The silken substance is then pressed by the more or less powerful contractions of the muscles of the filator, so that the passage of the threads is facilitated. If the muscles totally contract, the spinning canal is opened wide, the threads pass easily upwards and assume the form of a triangular prism (Fig. 336).

If this contraction diminishes, the chitinous wall of the spinneret comes together, owing to its elasticity; the ceiling of the canal approaches the floor; the cavity tends to take the form of a semicircular slit, and the threads are compressed, flattened. As each mass or thread of silk is much more voluminous than the canal, except when the latter is extremely dilated, it follows that the two threads are always compressed, or squeezed together, and that each of them is compelled to mould itself in the groove it occupies and to take its shape. Hence the variations in the appearance of the two masses or divided portions of silk, which as stated present all grades between the form of an isosceles-triangular prism and that of a nearly flat ribbon; but this last case is quite rare. The use of the spinneret, then, is to compress the thread and to change its form more or less considerably, at the same time as it diminishes its diameter.

Moreover, this constant compression of the thread as it passes through the press keeps it in a certain state of tension so as to allow the caterpillar while spinning to firmly hold its thread.

Finally, when the worm suspends the contraction of its spinning muscles, the press flattens, vigorously compresses the thread, and arrests its motion, in such a way that if there was a strain on the silken fluid (bave), it would break rather than oblige the caterpillar to let go any more of it.

The press does not act directly on the silken thread, but through the gummy layer (grès) which transmits over the whole surface of the silken fluid (brin) the pressure exerted on it. After having overcome this difficult passage, the silk thread has acquired its definite form; it rapidly passes out of the spinneret.

How the thread is drawn out.—Having seen, says Blanc, how the two masses of silk (brins), in passing through the spinning apparatus (or press), join each other, constituting the frothy silken fluid, thus becoming modified in form, it remains to examine the way in which the thread is drawn out of the spinneret. If we examine a caterpillar while spinning, it will be seen that in moving its head it draws on the frothy mass of silk fixed to the web of the cocoon. This traction certainly aids very much the exit of the thread, but it is not the only cause.

The silk, Blanc affirms, is pushed out by a force a tergo, developed by different agents, such as the pressure of the distended cuticle or the silky mass contained in the reservoir, as seen in the section of a worm which has spun its cocoon. But if we consider a caterpillar before it has begun to spin, it is difficult to explain the mechanism of spinning. As Blanc has often observed, in making sections of the heads of silkworms, two cases arise. Sometimes the worm has already spun a little, and a certain length of the frothy silk (bave) issues from the orifice of the spinneret, where it forms a small twisted bundle. At other times the worm has not spun since its last moult or the frothy mass of silk has broken within the head, and we find the end in the common tube. In the first of these two cases, the worm, dilating its press, is able by a general contraction to discharge a little of the gritty material (grès) which lines the ball of silk hanging at the end of the spinneret. It can also reject a certain quantity of the secretion of Filippi’s glands and thus soften the gritty substance. The little plug of silk can then adhere to the body with which it comes in contact.

In the same case it is necessary that the two bits or portions of silk traverse the press, and this normally has a calibre less than their diameter. The worm should then distend the spinning tube as much as is practicable, so as to make the openings as large as possible. It has been stated that the press is, in this condition, at least as large as the mass of frothy silk. This Blanc believes (although Gilson thinks otherwise) is pushed by a force a tergo, and reaches the funnel of the spinning canal; its two bits of silk (brins) unite there, penetrate into the canal itself, and, owing to successive impulses produced by the general contractions of the worm, press through and pass out of the spinneret.

While the silkworm is engaged in spinning its cocoon, the spinneret and press execute very varied movements, determined by the elevator, depressor, retractor, and protractor muscles of the labium, as well as those of the press. These movements, originally very numerous, may combine among themselves, so that the spinneret is susceptible of assuming during the process of spinning still more diverse positions.

Histologically the silk-glands are composed of three layers,—the outer, or tunica propria (Fig 337); the inner, the tunica intima; the middle layer being composed of extraordinarily large epithelial cells which can be seen with the naked eye, and are also remarkable for the branched shape of the nuclei (a, b, c, 337), the branches being more or less lobed, and the larger the cells the more numerous are the branches of the nucleus. Gilson^[53] finds that those of Trichoptera, Lepidoptera, Diptera, and Hymenoptera ordinarily consist of a small number of cells; and it is quite common, he says, to find only two cells in a transverse section (Fig. 338, A). In the Tenthredinidæ, however, “the organ still consists of a tube, the wall of which is composed of flat cells, but in addition to that, two series of spheroidal cells are attached to the sides. Each of these cells contains a system of tiny canals running through their cytoplasm (B, i. d). These cells are the secreting elements; they continually cast the silk substance into the tube.” A peculiarity of the tunica intima is its distinct transverse striation.

