The following paper contains the results of researches carried on at the Plymouth Laboratory of the Marine Biological Association, where, by appointment of the British Association Committee, I occupied a table for six weeks during the summer of last year (1892), my stay being prolonged for an additional month through the kindness of Mr. Robert Bayly, one of the Governors of the Association, who furnished me with an additional nomination.

The observations were made chiefly upon larvæ of Palæmonetes varians, but they were subsequently extended to the adult of this form, as well as to larvæ and adults of allied species, the latter part of the work having been done in Professor Weldon’s laboratory at University College, London.

Palæmonetes varians is found in the brackish waters of certain ditches and pools near the estuary of the Plym, which are subject to the influence of the tides.

The larvæ were at first obtained by keeping gravid females, whose eggs were in an advanced stage of development, in separate vessels in a mixture of one part of sea water to two of fresh (this mixture having about the same density as the water in which the animals were captured). They were afterwards found to hatch equally well in a stream of running sea water, and this method was then chiefly used.

Towards the end of the summer, larvae in all stages of development were obtained in large numbers by fishing with a fine net in certain sheltered pools, and further breeding in captivity became unnecessary.

The larvae were preserved in Flemming’s fluid (both the weak and strong mixtures gave good results), washed in water, and hardened in alcohol. The sections were stained with Delafield’s hæmatoxylin.

During the greater part of the larval life two pairs of nephridia are present; the green glands, which open at the bases of the second antennae, and the shell-glands, which open at the bases of the second maxillæ.

—A horizontal longitudinal section through the green gland of a Palæmonetes larva, which is three or four days old, is represented in fig. 1. At this stage the gland has a form which is similar to that described by Weldon (No. 20) and Marchal (No. 17) as persisting in the adult of Virbius, Pandalus, and Crangon, excepting that there is as yet no trace of the remarkable enlargement of the bladder which the former author designates as “nephro-peritoneal sac.”

The end-sac (fig. 1, e. s.) communicates by means of a U-shaped tube (tu.) with the bladder (bl.), which is in reality merely the enlarged distal portion of the tube. (The dotted lines in the figure [fig. 1] indicate the way in which tube and bladder join each other, as determined by the following sections of the series.) The bladder is placed in communication with the exterior by means of a very short ureter, which opens in the region indicated by the letter O in fig. 1.

It is not necessary for me to give an account of the histological structure of the various portions of the gland, as it has been so fully and accurately described for this and other species by Grobben (No. 10), Marchal (No. 17), and Weldon (No. 20). The only point upon which there appears to be a division of opinion is as to the nature of the most internal layer of the tube (cut.), which Grobben considers to be a cuticle pierced by fine pores, whilst Marchal regards it as made up of rows of minute vesicles, which subsequently enlarge and break away from the cells. Without expressing an opinion as to the latter point, I can say that Marchal’s figures (No. 17), especially his fig. 8, pl. v, represent much more nearly the appearance of this layer, as I have observed it, than the figures given by Grobben (No. 10).

The condition of the green gland represented in fig. 1 occurs, as I have stated, in larvae which are a few days old. At the time of hatching, however, a more primitive condition is found, for when this event takes place the gland is entirely without a lumen, although the ureter and external opening are present. A transverse section through it at this stage is represented in fig. 2. The gland has the same general shape as in the later stage, and the end-sac (e. s.) and bladder (bl.) are easily distinguished. They consist, however, of solid masses of cells, in which as yet no cavity has appeared. A few hours after the larva is hatched the cells begin to separate, and so give rise to the lumen of the gland. A transverse section at this time is shown in fig. 3.

The striation of the protoplasm of the bladder can be indistinctly seen in these early stages, but from the appearance of the gland I should not judge that it was yet functional.

