open with 
 
 * Zellenstrangen, Weismann. 
 
102 ON THE ANATOMY OF THE FLY. 
 
 the ducts of the testes into a common duct the vas deferens- 
 which terminates in an ejaculatory apparatus connected with 
 the intromittent organ by a very narrow ejaculatory duct. 
 
 The testes are surrounded by a thick layer of folliculate 
 glands, and are largely supplied with a plexus of nerves of 
 organic life. Each testis consists of a sac of structureless 
 membrane, lined with a bright reddish brown pigment mem- 
 brane formed of coalesced pigment cells. In the earlier stages 
 it contains cells in which the spermatozoa are developed, 
 but is afterwards entirely filled with spermatozoa. The testis 
 often presents an hour-glass contraction almost dividing it into 
 two cavities. The length of the organ in its long diameter is 
 about l-30th of an inch. The duct of the testis is narrow, 
 about l-30th of an inch in length, and lined with conical 
 epithelium . 
 
 The spermatozoa are very large, measuring l-100th of an 
 inch in length ; their large extremity is a flattened disc l-1000th 
 of an inch in diameter, but the filiform appendage is exceed- 
 ingly fine. 
 
 The albumen glands are sacs l-15th of an inch in length; they 
 are lined with cylindrical epithelium which secretes a very 
 coagulable fluid ; when the gland is mature its interior assumes 
 a bright brown colour like that of the pigment coat of the 
 testis, probably from an accumulation of pigment on the surface 
 of its epithelial lining. The albumen glands open directly 
 into the vas deferens by narrow necks. 
 
 The vas deferens is l-10thof an inch in length; it is widest 
 where it receives the ducts of the testes and albumen glands, 
 being nearly l-100th of an inch in diameter ; at its outlet, it has 
 only about half that calibre. It presents a very well defined 
 
THE MALE GENEBATIVE OBGANS 103 
 
 muscular coat of longitudinal and circular fibres, and is lined 
 with cylinder epithelium. 
 
 The ejaculatory organ (Plate IX., 13) is a peculiar chitinous 
 capsule, surrounded by a mass of muscles between the vas 
 deferens and ejaculatory duct. It is about 1-1 00th of an inch 
 in length in its longest diameter. It consists of a cup-shaped 
 cavity, the capsule, from the smaller extremity of which a 
 strong chitinous tube, the manubrium, projects. The manu- 
 brium terminates in a curved knob, which is perforated on its 
 convex side by six or eight small apertures, and marked by 
 two strong lateral ridges. These ridges are continuous with 
 a membranous sac, which forms the inner extremity of 
 the ejaculatory duct. The sac communicates with the in- 
 terior of the knob by the perforations on its convex side. 
 
 The vas deferens perforates the manubrium near its junction 
 with the capsule. The capsule is closed at its larger extremity, 
 and numerous bands of muscle which surround the whole organ 
 arise from the concave surface and edges of the knob-shaped 
 end of the manubrium, and are inserted into the edges of the 
 capsule close to its large extremity. 
 
 The ejaculatory duct is membranous, and not more than 
 l-1000th an inch in diameter ; it is about l-15th of an inch in 
 length from its commencement at the ejaculatory organ to its 
 termination at the end of the penis. The vas deferens, when 
 examined immediately after it ig removed, may be observed in 
 a state of vigorous peristaltic action, by means of which the 
 male fluid is, doubtless, conveyed to the ejaculatory organ. I 
 have not been able to satisfy myself as to the existence of a 
 muscular coat to the testes, but from the varieties of hour-glass 
 contraction they exhibit, and the large supply of nerves which 
 they receive from both senso-motor and sympathetic systems, 
 
104 ON THE ANATOMY OF THE FLY. 
 
 I cannot help suspecting their outer coat is contractile. The 
 ejaculatory organ is evidently an elastic sac within a muscular 
 covering, and aided by the peristaltic action of the vas def erens 
 undoubtedly ejects the male fiuid through the ejaculatory duct 
 with considerable force, as the mass of muscles is very large 
 indeed in proportion to the magnitude of the organ. 
 
 The four posterior abdominal segments (Plate IX., fig. 6 
 a. b. c. d.) form the external sexual apparatus of the male, and 
 are included within the genital fissure, which is formed by the 
 posterior margin of thejifth abdominal segment. This open- 
 ing is almost entirely closed by four valve-like pieces (Plate 
 IX., fig. 6 d), except during the sexual act, when the four 
 segments are partially protruded from the fissure, and the 
 valve-like pieces are directed backwards instead of forward. 
 
 The homologies of the several parts of these segments are 
 very difficult to make out, and will require a lengthened study 
 of the male and female organs of other insects before they 
 can be determined with certainty. 
 