The lining of the glands and of their common duct is moulted when the caterpillar casts its skin, and this, as well as the mode of development, shows that the glands are invaginations of the ectoderm. Gilson finds that the silk-glands and silk-apparatus of Trichoptera are very similar to those of caterpillars, and that the silk is formed in the same way.

Appendages of the silk-gland (Filippi’s glands).—In most larvæ there is either a single or a pair of secondary glands which open into the spinning glands near their anterior end. They are outgrowths of the gland provided with peculiarly modified excretory cells or evaginations of the entire glandular epithelium. Those of Bombyx mori (Fig. 340) are very well developed, and, according to Blanc, form two whitish, lobulated masses in the labium on each side of the common duct of the spinning gland. Externally they appear to be acinose; but their structure, as described by Blanc and by Gilson, is very peculiar. Helm thinks, with Cornalia, that the function of these glands is to secrete the adhesive fluid which unites the silk threads, and also to make the silk more adhesive in the process of spinning, but Blanc states that this is done before the thread passes into the common excretory tubes, and he is inclined to think that the secretion serves to lubricate the spinneret, and thus to facilitate the passage of the thread. On the other hand, in certain caterpillars these glands are situated quite far from the spinning apparatus.

The silk-glands in the pupa state undergo a process of degeneration, and finally completely disappear. They are specific larval organs evolved in adaptation to the necessity of the insect’s being protected during its pupal life by a cocoon. (Helm.)

Morphologically the silk-glands are by Lang regarded as modified coxal glands, and homologues of the setiparous parapodial glands of chætopod worms, the coxal glands of Peripatus, and the spinning glands of spiders.

In Scolopendrella, spinning glands are situated in the two last segments of the body, opening out at the end of the cercopods (Fig. 15, s.gl), and the larvæ of the true Neuroptera (Chrysopa, Myrmeleon, etc.) which spin cocoons, have spinning glands opening into the rectum. The silk forming the cocoon of the ant-lion, as Siebold and the older observers have stated, is secreted by the walls of the rectal or anal sac. Siebold (Anatomy of the Invertebrates, p. 445) states that in the larva of Myrmeleon, the silk-apparatus is very remarkable, “for the rectum itself is changed into a large sac and secretes this substance which escapes through an articulated spinneret projecting from the opening of the anus”^[54] (Fig. 307, e). The larvæ of the Mycetophilidæ have spinning glands at the hinder end of the body, as also the imago of the female of the tineid moth Euplocamus. (Kennel.) The larvæ of ichneumons, wasps, bees, of Cecidomyia, and other Diptera, spin silken cocoons, but their glands have not yet been examined.

It should also be observed that during the process of pupation the larvæ of butterflies, of certain flies (Syrphus), and beetles (Coccinellidæ and some Chrysomelidæ) attach themselves by silk spun from the anus, so that the pupa is suspended by its tail; such glands are probably homogenetic with the coxal glands.

The silk in its fluid or soft state is mucilaginous, and according to Mulder, in the silkworm consists of the following substances, varying somewhat in their relative proportions by weight:

Helm, E. Anatomische und histiologische Darstellung der Spinndrüsen der Schmetterlingsraupen. (Zeitschr. f. wissens. Zool., xxvi, 1876, pp. 434–469, 2 Taf.)

Lidth de Jeude, Th. W. van. Zur Anatomie und Physiologie der Spinndrüsen der Seidenraupe. (Zool. Anzeiger, 1878, pp. 100–102.)

Engelmann, W. Zur Anatomie und Physiologie der Spinndrüsen der Seidenraupe. (Onderz. Phys. Lab. Utrecht, iii, 1880, pp. 115–119.)

Joseph, G. Vorläufige Mitteilung über Innervation und Entwickelung der Spinnorgane bei Insekten. (Zool. Anzeiger, 1880, pp. 326–328.)

Poletajew, N. Ueber die Spinndrüsen der Blattwespen. (Zool. Anzeiger, 1885, pp. 22–23.)