The later development of the green gland consists chiefly in the enlargement of the bladder. The latter commences to grow in the region indicated by the letter E in fig. 1. At the end of a week after hatching its enlargement has already passed out of the basal joint of the second antenna, and by the end of a fortnight it has grown inwards beneath the circum-œsophageal nerve-commissure as far as the mid-ventral line. A still later stage is represented in fig. 4. Here the bladder has commenced to grow in a dorsal direction, anterior to the œsophagus, and between the two nerve-commissures. In the stage figured, it has reached a point which is at a distance from the ventral surface about a quarter as great as the height of the body in this region.

In a young Palæmonetes, which has apparently just assumed adult characters, the enlarged portions of the bladders of the two green glands are seen in sections (fig. 5, bl.) to lie side by side above the stomach in the position occupied by the single unpaired sac of the adult. The latter sac is formed by the fusion of the enlargements of the bladders of the two sides, and in the series from which fig. 5 was drawn this fusion had already commenced in the most posterior portions of the two enlargements.

The mode of development of this nephro-peritoneal sac (“vessiesus-stomacale impaire”—Marchal) proves it to be in reality derived from the fusion of the greatly enlarged bladders of opposite sides, a conclusion which has been already arrived at by Weldon (No. 20) and Marchal (No. 17) from a comparative study of the renal organs of the Decapods.

With regard to the plexus of tubules found in the adult between the end-sac and the bladder, I can only say that it is derived from the single short renal tube which in the larva connects the end-sac with the bladder; that is to say, the portion between the opening of the end-sac and the point indicated by the letter E in fig. 1. This tube enlarges considerably in size, but remains single until a late stage of larval life; whilst in the youngest adult of which I possess sections there are already several smaller tubules, which have obviously been derived by the splitting up of the single tube, but exactly how the process has taken place I do not know. What has in all probability happened is simply that opposite walls of the large tube have grown together along certain lines, and thus given rise to the complicated plexus of smaller ones.

As long ago as 1876, Claus (No. 2) predicted that a rudiment of the shell-gland would in all probability be found in Malacostraca larvæ, and he figures what he takes to be this gland in a stomatopod larva (No. 2, pl. iv, fig. 8). He has subsequently shown that the gland exists in Anisopods and Isopods (No. 7). In a figure of a Callianassa larva in the mysis stage, Claus also inserts and names the shell-gland, opening on the posterior surface of the second maxilla (No. 5, pl. v, fig. 42). I believe that this is the only recorded instance of the gland having been recognised in a Decapod at any stage. It occurs, however, in the larvæ both of Palæmonetes and of Palæmon, and at the time of hatching is in a much more highly developed state than the green gland. It is, in fact, at this time the functional renal organ of the animal, the green gland being still without a lumen, and only showing very faintly the characteristic striation of the protoplasm.

The shell-gland of Palæmonetes has the typical structure of a crustacean nephridium (fig. 6, sh. gl.; figs. 7 and 8). It consists of a comparatively short renal tube (tu.), with a considerable lumen, which communicates internally with the end-sac (e. $.), and opens externally at the base of the second maxilla (fig. 8, B, ex. o.). Fig. 6, sh. gl., and fig. 7 represent transverse sections through the gland; whilst fig. 8, A, B, and c, shows three of a series of sagittal sections, of which A is more internal and c more external than B. The general form of the tube may be expressed by saying that it is Y-shaped, the two arms of the Y being in a horizontal plane, with the end-sac attached to the internal one, whilst the leg of the Y is curved in a vertical plane, the concavity looking downwards and backwards. Fig. 6, sh. gl., and fig. 7 both pass through the two arms of the Y, and through the point of attachment of the end-sac.

The end-sac (e. s.), which is suspended from the body-wall by bands of connective tissue, has a similar structure to the end-sacs of the green glands of Estheria and Mysis as figured by Grobben (No. 10), and to those of the green gland and shell-gland of Branchipus as figured by Claus (No. 6). It will be observed, however, that in all my figures the cavity of the end-sac is not in free communication with that of the tube, the entrance from the one to the other being guarded by certain elongated cells of the end-sac which project into the lumen of the tube. Two such cells appear in each section. This arrangement of cells is invariably found at the point where the end-sac joins the tube, and appears to constitute a valve which would allow fluid to pass out of the end-sac, but would prevent it from returning in the opposite direction. The structure of the walls of the renal tube of the shell-gland is the same as that of the corresponding parts of the green gland.