 There is no difficulty in recognising the dorsal plates of 
 three segments, but the fourth seems to be represented only 
 above, by the inner pair of valve-like pieces already mentioned. 
 The ventral pieces are, however, considerably modified, and 
 two appear to be fused together, quite an exceptional condition, 
 for, as has already been remarked, the ventral are normally 
 more distinct than the dorsal plates. Another anomaly is 
 found in the unsymmetrical development of these plates, 
 affording the only example of the want of perfect lateral 
 symmetry in the integumental organs, and if the seminal 
 receptacles of the female are exCepted, in the structure of the 
 insect. 
 
 The sixth abdominal segment presents only a very narrow thin dorsal 
 
THB MALE GENEBATIVE OBOANS. 105 
 
 plate which seems to coalesce laterally with the seventh, at least on the 
 right side. The seventh and eighth are each very broad and strong, and 
 are minutely tuberculated or shagreened on their surface. The eighth 
 is deeply notched behind, the edges of the notch being produced laterally 
 to form the external pair of valve-like pieces already mentioned. The 
 internal pair of valve-like pieces are quite distinct from this plate, and 
 surround the extremity of the rectum which opens between them ; they are 
 united with the notch by a thin membrane. I believe they form part of 
 the ninth segment. 
 
 The junction of the sixth and seventh segments on the right side is very 
 considerably thickened, and supports, or more correctly these segments 
 are supported, by a broad process which becomes gradually narrower, and 
 curving round their ventral aspect forms a strong chitinous loop ; a very 
 rudimentary process exists on the left side, but never joins this loop. 
 Numerous muscular bands are inserted into the edge of the loop, and a 
 strong muscle is attached to its thick extremity. The loop forms a hinge 
 upon which the four posterior segments move. It is united loosely by 
 membrane to the ventral plate of the fifth and to the ventral plate of 
 the eighth segment. The membrane between it and the eighth segment 
 forms a pouch, which contains the male organ when the segments 
 are retracted ; the loop then lies along the inner margin of the notch in 
 the fifth abdominal ventral plate (Plate DL, fig 6 represents the position 
 of the loop both during the exsertion and retraction of the four posterior 
 segments). 
 
 The eighth segment is continuous with the outer valve-like appendages 
 (Plate IX., figs. 8 & 9), which are curved and covered upon their 
 outer surface with hairs directed forwards ; they probably embrace 
 the ovipositor during copulation.* Externally each is prolonged for- 
 ward as a slender but strong-curved rod, which unites it with a heart- 
 shaped ventral plate. The rods lie along the lateral margins of the genital 
 fissure (Plate IX., fig. 9), and the ventral plate with which they are 
 united, lies within its anterior border ; it ia united by the membranous 
 pouch with the loop. 
 
 The posterior prolongations of the dorsal plate are probably not to be 
 
 * I have never been able to observe the sexual act in the blow-fly. 
 When kept in captivity I have never known them either to perfect ova or 
 spermatozoa, neither have I been able to hear of anyone having observed 
 the act. By comparing the organs carefully, I have attained the results 
 which I have set down as the probable relations and uses of the various 
 parts concerned. 
 
106 ON THE ANATOMY OF THE FLY. 
 
 regarded as the dorsal lateral appendages, as they are continuous with the 
 plate itself as well as with the lateral rods. But the ventral plate of this 
 segment is deeply notched on either side behind, and a pair of curved hollow 
 nipple-like organs, each bearing several setee, are articulated with these 
 notches; they evidently represent the inferior lateral appendages and 
 probably curve over the leaf -like organs on the dorsal aspect of the 
 ovipositor. The dorsal prolongations of this segment act as springs ; when 
 at rest they are directed forward, and lie with their broad surface in a 
 horizontal position ; when the segments are partially protruded, they are 
 turned back with their broad edges directed vertically ; this is effected by 
 muscular action through the medium of the lateral rods. 
 
 The remainder of the sexual apparatus consists of a single segment. 
 The internal valve-like pieces which surround the rectum are probably its 
 dorsal lateral appendages ; this will appear more likely by comparing 
 them with their homologues in the female which are separated by a small 
 triangular dorsal plate. (Plate X., fig. 5.) 
 
 The intromittent organ (Plate IX., fig. 10) probably represents the ventral 
 lateral appendages of the same segment, fused more or less into one. It 
 presents perfect lateral symmetry, and is supported by a hinge- joint upon 
 a very large apode"me (Plate IX., fig. 11) ; each lateral half is articulated 
 separately with this apod6me, and each bears at its base a palpus-like 
 organ articulated with it by a distinct ginglymus joint. The two lateral 
 halves are united by a curved spine or crest, which arches over the ejacu- 
 latory duct. 
 