Meinert, Fr. Contribution à l’anatomie des fourmilions. (Overs. Danske Vidensk. Selsk. Forh. Kjöbenhavn, 1889, pp. 43–66, 2 Pls.)

Blanc, Louis. Étude sur la sécrétion de la soie et la structure du brin et de la bave dans le Bombyx mori. Lyon, 1889, pp. 48, 4 Pls.

—— La tête du Bombyx mori à l’état larvaire. Anatomie et physiologie. (Extrait du volume des Travaux du Laboratoire d’Études de la Soie. Années 1889–1890, Lyon, 1891, pp. 180, 95 figs.)

Gilson, G. Recherches sur les cellules sécrétantes. La soie et les appareils séricigènes: I. Lépidoptères. (La Cellule, 1890, vi, pp. 115–182, 3 Pls. I, Lépidoptères (suite); II, Trichoptères. Ibid., x, pp. 71–93, 1893, 1 Pl.)

Garman, H. Silk-spinning dipterous larvæ (Science, xx, 1893, p. 215).

Also the writings of Meckel, Pictet, Duméril, Klapálek, Wistinghausen, Loew, Hagen, Fritz Müller, Kolbe, McLachlan, de Selys-Longchamps.

c. The cæcal appendages.

These diverticula of the mid-intestine (“stomach”) are appended to the anterior end, and in the living, transparent larva of Sciara, which has two large, long, slender cœca (Fig. 341), the partly digested food may be seen oscillating back and forth from the anterior end of the stomach into and out of the base of each cæcum. In the Locustidæ (Anabrus, Fig. 299) and Gryllidæ (Fig. 344, e) there are two large, short cæca, and in the locusts (Caloptenus) there are six cæca, while cockroaches have eight. In the Coleoptera (Carabidæ and Dyticidæ) these large cæca appear to be replaced by very numerous slender, minute villi or tubules, which arise from the anterior part of the stomach (Figs. 317, r, also 342).

These cæca differ in structure from the stomach, as shown by Graber, as well as by Plateau and by Minot. The latter states that a single transverse section of one of the diverticula of the locust demonstrates at once that its structure is entirely different from that of the stomach.

Its inner surface is thrown up into longitudinal folds, generally twelve in number. These folds shine through the outer walls, and are accordingly indicated in the drawings of Dufour, Graber, and others. The entire cæcum has an external muscular envelope, outside of which are a few isolated longitudinal muscular bands. The folds within are formed mainly by the high cylindrical epithelium which lines the whole interior of the cavity. Tracheæ ramify throughout all the layers outside the epithelium. There are appearances of glandular follicles in the bottom of the spaces between the folds. (Minot.)

Burmeister supposed that these cæca were analogous to the pancreas, and this view has been confirmed by Hoppe Seyler, Krukenberg, Plateau, and others, who claim that the digestive properties of the fluid secreted in them agrees with the pancreatic fluid of vertebrates.

d. The excretory system (urinary or Malpighian tubes)

The excretory matters or waste products of the blood tissue of worms are carried out of the body by segmentally arranged tubes called nephridia. As a rule they arise in the blood sinuses of the body and open externally through minute openings in the skin. As there is a pair to each segment (in certain oligochete worms two or three pairs to a segment), they are often called segmental organs. In the annulate worms each segment of the body, even the cephalic or oral segment, originally contains a pair of these excretory organs. These vessels may have survived in myriopods and perhaps do exist in insects as urinary tubes, and also occur in many of the Arachnida, and thus are characteristic of each important class of land arthropods, but are either wanting or are very rudimentary or much modified in the marine classes, notably the Crustacea and Merostomata (Limulus), where they are represented by the shell-glands of Copepoda, green glands of the lobster, and the brick-red glands of Limulus.

In the earliest tracheate arthropod, Peripatus, these tubes are well developed and are highly characteristic, each segment behind the head bearing a pair (Fig. 4, so₄-so₉). It has been suggested by some, but not yet proved, that the urinary tubes of insects are morphologically the same as the segmental organs of worms and of Peripatus; but there are no facts directly supporting this view, and, as Sograff states, it is a pure hypothesis and can only be confirmed or disproved by very detailed researches on the development of the urinary tubes of myriopods and of insects. Others regard them as probably homologous with the tracheæ, since they have a similar origin. As, however, they arise in the embryo as outgrowths of the proctodæum they may have arisen in myriopods and insects independently, and not be vermian heirlooms.