In his account of the development of Eriphya spini-frons, Lebedinski (No. 16) describes a pair of “segmental organs “which develop as evaginations of the somatopleure, and which, according to him, communicate directly with the body-cavity, and open at the bases of the first pair of maxilli-pedes. Korschelt and Heider (No. 14) have already suggested that this organ may be the shell-gland. If this suggestion be correct, Lebedinski must either have mistaken the appendage on which the gland opens and overlooked the end-sac, or the arrangement in the species examined by him must vary from that which exists in Palæmonetes and Palæmon, and which upon theoretical grounds appears the more probable.

It is a fact of interest, and one which is of some importance in a consideration of the phylogeny of the various groups of Crustacea, that in the Entomostraca (Branchipus, Claus, No. 6; Cetochilus, Grobben, No. 11) the green gland is the first of the two pairs of nephridia to become functional, the full development and activity of the shell-gland being deferred to the post-larval period when the green gland has almost disappeared, whilst among the Decapods the periods of development of the two glands are reversed.

The shell-gland is, I believe, functional in Palæmonetes even before hatching, as I have observed it in sections of embryos, some time before this event, with a well-developed lumen, with the characteristic striation of the protoplasm, with a definite external opening, and with a mass of yellowish concretion in the distal portion of the tube. On the other hand, in young specimens which have attained adult characters no trace of this gland has been detected. In them the green gland, which did not appear to be functional at the time of hatching, has become remarkably developed, its activity having increased as that of the shell-gland diminished.

Claus has shown that shell-gland and green gland are developed in the same order in Nebalia as that which occurs in Decapods, and the union of that animal with the Thoracostraca rather than with the Entomostraca is favoured (cf. Claus, No. 8, p. 100).

The different behaviour of the two pairs of nephridia in Entomostraca and Malacostraca renders it unlikely that the latter have descended directly from the former. We should rather suppose that the two groups possessed a common ancestor, in which both pairs of nephridia were equally developed.

THE ANTERIOR REGION OF THE THORAX.—Fig. 6 represents a transverse section through the region of the second maxillæ of a Palæmonetes larva which is ten days old. The body of the animal is surrounded by a layer of thin chitin, beneath which the ectoderm lies. The latter consists of a layer of flattened cells with oval nuclei. A considerable thickening of the ectoderm is generally found in the middle dorsal line (figs. 9, 16), whilst along the ventral surface the layer presents a peculiar granular appearance (figs. 6, 18, 19, ect.) which I have been unable to account for. The cavity enclosed by the chitin and ectoderm may be divided into four regions: a dorsal sac (c.) surrounded by a definite layer of epithelium, and in which the cephalic aorta (Claus; “artère ophthalmique”—Milne Edwards) lies (Ao. c.), but which does not itself contain blood; a central cavity (b. c.), in which the liver, intestine, and nerve-cord are found; two lateral cavities (b. lat.), separated from the central cavity by masses of muscle and bands of connective tissue, and which in the region of the second maxillæ contain the proximal ends of the shell-glands; and fourthly, the cavities of the limbs, which contain the distal ends of the same organs. The cavities of the limbs communicate with the lateral cavities, and the latter frequently communicate with the central cavity by the disappearance of the connective-tissue bands. The three latter cavities all contain blood. The intestine, the liver lobes, the nerve-cord, and the muscles are each surrounded by a definite layer of mesoderm. That surrounding the liver is strongly muscular, the lobes in the living larva being often seen to vigorously contract. Around the shellgland I have been unable to detect any similar mesodermal investment.