 Each lateral half consists of a slightly curved hard rod of chitine, broad 
 at its base, but gradually becoming thin towards its extremity, which is 
 terminated by a small triangular knob. About half way between the base 
 and apex of each of these rods on its under side is a proj ecting process 
 which bears a curved plate, serrated along its inferior border, and uniting 
 with its fellow in the mesial line. These plates terminate in pointed pro- 
 cesses at their outer posterior angle, but their middle portion is continued 
 backward, and turns up to surround the ejaculatory duct, which lies upon the 
 upper surface of these plates except at their posterior extremity, where 
 they form a complete tube, united anteriorly with the duct, and opening 
 posteriorly in a broad trumpet-shaped mouth. (Plate IX., fig. 12.) 
 
 The intromittent organ is moved by a mass of muscles arising from the 
 handle-like apod&ne which supports'it ; these are inserted into the lateral 
 halves of the organ as well as into the bases of its palpus-like appendages 
 other muscles arise from the eighth ventral plate, and are inserted into 
 the apode"me. 
 
 Unless the apodSme represents the ventral plate of the ninth segment, 
 
THE FEMALE GENERATIVE OBGANS. 107 
 
 which I strongly suspect it does, this is only represented by its four lateral 
 appendages. 
 
 The intromittent organ is l-30th of an inch in length, which corresponds 
 with the distance between the external opening and the neck of the bursa 
 in the female. 
 
 The relations of the crest and palpus -like appendages during the sexual 
 act, are, probably, as follows : The crest is brought into relation with the 
 interior of the last joint of the ovipositor, distending the extremity of the 
 oviduct, whilst the palpus-like organs are in relation with its lateral 
 surfaces externally. 
 
 The palpus-like organs are apparently sensory. They are hollow, 
 slightly curved, and terminated by thin rounded spoon-shaped blades 
 the concave or posterior border is furnished with a single row of stiff hairs. 
 The base presents a welUmarked articular surface formed of two parts at 
 right angles with each other. 
 
 Section XXI. The Female Generative Organs. 
 PLATE X., Fios. 17. 
 
 The female generative organs (Plate X., fig. 1) consist of 
 the ovaries and their ducts, the common oviduct, the bursa, the 
 seminal receptacles, the albumen glands, and the ovipositor. The 
 ovaries are circular discoid masses of ova, situated at the sides of 
 the abdomen, and largely supplied with tracheal vessels from the 
 second abdominal spiracles. They consist of numerous follicles 
 which terminate the oviducts, each follicle containing a series 
 of ova, one behind the other, in various stages of development. 
 The follicles are so arranged that the ova are packed side by 
 side with their long diameters vertical to the surfaces of the 
 disc of the ovary. The oviduct commences upon the inner surface 
 of the disc by the union of the follicles, and the outer surface 
 
108 ON THE ANATOMY OP THE FLY. 
 
 DESCRIPTION OF PLATE X. 
 
 Fig. 1. The internal generative orgaes of the female seen from below 
 only a portion of the ovaries is represented, from a preparation stained 
 with carmine, x 15 diam. 
 
 Fig. 2. The common oviduct and its appendages, viewed laterally, to 
 to show its muscles and bursa. x 30 diam. 
 
 Fig. 3. The ovipositor seen laterally, during the deposition of an egg. 
 X 30 diam. 
 
 Fig. 4. Plate between the rectum and ovipositor, x 30 diam. 
 Fig. 5. The last segment of the ovipositor, x 30 diam. 
 Fig. 6. The bursa copulatrix. x 40 diam. 
 
 Fig 7. The seminal receptacles of the female from the right side 
 X 60 diam. 
 
 Fig. 8. One of the ovarian tubes or follicles exhibiting four ova in 
 various stages of development a the germinal vesicle surrounded by 
 epithelium -like cells ; b the first stage of yolk segmentation ; c the 
 second stage of yolk segmentation, d the segmented yolk broken up and 
 surrounded by a zona pellucida. X 60 diam. 
 
 Fig. 9. Yolk cells undergoing degeneration, a mature yolk cell, 
 b c & d the same undergoing degeneration, x 250 diam. 
 
 Fig. 10. Upper part of an ovum, showing the first stage of the for- 
 mation of the germinal disc and yolk canal, x 60 diam. 
 
 Fig. 11. The ovum just before impregnation, x 60 diam. 
 
 Fig. 12. The upper extremity of the same, showing the micropyle, 
 germinal disc, and upper part of the yolk canal, x 150 diam. 
 
 Fig. 13. The ovum, eight hours after impregnation, showing the larval 
 integment in the interior. X 60 diam. 
 
 Fig. 13a. Cells from the larval integment of the same age. x 500 diam. 
 
 Fig. 14. The ovum, twenty-four hours after impregnation, showing the 
 tracheal system of the larva just before its escape from the egg. x 60 
 diam. 
 
 Fig. 15. A small portion of the ovicapsule at the time of impregnation. 
 X 200 diam. 
 