While in worms and in Peripatus a pair of these segmental organs occur in each segment, in insects this serial arrangement is not apparent; those with a purely excretory function are not segmentally arranged, with outlets opening externally, but arise as outgrowths of the hind-intestine or proctodæum of the embryo, not being segmentally arranged. The place of their origin is usually the dividing line between the mid and hind intestine (Fig. 343, mp); this applies to Scolopendrella (Fig. 15, urt) as well as to insects.

The urinary tubes are usually long, slender, blind, tubular glands varying in number from two to over a hundred, which generally arise at the constriction between the mid and hind intestine, and which lie loosely in the cavity of the body, often extending towards the head, and then ending near the rectum (Figs. 301, 310, vm). They were first discovered by the Italian anatomist Malpighi, after whom they were called the Malpighian tubes. While at first generally regarded as “biliary” tubes, they are now universally considered to be exclusively excretory organs, corresponding to the kidneys of the higher animals.

Usually arising from the anterior end of the hind-intestine where it passes into the mid-intestine, in certain forms they shift their position, in some Hemiptera (Lygæus, Cimex) opening into the rectum, while in the Psyllidæ they arise from the slender hinder part of the mid-intestine, being widely separated at their origin. (Fig. 321.)

The length varies in different groups; where they are few in number (two to four, six to eight), they are very long, but where very numerous they are often short, forming dense tufts, each tuft connecting with the intestine by a common duct (ureter), or, as in the mole-cricket, the numerous tubes empty into a single duct (Fig. 344); in the locusts (Acrydiidæ), however, they are arranged in 10 groups, each group consisting of about 15 tubes, making about 150 in all; and are much convoluted and wound irregularly around the digestive canal, and when stretched out being about as long as the entire body.

The urinary tubes occur in twos, or in multiples of two, though a remarkable exception is presented in the dipterous genera Culex and Psychodes, in which there are five tubes; the young and fully grown larvæ, as well as the pupa and imago of Culex, having this number (Fig. 433, mg.)

In many insects (Pentatoma, Cimex, Velia, Gerris, Haltica, Donacia, and often in caterpillars), the vessels open into a sort of urinary bladder connecting with the intestine on one side.

In the larvæ of some insects the blind ends of the tubes are often externally bound to the rectum, in the silkworms being attached by fine threads to the intestine, while in some flies (Tipula and Ctenophora), two vessels may unite to form a loop. In all larval Cecidomyiæ, the two tubes are united to form a loop which curves backward, opening near the vent, the proctodæum being very short. (Giard.)

While usually the urinary vessels form simple tubes, in many species of Lepidoptera and Diptera they are branched, thus resembling those of spiders and scorpions. Moreover, in many Lepidoptera and Diptera (Fig. 308), the tubes are not simple, but are lobulated, and in some Hemiptera (Pentatoma, Notonecta, and Tettigonia) are twisted or lace-like. In rare cases there are two kinds of urinary tubes; in Melolontha vulgaris, two of them are partly lobulated and yellow, while the other two are simple and white. Their color in beetles varies, some being whitish or yellowish; in Geotrupes, Dyticidæ, Hydrophilidæ, etc., reddish brown; in Gryllotalpa as well as Locusta viridissima, there are two different kinds of vessels, differing in contents and in color (white or yellow), as well as histologically. (Schindler.)

The exterior of the tubes is richly provided with tracheæ, which often form a web around them, and the fine branches often seem to attach them to the intestine. In Acheta they are enveloped by a very delicate, loose network of muscular fibres. (Schindler.)

The urinary tubes consist, according to Schindler, of at least three cellular layers (Fig. 345):—

1. An external, connective, nucleated membrane, the peritoneal membrane.

2. A very delicate homogeneous basal membrane, the tunica propria.

3. A single layer of large polygonal excretory cells.

4. Lining the internal canal a chitinous layer penetrated by pore-canals, the intima often wanting.

The secretory cells are usually of the same size, but in many cases are relatively small; sometimes four to six or more form the periphery of the canal, sometimes three or only two. In some insects the cells are so very large that a single cell forms the entire periphery. The nuclei in the Lepidoptera (Papilio, Pontia, Cossus) are large and irregularly branched.