In the parts of the thorax anterior to the region of Maxilla II a similar condition of things occurs (fig. 9), except that the central cavity is somewhat broken up by the masses of muscle which move the mandibles. No trace of nephridia has been seen in the segments between Antenna II and Maxilla II, but a section through the region of the first maxillae (fig. 9) shows in the lateral cavities a single pair of large glands (rt. gl.), which have the same structure as the spherical glands which I have recently described in the axis of the gill of the adult Palæmonetes (No. 1). These glands open by short ducts at the bases of the first maxillae, and they, together with the smaller salivary glands (sal. gl.), which are present in considerable numbers in the upper lip, the paragnaths, and the maxillæ, have a great affinity for hæmatoxylin, and assume a much brighter tint when that stain is used than the surrounding tissues.

I have found the dorsal sac in Palæmon serratus, Palæmonetes varians,and Crangon vulgaris, and in the adult it attains a considerable size. If a dissection be made of an adult Palæmon the sac is readily seen (fig. 10). Anteriorly it appears as a long cylindrical tube (c.) lying upon the dorsal enlargement of the bladder (bl.), and containing within it the cephalic aorta (Ao. c.). Posteriorly it is very much enlarged, covering the front part of the ovaries, and running downwards on either side into the cavity which surrounds the intestine and liver. The sac extends backwards to near the front end of the pericardial cavity, but it is completely closed, and has no communication with the latter.

By means of a series of sections through a fully grown adult specimen of Palæmonetes varians, I have ascertained that exactly similar relations exist in that form, and the sections entirely confirm the results of dissection.

The position of the sac in larvæ is also similar to that which it occupies in adults. It is seen in transverse section in figs. 6 and 9, c., and in sagittal section in fig. 11, c. It extends from the base of the rostrum to a point in front of the anterior end of the pericardial septum. The sac appears to be completely closed, and in preserved specimens contains a clot, which can generally be distinguished from the surrounding blood-clot, and which contains no corpuscles.

I have only met with one reference to the existence of this sac. Weldon (No. 20) describes the nephro-peritoneal sac of Palæmon as lying “ventral to the ophthalmic artery and to the median dorsal blood-sinus.” The dorsal sac above described is what is here referred to as the “median dorsal blood-sinus.” It does not, however, contain blood. I have been led to this conclusion for the following reasons:—(1) In a large number of series of sections, both of larvæ and adults, I have never seen a blood-corpuscle within the sac. (2) The sac is closed, and has no communication with the blood-sinuses of the body. This point can be definitely established for larvæ. With regard to adults I cannot make such a positive statement, on account of the difficulty of obtaining a perfect series of undamaged sections, but I have seen nothing which would suggest a contrary conclusion. (3) I have observed carefully and for a long time living larvæ under the microscope. If the larvæ are lying upon their sides blood-corpuscles are readily seen passing along the cephalic aorta, and corpuscles returning from the head are also observed; but the latter keep always quite close to the anterior lobes of the liver. The space occupied by the dorsal sac—that is to say, the space between the cephalic aorta and the row of corpuscles close to the liver—is absolutely free from corpuscles.

At its anterior end the dorsal sac is surrounded by a curious mass of tissue, which is seen in transverse section in fig. 12 (compare also figs. 10, 11, and 15,1, g.). Before commencing this research, Professor Weldon, to whom the existence of this mass of cells was known, and who has indicated it in his figures (No. 20, pl. i, figs. 1—3), suggested to me that blood-corpuscles were being budded off by the tissue, and my preparations certainly tend to support this view. In fig. 121 have endeavoured to represent as nearly as possible the appearance presented by a section, and it is difficult to account for it in any other way than that which Professor Weldon suggests, since the outlying cells cannot be distinguished from blood-corpuscles found in other parts of the body.