 Fig. 16. Cells from the fat bodies of the adult larva, x 30 diam. 
 a, the same in the first stage of degeneration. 4, c, growth -corpuscle formed 
 from the same, and nucleated cells originating in the growth-corpuscles, 
 The former x 30, the latter X 300 diam. 
 
f.T.Z.&ir&S 
 
THE FEMALE GENERATIVE OBQANS. 109 
 
 is covered by a dense mesh of trachea! vessels. When the 
 insect first emerges from the pupa, the ovaries are not l-30th of 
 an inch in diameter, but when the ova are mature they occupy 
 the greater part of the abdominal cavity ; for the mature ova 
 are l-20th of an inch in length, and on the most moderate com- 
 putation, each ovary contains 300 mature ova, so that a single 
 female is capable of producing at least 2,400 ova, supposing 
 the four sets in the ovary to be matured in succession. 
 
 The ovarian follicles appear to consist of structureless 
 membrane, largely supplied with delicate trachea! vessels, and 
 the oviducts in the young insect consist entirely of an exceed- 
 ingly fine membrane. When mature, they exhibit a thick 
 muscular coat of transverse and longitudinal fibres. They 
 unite with each other at the base of the ovipositor, and are 
 continued to its extremity as a common oviduct. 
 
 About l-20th of an inch from the extremity of the ovipositor, 
 the albumen glands and seminal receptacles open into the 
 common oviduct. 
 
 The albumen glands are simple sacculate glands, l-15th of an 
 inch in length, which open into the oviduct by narrow necks. 
 They secrete a viscid fluid, which covers the ova after impreg- 
 nation, rendering them opaque. It is quite insoluble in water, 
 but is immediately dissolved by aether, rendering the ova 
 again transparent. 
 
 The seminal receptacles (Plate X., fig. 7) are three in 
 number, two upon ihe right and one upon the left side. Each 
 consists of an ovate dark brown chitinous capsule, measuring 
 about 1-1 20th of an inch in its long diameter, connected with 
 the oviduct by an elastic tube, l-20th of an inch in length, 
 continuous with the wall of the capsule. 
 
110 ON THE ANATOMY OF THE FLY. 
 
 The capsule exhibits well marked rings of chitine and numerous nuclei. 
 It is easily ruptured by pressure. The elastic tube connecting the capsule 
 with the oviduct has a very thick wall, of a light yellow tint, marked 
 with transverse wrinkles. The calibre of the tube is greatest in the 
 middle, where its diameter does not exceed l-400th of an inch. It tapers 
 gradually to a very fine opening at either extremity. Numerous small 
 cells may be observed adhering to the duct, which give it a beaded 
 appearance. They are probably concerned in the development of the 
 muscular coat which surrounds both the capsule and its duct, but which is 
 not well developed until the ova are mature. 
 
 A little below the openings of the albumen glands and 
 seminal receptacles, about l-30th of an inch from the outlet of 
 the common oviduct, its membranous coat forms a pouch in 
 its ventral wall. This pouch is called the bursa copulatrix (Plate 
 X., figs. 1, 2, and 6). It is about l-80th of an inch in diameter. 
 The bursa has a narrow neck, which opens backwards into 
 the oviduct. It exhibits two curious lateral folds, one on 
 each side of the neck, which project into its interior, as if the 
 wall had been pushed inwards, and a third fold in the mesial 
 line, which forms a partial septum, dividing the pouch into 
 two lateral halves. The bursa is very elastic, resuming its 
 shape after dilatation or compression, and refracts light 
 powerfully. The neck of the bursa is studded with minute 
 corpuscular bodies, with a bright transparent centre. These 
 are not more than l-5000th of an inch in diameter, and may 
 be either papillse, connected with nerves, or merely centres 
 of chitinous deposit. I am decidedly inclined to the latter 
 view. 
 
 The arrangement of the muscular fibres of the outer coat of the oviduct 
 below the entrance of the ducts of the albumen glands and seminal recepta- 
 cles, needs special notice. The muscular coat is very highly developed (Plate 
 X , fig. 2), and its outer layers form broad bands of muscle, chiefly having an 
 oblique direction round the oviduct. A well marked broad band surrounds 
 both the bursa and oviduct in a transverse direction, and conceals other layers 
 which have an oblique direction ; and others again, which surround only the 
 bursa or oviduct separately. A pair of large muscles, which arise from 
 the dorsal part of the second segment of the ovipositor the seventh ab- 
 
THE DEVELOPMENT OP THE OVUM. Ill 
 
 dominal segment are inserted into the oviduct before and above the bursa, 
 and a second pair from the same part of the eighth abdominal segment 
 are inserted into the oviduct between the bursa and its outlet. The 
 internal or membranous coat of the oviduct is continuous with the 
 integument at the extremity of the ovipositor. 
 