The excretions of the Malpighian vessels, derived from the blood and from the fat-body, are more or less fluid and granular, sometimes pulpy. From the cells they pass into the canal, thence into the intestine, and thence out of the body. How, says Kolbe, the secretion passes into the intestine, whether by the contraction of the fine fibrillæ of the peritoneal membrane, or by the external pressure of the other organs, or by the pressure of the secretory matter behind, is not yet known. Grandis observed in living Hydrophilus that the urinary tubes moved, without the muscles seeming to show what caused the motion. Moreover, the cells incessantly changed their form. At a lower temperature such motions ceased. The tracheæ, ending freely in the cells, did not anastomose. (Kolbe.)

The different colors of the tubes (white, yellow, red, brown, or green) is due to the hue of the excretions, and is independent of the color of the blood and of the urinary substances held in the secreted matter.

Schindler found that insects of different stages, collected in winter, differed very much in their urinary secretions, the tubes in the adults being entirely empty, while in the larvæ they were filled full, so that he concluded that in the former the process of excretion during the winter hibernation is very slow, but in the latter very rapid.

As to the activity of the urinary vessels the following experiments will throw some light. Tursini fed a Pimelia with fuchsin; its urinary tubes were consequently colored red. Schindler fed insects with indigo-carmine, which was excreted by the urinary tubes; Kowalevsky arrived at the same results, which seems to prove that these vessels are analogous to the kidneys of vertebrates. Moreover, Schindler injected through the side of the first abdominal segment into the cavity of the body of a Gryllotalpa a concentrated solution of sodium salt of indigotin-disulphonic acid. After one or two hours the external portion of the epithelium of the urinary vessels was stained deep blue, while the inner portion remained of the normal transparency; the nuclei being for the most part deeply stained. Between one and two days after, the staining matter had not yet wholly passed through the central canal, the surface recently stained still appearing light blue.

The solid contents of the urinary tubes consist partly of crystals, which occur singly in the epithelial cells, or form scattered masses when situated in the central canal. Besides tabular rhombic crystals, there occur concretions which contain uric acid, and probably consist of urate of soda, also octahedral crystals of chloride of soda, and quadro-pyramidal crystals of oxalate of lime. Also acicular prisms occur; besides chloride of soda, phosphates, carbonate of lime, oxalate of lime in quantity, leucine, coloring matters, etc.; while the fluid secretion also contains urea (?), uric acid, and abundant urates; uric acid crystals were precipitated by the addition of acetic acid, and by adding hydrochloric acid crystals belonging to the dimetric system were formed. The often numerous spheroidal small granules are biurate of soda and biurate of ammonia. Pale, concentrically banded concretions are leucine pellets.

According to Kölliker the contents of the urinary vessels^[55] in general are: (1) round granules of urate of soda and urate of ammonia; (2) oxalate of lime; and (3) pale transparent concretions of leucine. Crystals of taurin are also said to occur. (Claus’ Zoölogy, p. 531.)

Although uric acid is characteristic of the urinary tubes, yet sometimes it is wanting in them, while uric acid substances in quantity occur in the fat-body or in the mid-intestine.

In the living insect the urinary tubes remove urates from the blood; “the salts are condensed and crystallized in the epithelial cells, by whose dehiscence they pass into the central canals of the tubules and thence into the intestine.” (Miall and Denny.)

The process of excretion is carried on not only by the urinary tubes, but also, as Cuénot has recently shown (1896) in Orthoptera, by the pericardial cells and certain cells of the fat-bodies. In the last-named cells urates are stored throughout life; the pericardial cells apparently secrete but do not store waste products, which are finally eliminated by the urinary tubes, the latter constantly eliminating waste.

Primitive number of tubes.—Wheeler considers the primitive number of urinary tubules to be six, other authors regarding two pairs as the primary or typical number; and while Wheeler agrees that the more ancestral tracheate arthropods had but a single pair, Cholodkowsky supposes the primitive number in insects themselves to be a single pair. This view is strengthened by the fact that Scolopendrella has but a single pair (Fig. 15).

While Peripatus has no urinary tubes, in Myriopods a single pair arises, as in insects, from the hind-intestine.

When in insects the number of these tubes is few, they are, with rare exceptions, arranged in pairs, so that Gegenbaur and others have considered this paired arrangement as the primitive one. When the tubules are very numerous in the adult, as in Orthoptera, the embryos and larvæ have a much smaller number, Wheeler stating that “in no insect embryo have more than three pairs of these vessels been found.” We have observed 10 primary tubes in the embryo of Melanopus (Fig. 347), from each of which afterwards arise 15 secondary tubules. In the Termites, only, do the young forms have more urinary tubes than the adults.