A pair of muscles must also be mentioned, which run on either side of the cephalic aorta in the anterior portion of the dorsal sac. These muscles (figs. 11, 12, 15, c. m.) run, in the adult, from the median dorsal surface of the carapace to points just below the angle made by the inferior surface of the rostrum with the front end of the head. What their function is in Palæmonetes is not apparent, but they would seem to be homologous with the muscles which move the rostrum in other forms.

According to Reichenbach for Astacus (No. 18), and Kingsley for Crangon (No. 13), the cephalic aorta of Decapods is formed from mesoderm-cells which have broken away from the ventral bands and wandered to the dorsal surface, where they unite to form a tube. Although I have not specially studied these stages, I may say that one series of sections which I have made tends to confirm the view that these cells travel up on either side between the liver lobes and the ectoderm; and the mode of development of the heart of Branchipus, as described by Claus (No. 6), in which animal solid masses of mesoderm grow up on both sides, appears to be a more primitive form of the same process.

In embryos of Palæmonetes, in which the cephalic aorta is already formed, the latter is not surrounded by a single layer of cells, but the nuclei are in reality arranged in two layers, an internal and an external. This may be seen from fig. 13, A and B, which are drawn from a late embryo of Palaemon etes; but the arrangement in two layers becomes even more obvious when a series of sections is examined. Fig. 13, B, represents the condition of things along the greater part of the length of the aorta, but a few anterior sections appear as in fig. 13, A, where a tongue of mesoderm extends downwards between the liver lobes. This condition of things will be explained at a later stage. In both figures (fig. 13, A and B) the nuclei n′ belong to the internal layer, the nuclei n ″ to the external.

Before the time of hatching arrives, the cells which surround the aorta become considerably enlarged, and give rise to the appearance of solid masses of mesoderm upon either side. Fig. 14, E, may be taken to represent the condition of things throughout the greater part of the length of the aorta just before the larva leaves the egg. In this figure the blood is coloured black, the central mass (Ao. c.) being in the cavity of the aorta, the remainder being in the cavity of the body which lies between the liver lobes (li.) and the ectoderm (ect.); mes. represents the mass of mesoderm on either side of the aorta.

Soon after hatching a change takes place, which leads to the formation of the dorsal sac. Fig. 14 represents sections through different regions of the thorax of a larva very soon after it has become free. The second section drawn (fig. 14, B) passes through the region of the paragnaths. The cells of the external layer (cf. fig. 13, B) are here seen to have drawn away from those of the internal layer which form the walls of the aorta, leaving a clear space (c), which extends almost completely around the latter. A few sections further back this cavity begins to close, until the condition represented in fig. 14, c, is reached. A comparison of this section with those which precede and follow it renders it highly probable that there is a complete closing of the cavity at this point, which is situated in the anterior portion of the segment of the first maxillae. Behind this point sections similar to that seen in fig. 14, D, are met with, and this stage is specially important and interesting. Two distinct cavities have here formed in the masses of mesoderm-cells, and these cavities are separated from each other by the cephalic aorta. After persisting throughout the region of the first maxillae these cavities close, and the sections appear as in fig. 14, E; that is to say, solid masses of mesoderm still lie on either side of the aorta. This condition of things continues through the region of the second maxillae until the point is reached at which, in the following stages, the dorsal sac is found to end. This point is easily determined, because the intestine is there attached by bands of connective tissue to the dorsal surface of the carapace, the bands passing close to the cephalic aorta.

The most anterior region of the aorta and sac remains to be considered. A transverse section through the parts, which is nearly at the level of the mouth, is shown in fig. 14, A. The appearance here presented is explained by fig. 15, which represents a horizontal section through the front part of the head. The cephalic aorta, which has run horizontally from the heart, suddenly bends downwards, and runs almost perpendicularly towards the brain. It is surrounded during this part of its course by the cells already described (fig. 12) as giving rise to blood-corpuscles (Z. g.), and has upon either side of it the two muscles (c. m.) which are also found in the adult. The dorsal sac is already open behind the aorta (fig. 15, c), and the transverse section (fig. 14, A, C) passes along its length.