 The function of the bursa is clearly to receive the male 
 fluid, but a few hours after copulation all the spermatic fila- 
 ments disappear from both the bursa and the oviduct. They 
 are then found in a dense mass, closely coiled together in the 
 capsules of the seminal receptacles. Nothing, then, remains 
 in the bursa but the two masses of coagulated albumen, one of 
 which occupies each lateral half. These gradually disappear, 
 either by deliquesence or absorption. The manner in which 
 the spermatic filaments are transferred from the bursa to the 
 receptacles is unknown, but the transference is probably 
 effected by their own movements. Upon one occasion I found 
 the upper part of the oviducts of a fly, which had been im- 
 mersed in spirit for four-and- twenty hours, filled with sper- 
 matic filaments, which were still in a state of active movement. 
 I suppose the insect had been captured immediately after 
 copulation, and that the spirit had not permeated the oviducts, 
 owing to the coagulation of the albumen around them, so that 
 the filaments, no longer directed into the ducta of the sper- 
 matic receptacles by the contraction of the oviduct above 
 their openings, found their way into the upper part of the 
 oviducts.* 
 
 The external organ in the female consists of a lengthened 
 ovipositor, which is retracted when at rest within the posterior 
 
 * In Burnett's Translation of Siebold, p. 450, note v., the spermatic 
 filaments are stated to be conveyed from the bursa of insects to the recep- 
 tacles by their ovrn movements, which agrees perfectly with the result of 
 my researches. 
 
112 ON THE ANATOMY OP THE FLY. 
 
 opening of the fifth abdominal segment. It consists of four 
 segments, the anterior three of which are separated by mem- 
 branous interspaces equal in length to the segments, and when 
 the organ is retracted each segment is withdrawn within the 
 next anterior to it, like the draw-tube of a telescope. The 
 posterior segment is attached closely to the penultimate, and, 
 like the ninth abdominal segment in the male, is considerably 
 modified, and bears a pair of leaf -like appendages which termi- 
 nate the rectum. 
 
 The first and second segments of the ovipositor are like the abdominal 
 segments ; they consist of a dorsal and ventral plate, and have a fringe of 
 long vibrissas-like setae along their posterior margin. The third consists of 
 three plates, two of which are dorsal or lateral, and the third is ventral ; 
 it is slightly cleft posteriorly, and its extremity is covered with hairs. 
 
 The fourth segment (Plate X., fig. 6) consists of a small triangular dorsal 
 plate which bears a pair of leaf -like lateral appendages covered with hairs. 
 These surround the dorsal part of the opening of the rectum. The intestine 
 is separated from the oviduct by a curved semicircular plate (Plate X., fig. 
 4), the angles of which are anterior and prolonged as apode"mes which 
 give insertion to a pair of muscle?, which arise from the ventral plate of 
 the third segment, and embrace the oviduct. Its curved border is posterior 
 and is fringed with hairs. It is probably the ventral plate of the last 
 abdominal segment ; if such is the case, the oviduct opens between the 
 last and penultimate segments of the ovipositor on its ventral aspect. 
 
 The development of the sexual organs commences in the 
 larva, and so presents an exception to the other organs of the 
 imago. The ovaries or testes are found as a pair of cellular 
 bodies adhering together on the under side of the rectum, near 
 its termination. Dr. Weismann* figures both male and female 
 organs in this condition, and asserts that even at that early 
 period they present distinctive characters. I have been 
 unable to verify this assertion, although I believe it is likely 
 that Dr. Weismann is quite correct in this, as in most of his 
 elaborate researches. 
 
 * Kolliker's Zeitschrift. Band XIV. 
 
THE DEVELOPMENT OP THB OVUM. 113 
 
 Section XXU.The Development of the Ovum. 
 PLATE X., Fias. 8 TO 14. 
 
 The ovum in the earliest stage of development in which I 
 have yet succeeded in observing it, is ovate in form, and mea- 
 sures about l-300th of an inch in its long diameter. It 
 consists of a single cell, the germinal vesicle, which occu- 
 pies the greater part of its interior, and is surrounded by a 
 thin layer of clear fluid. This is covered by an external en- 
 velope of minute cells the yolk sac 1-30 00th of an inch in 
 diameter, loosely adherent to each other, and contained in an 
 extremely delicate membrane the ovi-capsule. These cells 
 refract light strongly, and make it very difficult to observe the 
 germinal vesicle. 
 
 The ovum soon after becomes globular; the external cells 
 adhere firmly to each other, and become prismatic by pressure ; 
 and the yolk, instead of containing a single vesicle, consists of 
 six to eight cells, which fill the whole interior of the yolk sac. 
 I am doubtful whether this multiplication of cells is analogous 
 to yolk-segmentation in vertebrates, but I suspect that it is. 
 