In Campodea there are about 16 urinary tubes and in Machilis either 12 (Grassi) or 20 (Oudemans); but in other Thysanura the number is much less, Lepisma having either four, six, or eight, according to different authors, and both Nicoletia and Lepismina having six, opening separately into the hind-intestine. On the other hand, these organs have not yet been detected in Japyx. Whether they exist at all in the Collembola, which are degenerate forms, is doubtful. The weight of opinion denies their existence, though they may yet be found existing in a vestigial condition. They are said by Tullberg and by Sommer to exist in Podura, but are of peculiar shape.

Coming now to the winged insects, in what on the whole is perhaps the lowest or most generalized order, the Dermaptera, the number is over 30, and their insertions regularly encircle the intestine. (Schindler.) In the most ancient and generalized family of Orthoptera, the Blattidæ, Schindler detected from 60 to 70 tubes, but in a nymph of Periplaneta not quite 10 mm. in length he found from 16 to 18, and in nymphs 4 to 5 mm. long there were only eight vessels; while Wheeler has found in the embryo of Phyllodromia germanica but four tubes. In the adult Acrydiidæ there are as many as 150, in the Locustidæ between 40 and 50, and in the Gryllidæ about 100.

The Ephemeridæ with about 40, the Odonata with 50 to 60 tubules, the Perlidæ with from 50 to 60, are polynephrious; while the Termitidæ and Psocidæ are oligonephrious, the former having from six to eight and the Psocidæ only four tubes. So also all the other orders not mentioned, except the Hymenoptera, have few of these tubes. The Hemiptera, with none in Aphidæ, a single pair in the Coccidæ, and two in all the rest of the order, have the fewest number.

In the Neuroptera there are from six to eight, while in a larva, possibly that of Chauliodes, Wheeler finds the exceptional number of seven.

The closely allied order Mecoptera (Panorpidæ), and also the Trichoptera, agree with the Neuroptera (Sialis) in having six. According to Cholodkowsky all Lepidoptera have six of these vessels, except Galleria, which has but four. He finds that in Tinea biselliella (also T. pellionella and Blabophanes rusticella) the larva has six vessels, which, however, undergo histolysis during pupation, a single pair arising in their stead. On this account he regards the primitive number of urinary tubes as two, or a single pair, this return from six vessels in the larva to two in the imago being considered a case of atavism.

In the Coleoptera, the number of urinary tubes is from four to six; in what few embryo beetles have been examined (Doryphora, Melolontha), there are six vessels, but in the embryo of Dyticus fasciventris, Wheeler has detected only four, this number being retained in the adult. He thinks that in beetles in general, a pair of vessels must be “suppressed during post-embryonic development, presumably in early larval life.”

In Diptera and Siphonaptera, the number four is very constant, there being, however, a fifth one in Culex and Psychoda (Fig. 400.)

The number of these vessels is very inconstant in the Hymenoptera, varying from six (Tomognathus, an ant, worker) to 12 (Myrmica), and in Apis reaching the number of 150.

In the embryo of the honey-bee and wall-bee (Chalicodoma), there are only four; we still lack any knowledge of the number in embryo saw-flies.

The following is a tabular view of insects with few urinary tubes (Oligonephria) and many (Polynephria). It will be seen that the number has little relation to the classification or phylogeny, insects so distantly related as the Orthoptera and Hymenoptera being polynephrious:—

Here should be mentioned the singular fact discovered by Koulaguine that in the larva of Microgaster, the urinary tubes have no connection with the intestine, but open dorsally on the outside of the body on each side of the anus. Ratzeburg had stated that the last segment of the body was in the form of a vesicle. Koulaguine now shows that this vesicle is in reality the end of the intestine opening upwards; as the result of this dorsal opening of the intestine the Malpighian vessels open on the sides of the oval vent, and have no connection with the intestinal canal. Whether this is of morphological import, or is only a secondary adaptation, Koulaguine does not state, his paper being a preliminary abstract.

Wheeler thus sums up our present knowledge regarding the number and homologies of the Malpighian or urinary tubes:

1. It is very probable that the so-called Malpighian vessels of Crustacea and Arachnida are not the homologues of the vasa Malpighi of the Eutracheata (insects and myriopods).