 The ovum now enlarges rapidly, assumes an ovate form, and 
 the yolk cells become spheroidal, multiplying until about 
 thirty-two are contained within the ovum (Plate X., fig. 8 c), 
 after which they degenerate into a homogeneous molecular 
 mass. A clear space is then gradually formed at the upper 
 part of the yolk, which is the representative of the germinal 
 disc in vertebrates. 
 
 The changes which take place in the yolk cells are very remarkable : at 
 first they exhibit a cell wall and a large nucleus, but after a time the 
 
114 ON THE ANATOMY OF THE PLY. 
 
 contents of the cell form a well-marked mass around the nucleus separated 
 from the cell wall by a clear space (Plate X., fig. 9 a). Sometimes one or 
 more small shining nuclei appear in the clear space, and the central mass 
 becomes nebulous, and degenerates into molecular granules. Occasionally 
 it becomes segmented into two or four, previously to its degeneration (Plate 
 X., fig. 9, b, c, d). Ultimately the cell wall disappears, and the whole 
 yolk becomes homogeneous (Plate X., fig. 8d, represents the yolk during the 
 degeneration of its segmented cells). 
 
 The ovum has by this time attained its full size. The ovi- 
 capsule becomes separated from the cellular layer or yolk sac be- 
 neath, by a clear area corresponding to the zona pellucida, and the 
 wedge-shaped cells of the external layer of the yolk, the yolk 
 sac, become converted into hexagonal chitinous cells of pave- 
 ment epithelium. (Plate X., fig 15). 
 
 The ovicapsule forms a bladder-like projection at this period 
 at the upper extremity of the ovum, immediately above the 
 germinal disc, and a groove commencing at the germinal disc 
 is gradually formed between the yolk sac and the yolk. (Plate 
 X., fig. 10). The germinal disc and the groove or yolk canal 
 may now be observed to be surrounded by rapidly growing 
 cells ; and a group of cells appears at the upper extremity of 
 the germinal disc. This may be distinctly seen to surround a 
 minute canal the micropyle by means of which the germinal 
 disc communicates perhaps directly with the inflated portion 
 of the ovicapsule. These changes take place before impregna- 
 tion. 
 
 The groove or yolk canal is the only part within the yoke 
 sac which remains unaltered after impregnation, and it may 
 be observed, even after the larva has escaped from the ovum, 
 as a distinct line on the interior of the yolk sac. 
 
 The length of time required for the development of the 
 ovum up to this point is unknown ; for, although I have 
 
THE FORMATION OF THE PUPA. 115 
 
 repeatedly kept flies for the purpose of discovering it, I have 
 never succeeded in getting the ova developed in captivity, per- 
 haps from the want of a proper kind of food. However well fed, 
 the insects only laid by large stores of fat in the ductless 
 glands. 
 
 Impregnation is usually effected immediately before the eggs 
 are laid, the spermatozoa being probably discharged from the 
 spermatic receptacles upon the upper extremity of the ovum, 
 and brought into immediate contact with the whole length of 
 the yolk through the medium of the micropyle and yolk canal. 
 Immediately after the discharge of sperm, or simultaneously 
 with it, a viscid fluid is poured upon the ovum probably from 
 the albumen glands, which renders it glossy and opaque. The 
 act must be rapidly performed, for although the ova are usually 
 extruded from tke ovipositor in quick succession, none escape 
 impregnation. 
 
 -Soon after impregnation, a layer of rapidly- growing cells 
 appears in the place of the germinal disc, and along the inner 
 surface of the yolk canal ; this layer forms tke ventral portiom 
 of the integument of the larva. At the end of eight hours it 
 kas extended so as to aurroami the entire yolk. The whole 
 integument is then sketched out by this oellwkr layer, and ita 
 segments are divided by well-marked folds* 
 
 At this period, or foen after, tfee alimentary canal appears 
 sketched out by a cellular membrane the inaer or mucous 
 layer of the cells which surround the yolk. At first ft 
 is straight and very capacious, i&clmding the whole yoik, 
 but it becomes gradually narrow and tortuous during the 
 formation of intermediate layers of cells, from which the other 
 internal organs are developed. When the larva is hatched, * 
 still contains the remains of the yolk. 
 
116 ON THE ANATOMY OP THE PLY. 
 
 Beyond these facts, I can add little with any degree of cer- 
 tainty, except that twenty-four hours after impregnation, the 
 newly developed tracheal vessels of the embryo become suddenly 
 filled with air, and the larva commences to struggle in the egg, 
 but soon ruptures the ovisac and escapes. 
 
 Dr. Weismann has traced the development of the embryo larva far more 
 minutely than I have, in his " Entwicklung Der Dipteren ;" but I confess I 
 do not know how he has attained such well-defined results from the very 
 delicate cellular structures of the embryo, nor do his figures appear to me 
 so accurate as those of the imaginal discs, and the development of the 
 pupa. 
 
 Section XXIII. The Formation of the Pupa. 
 PLATE VI., FIGS. 1 TO 4. 
 
 The nature of the changes which take place during the 
 formation of the pupa have been already briefly indicated. 
 They closely resemble those in the Echinodermata, where 
 the perfect form is developed in the interior of the larva. 
 But the change is even more remarkable in the fly, as 
 all the tissues of the larva undergo degeneration, and the 
 imaginal tissues are re-developed from cells which originate 
 from the disintegrated parts of the larva, under conditions 
 similar to those appertaining to the formation of the embryonic 
 tissues from the yolk. 
 
 The formation of the pupa^ may be conveniently studied 
 under two heads the formation of the pupa skin, and the 
 transformation of the larval tissues. 
 
 The pupa skin is partly formed from seven pairs of delicate 
 
THE FORMATION OP THE PUPA. 117 
 
 cellular expansions, called by Dr. Weismann imaginal discs, 
 and partly from cells formed upon the inner surface of the 
 larval integument. When mature, it is an extremely delicate, 
 apparently structureless transparent membrane, presenting no 
 trace of its cellular origin, and, after the escape of the imago, 
 remains as a fine silvery membrane in the interior of the pupa 
 case. 
 
 The imaginal discs have received the following names from Dr. Weis- 
 mann, which I have retained, and simply translated into English, the 
 antennal and optic discs* which enter into the formation of the head, and 
 the inferior pro-thoracic and meso-thoracic discs f which are chiefly con- 
 cerned in forming the anterior and intermediate limbs. These are all 
 attached firmly through the medium of short nerve trunks to the nerve 
 centres of the larva. (Plate VI., fig. 1 ) 
 
 The superior meso-thoracic disc, J and the superior and inf erior meta- 
 thoracic || discs are firmly attached to the lateral tracheal vessel near the 
 anterior spiracle. The superior discs are chiefly concerned in forming the 
 wings and halteres, and the inferior in the formation of the posterior limbs. 
 (Plate VI., fig. 2) . Beside these a horse-shoe shaped arch of similar 
 tissue is attached to a pair of tracheal vessels, in front of the supra- 
 cesophageal ganglia, and above the antennal discs. It is called the ring 
 by Dr. Weismann, but I believe it is the upper pro-thoracic disc. 
 
 The abdominal portions of the pupa skin are formed from cells developed 
 upon the inner surface of the corresponding larval segments. 
 
 Each imaginal disc is enclosed in a capsule of structureless 
 membrane. The capsules of those attached to the nerve centres 
 appear to consist of the neurilemma of the nerves which sup- 
 port them, but the capsules of those attached to the trachea 
 seem to be derived from the outer coat of the tracheal vessels. 
 The discs are formed of minute cells, having a granular appear- 
 
 * Autennen and Augensheiben. 
 
 t Unter prothoral scheibe,and meso-thoral sheibe. 
 
 J Ober meso-thoral sheibe. 
 
 II Ober and unter meta-thoral eheiben. 
 
118 ON THE ANATOMY Ofl THE 
 
 ance. The rapid development of the discs ruptures the capsules 
 which contain them ; but during their development, before the 
 rapture of the capsule, they present the appearance of spiral 
 folds and concentric rings the optic and superior thoracic 
 discs having the former, and the inferior thoracic discs having 
 the latter kind of folds. ^The antennal discs present the appear- 
 ance of concentric rings at their anterior extremity. 
 
 The unfolding of the concentric ring-like folds is exceedingly 
 curious, the internal ring being pushed out first, the rest 
 following in succession, These form the joints of the limbs 
 and antenna 
 
 The development and unfolding of the inferior thoracic discs is repre- 
 sented in Plate VL, figs. 1, 3, and 4. At first the several discs are 
 separated entirely from each other, but as they are developed and enlarged, 
 the discs belonging to each pair become closely united by the common 
 origin of their capsules (fig. 3). The concentric rings each form a tarsal 
 joint, and other rings which are developed later, form the tibia, femur, and 
 ooxa> After the rupture of the capsule, the disc presents the appearance 
 represented in fig. 4. 
 
 The superior thoracic discs, after rupturing their capsules, form sacs* In 
 the interior of which the wings and halteres are developed. 
 
 The pupa covering of the thorax is probably formed entirely from these 
 discs ; but according to my observations the inferior meso- and meta- 
 thoracic discs take no part in the formation of the proper thoracic walls* 
 These seem to be formed entirely from the inferior pro-thoracic and 
 superior meso- aud metathoracic discs. If such is the case, my views con- 
 cerning the number of thoracic segments would acquire confirmation. The 
 pro-thorax being formed from the pro-thoracic discs ; the anterior spiracle, 
 which opens between the pro-thorax and meso-thorax, would belong to the 
 next thoracic segment, undeveloped in the imago, but forming the sixth 
 segment in the larva, and bearing, its anterior spiracles, whilst the 
 superior meso- and meta-thoracic may belong to segments which alternate 
 with the inferior thoracic discs. 
 