2. The Malpighian vessels of the Eutracheata arise as paired diverticula of the hind-gut and are, therefore, ectodermal.

3. In no insect embryo are more than six vessels known to occur; although frequently only four are developed.

4. The number six occurs either during embryonic or post-embryonic life in members of the following groups: Apterygota, Orthoptera, Corrodentia; Neuroptera, Panorpata, Trichoptera, Coleoptera, Lepidoptera, and Hymenoptera.

5. The number four seems to be typical for the Corrodentia, Thysanoptera, Aphaniptera, Rhynchota, Diptera, and Hymenoptera.

6. The embryonic number in Dermaptera, Ephemeridea, Plecoptera, and Odonata has not been ascertained, but will probably be found to be either four or six.

7. There is evidence that in at least one case (Melolontha), the tetranephric is ontogenetically derived from the hexanephric condition by the suppression of one pair of tubules.

8. It is probable that the insects which never develop more than four Malpighian vessels have lost a pair during their phylogeny.

9. The post-embryonic increase in the number of Malpighian vessels in some orders (Orthoptera, Odonata, Hymenoptera) is secondary and has apparently arisen to supply a demand for greater excreting surface.^[56]

LITERATURE ON THE EXCRETORY (URINARY) ORGANS

Malpighi, M. Dissertatio epistolica de Bombyce, Societati regiæ Londini ad scientiam naturalem promovendam institutæ dicata. (Londini, 1669, 12 Pls.)

Herold, M. J. D. Entwicklungeschichte der Schmetterlinge. 1815.

Rengger, J. R. Physiologische Untersuchungen über den tierischen Haushalt der Insekten. Tübingen, 1817, pp. 82.

Wurzer. Chemische Untersuchungen des Stoffes in den Gallgefässen von Bombyx mori. (Meckel’s Archiv f. Physiol., iv, 1818, pp. 213–215.)

Gaede, H. M. Physiologische Bemerkungen über die sogenannten Gallgefässe der Insekten. (Nova Acta Acad. Caes. Leopold.-Carolin., 1821, x, Pars II, pp. 186–196.)

Meckel, J. F. Ueber die Gallen- und Harnorgane der Insekten. (Meckel’s Archiv, i, 1826, pp. 21–36.)

Audouin, J. V. Calculs trouvés dans les canaux biliaires d’un cerf volant. (Ann. sc. nat., 2 Sér., 1836, v, pp. 129–137.)

Frey und Leuckart. Anatomie und Physiologie der Wirbellosen. 1843.

Dufour, L. Mémoire sur les vaisseaux biliaires ou le foie des Insectes. (Ann. sc. nat., 1848, Sér. 2, xix, pp. 145–182, 4 Pls.)

Karsten, H. Harnorgane von Brachinus complanatus. (Müller’s Archiv f. Anat. und Physiol., 1848, pp. 367–374.)

Fabre, J. L. Étude sur l’instinct et les metamorphoses des Sphégiens. (Ann. d. sc. nat., 4 Sér., 1856, vi, pp. 137–189.)

—— Étude sur le rôle du tissu adipeux dans la sécrétion urinaire chez les Insectes. (Ibid., 4 Sér., xix, pp. 351–382.)

Schlossberger, J. E. Untersuchungen über das chemische Verhalten der Krystalle in den Malpighischen Gefässen der Raupen. (Archiv f. Anat. und Physiol., 1857, pp. 61–62.)

Leydig, F. Lehrbuch der Histiologie. 1857.

Sirodot, S. Recherches sur les sécrétions chez les Insectes. (Ann. sc. nat., 4 Sér., Zool., 1858, x, pp. 141–189, 251–334, 12 Pls.)

Kölliker, A. Zur feineren Anatomie der Insekten (Ueber die Harnorgane, u.s.w.) (Verhandl. d. Physikal.-medizin. Gesellsch. in Würzburg, viii, 1858, pp. 225–235.)

Schindler, E. Beitrage zur Kenntnis der Malpighischen Gefässe der Insekten. 3 Taf. (Zeitschr. f. wiss. Zool., xxx, 1878, pp. 587–660.)

Chatin, G. Note sur la structure du noyau dans les cellules marginales des tubes de Malpighi chez les Insectes et les Myriapodes. (Ann. d. sc. nat., 6 Sér., xiv., 1882, pp. 7, 1 Pl.)

Witlaczil, E. Zur Anatomie der Aphiden. (Arbeiten a. d. Zool. Instit. d. Univers. Wien., iv, 1882, pp. 397–441, 3 Taf.)