 The antennal and optic discs form the pupa covering of the head, and the 
 optic discs present a strongly-marked spiral fold before the rupture of 
 their capauJ.es. 
 
TSI FORMATION OF THE PUPA. 119 
 
 The semi-fluid cellular matter, from which the imago is 
 developed, is derived partly from the disintegrated tissues of 
 the larva, and partly from the fat bodies or omenta. 
 
 After the larva ceases to feed, if the temperature be suffi- 
 ciently high, the tissues begin to degenerate. The muscles 
 may be observed at this time in a state of continuous activity, 
 rythmic contractions commencing at one extremity of each set 
 of fibres, and passing regularly with a wave-like motion to the 
 opposite extremity. At the same time, large bright nuclei, 
 1-1 000th of an inch in diameter, appear in rows in the centre 
 of the muscular fibres. These are ultimately set free by the 
 degeneration and waste of the muscles, and exhibit a granular 
 appearance, but are readily distinguished by their great trans- 
 parency and low refractive power. 
 
 At the same time a series of remarkable changes take place 
 in the fat bodied, which consist in the adult feeding larva of 
 flattened hexagonal jells filled with very opaque, highly refrac- 
 tive white granular matter. These cells now begin to exhibit 
 clear spaces in their centre, which presently become converted 
 into nuclei exactly like those formed in the muscles. The granu- 
 lar matter of the omental cell then becomes condensed about the 
 nucleus, leaving a clear space around the circumference of the 
 cell; the cells separate from each other, and the cell wall 
 undergoes disintegration. 
 
 The free nuclei developed from the muscular fibres of the 
 larva now begin to collect around them aggregations of mole- 
 cular matter, derived from the degeneration of the muscles 
 and other larval tissues, so that all the nuclei are soon sur- 
 rounded by similar molecular aggregations, each about l-150th 
 of an inch in diameter. 
 
 The precise nature of the changes which take place imme- 
 
120 ON THE ANATOMY OP THE FLY. 
 
 diately afterwards are more difficult to observe, but after the 
 second day of the pupa state, numerous delicate nebulous- 
 looking cells, about l-1000th of an inch in diameter, replace 
 some of these aggregations, and bright nuclei, l-3000th to 
 l-5000th of an inch in diameter, make their appearance amongst 
 them. The majority of the aggregations remain, however, 
 and become more dense toward their circumference. The 
 growth of the imaginal tissues evidently proceeds at the ex- 
 pense of some of these aggregations, whilst those which remain, 
 undergo marked changes ; they increase in size, lose their 
 original nuclei, and become invested by a delicate membrane. 
 When the imago emerges from the pupa, a large number of 
 these corpuscular aggregations remain in all parts of the 
 insect ; they disappear during the development of the imago, 
 and when it is mature, not one can be detected. I propose the 
 term growth-corpuscles to designate these remarkable bodies. 
 
 The growth-corpuscles of the imago, when it emerges from 
 the pupa, vary from l-80th to l-200th of an inch in diameter. 
 They contain molecules and granules of various sizes, besides 
 numerous nucleated cells, l-1000th of an inch in diameter, with 
 a well-defined cell wall, and from one to three large nuclei. 
 The exact manner in which these corpuscles are concerned 
 in the development of the tissues is, at present, unknown ; their 
 functions may be similar to those of the folliculate glands, the 
 preparation of the nutritive fluid or formative plasma. I do not 
 think they are directly converted into tissue either in the pupa 
 or imago. 
 
 All these changes are much influenced by temperature, and 
 the entire development of both the mature larva and the pupa 
 in all its stages is immediately stopped when the temperature 
 falls below 45 Fahrenheit. This circumstance alone preserves 
 the species through the winter, all the larvae and pupse existing 
 
THK FORMATION OF THR PUPA. 121 
 
 being entirely checked in their development, undergo hy- 
 bernation. Their dormant vitality remains, however, to be 
 awakened by the warmth of spring. 
 
 When the imago is ready to burst from the pupa case, the 
 anterior extremity is forced off by the dilatation of the frontal 
 sac. This is effected by the contraction of all the muscles then 
 existing, which forces the greater part of the circulating fluid 
 into the frontal sac, and distends it until it far exceeds the 
 head in size. At this period a pair of large muscles extend 
 from the edges of the frontal sac to the margin of the occipital 
 foramen. They assist in the dilatation of the sac by diminish- 
 ing the cavity of the head by their contraction. They after- 
 wards undergo degeneration, and may be looked for in vain in 
 the mature insect. 
 
 FINI3. 
 
 THOMSON AND FINDER, PRINTERS, BOLT COURT, FLEET STREET.