Cholodkowsky, N. Sur les vaisseaux de Malpighi chez les Lépidoptères. (Compt. rend. Acad. d. Sc., Paris, xcix, 1884, pp. 631–633.)

—— Sur la morphologie de l’appareil urinaire des Lépidoptères. (Archives de Biologie, 1887, vi, pp. 497–514, 1 Pl.)

Loman, J. C. C. Ueber die morphologische Bedeutung der sogenannten Malpighischen Gefässe der echten Spinnen. (Tijdschr. Nederl. Dierk. Ver. (2) Deel 1, 1887, pp. 109–113, 4 Fig.)

Marchal, P. Contribution à l’étude de la désassimilation de l’azote. L’acide urique et la fonction rénale chez les Invertébrés. (Mém. Soc. Zool. de France, 1889, iii, pp. 42–57.)

Kowalevsky, A. O. Ein Beitrag zur Kenntnis der Exkretionsorgane. (Biol. Centralbl., ix, 1889–90, pp. 33–47, 65–76, 127–128.)

—— Sur les organes excréteurs chez les arthropodes terrestres. (Congrès international de Zool., 2^me Session à Moscou, 1892, Pt. I, pp. 186–235, 4 Pls.)

Griffiths, A. B. On the Malpighian tubules of Libellula depressa. (Proc. Roy. Soc., Edinburgh, 1889, xv, pp. 401–403, Figs.)

Grandis, V. Sulle modificazioni degli epitelii ghiandolari durante la secrezione. (Atti Accad. Torino, 1890, xxv, pp. 765–789, 1 Pl.; Archiv Ital. Biol., 1890, xiv, pp. 160–182, 1 Pl.)

Koulaguine, N. Notice pour servire à l’histoire du développement des hyménoptères parasites. (Congrès internat. de Zool., 2^me Session à Moscou, 1892, Pt. I, pp. 253–277.)

Sograff, Nicolas. Note sur l’origine et les parentés des Arthropodes, principalement des Arthropodes trachéates. (Congrès internat. de Zool., 2^me Session à Moscou, 1892, Pt. I, pp. 278–302.)

Giard, Alfred. (Note on the urinary tubes of larval Cecidomyia. Annals Ent. Soc., France, lxii, 1893, pp. lxxx-lxxxiv, 1 Fig.)

Wheeler, William M. The primitive number of Malpighian vessels in insects. (Psyche, vi, May-December, 1893, Parts 1–6, pp. 457–460, 485–486, 497–498, 509–510, 539–541, 545–547, 561–564.)

Metalnikoff, C. K. Organes excréteurs des insectes. (Bull. Acad. imp. Sci. St. Pétersbourg, 1896, iv, pp. 57–72, in Russian, 1 Pl.)

See also the works of Straus-Dürckheim, Will (Müller’s Archiv. 1848, p. 502), Brugnatelli, Leidy, Dufour, Ramdohr, Basch, Davy, Grassi, Minot, Berlese, Adlerz, Marchal (Bull. Ent. Soc. France, 1896, p. 257); Bordas (Appareil glandulaire des Hyménoptères, 1894), also C. R. Acad. Sc. Paris, 1897.

e. Poison-glands

Poison-glands are mainly confined to the stinging Hymenoptera, i.e. certain ants, and the wasps and bees, but also occur in the mosquito, while many, if not most bugs, seem to instil a drop of poison into the punctured wounds they make.

In the honey and other bees the poison apparatus consists of two poison-glands whose secretion passes by a single more or less convoluted efferential duct into the large poison-sac, and thence by the excretory duct, which is enlarged at the base of the sting (Figs. 194, 195), out through the sting by the same passage as the eggs. According to Carlet, the poison apparatus of bees consists of two kinds of glandular organs, of which one kind secretes a feebly alkaline fluid, the other an acid product. The poison is only effective when both fluids are mixed. The resultant venom is always acid. The action of this venom upon some animals, as rabbits, frogs, and certain beetles, is slight; but the domestic fly and the flesh-fly are immediately killed by it. The inoculation of a fly with the secretion of one of the glands does not produce death until after a considerable time, but death follows very quickly if the same fly is subjected to a second inoculation, this time with the secretion of the other gland. The alkaline glands are in bees and all poisonous Hymenoptera strongly developed, but become vestigial in those forms which sting their prey to serve as food for their larvæ. The poison which the solitary sand and wood wasps and Pompilidæ inject into their victims only paralyzes them.