The following numbers, obtained from the series of which figs. 14, A—E, represent selected sections, will give an idea of the extent of the different regions. The cavity of the sac is seen below the aorta as in fig. 14, A, in three sections at the level of the mouth, this condition passing into that of fig. 14, B, where the cavity is below and on both sides of the aorta, which continues for six sections. The next section is represented in fig. 14, c. The condition with two lateral cavities (fig. 14, D) then appears, and persists in eleven sections in the region of the first maxillæ. The appearance seen in fig. 14, B, in which no cavity is yet formed, runs through the next nine sections in the region of the second maxillaæ to the point where the bands of connective tissue suspend the intestine.

In a larva only a few hours older than the one just described, what appears at first sight to be a very considerable development has taken place. From the region of the mouth to that of the second maxillae the parts under consideration have the appearance represented in fig. 16, A, a few sections only at the posterior end presenting that shown in fig. 16, B. There is now one cavity, which surrounds three sides of the aorta and extends along its whole length. But no great rapidity of growth has been necessary to produce this marked result. By comparing the sections of fig. 14 with those of fig. 16 it will be seen that what has taken place is that the protoplasm of the cells, instead of being gathered into masses around the nuclei, has spread out into a thin sheet, thereby enlarging considerably the cavity, and drawing its walls away from the wall of the aorta.

The sections at the posterior end of the sac in this and the following stages are of special interest, because in them the cavity is divided by a vertical septum (fig. 16, B, S.) into two lateral portions. This fact tends to confirm what is already apparent from fig. 14, D, of the previous stage, that the cavity of the dorsal sac is formed by the union of two lateral cavities, which lie on either side of the cephalic aorta.

The further development of the dorsal sac consists mainly of an increase in its size. Later stages are seen in fig. 9 and fig. 6, c. Its walls are very thin, and the nuclei spindleshaped. At its posterior end the sac grows backwards as a pair of lobes, which are not attached to the wall of the aorta, and which extend as far as the front end of the pericardium. A section through these lobes is seen in fig. 17. This posterior portion of the dorsal sac becomes very much enlarged in the adult, as has already been shown, and lies upon the front end of the ovary (fig. 10).

The condition of the body-cavity in this region may be gathered from figs. 18 and 19. The central and lateral cavities {bo., b. lat.) are similar to those in the anterior region, whilst dorsal to these the pericardial chamber lies (per.). This chamber is separated from the central body-cavity, as is already well known, by the pericardial septum (p. sep.), and it contains the heart. In the larva the pericardial cavity appears to be bounded dorsally by the ectoderm, but I consider it probable that a layer of mesoderm lies within the ectoderm, although I have been unable to detect it. Such a layer is undoubtedly present in the adult, although it is even then very thin, and has extremely small nuclei.

At the front end of the pericardial cavity, immediately beneath the pericardial septum, the genital organs are situated (fig. 18, gen.). In the just hatched larva and in late embryos these consist of two masses of cells with large nuclei, each mass being enclosed in a sheath of mesoderm. At this stage the presence of the latter, on account of the extreme tenuity of the layer, can only be ascertained by the presence of occasional spindle-shaped nuclei which lie outside the larger cells, but in later stages this mesodermal investment becomes much more marked. Bobretzky (see No. 14, p. 379) has already stated that the genital organs originate in the region in which they are found in the larva.

I have not been able to find any trace of the genital ducts in young larvæ, and can give no information as to how these organs originate.

THE ABDOMEN

With regard to the abdomen I have no new facts to add, but my sections confirm the accounts given by Milne Edwards (No. 9) and Claus (No. 4). There are two principal sinuses which run along its whole length—a dorsal sinus in which the intestine lies, and a ventral one which contains the nerve-cord. These two sinuses are generally separated by masses of muscle, but they communicate at intervals by means of lateral sinuses.

THEORETICAL CONSIDERATIONS

We may seek for an explanation of the relations above described by comparing them with the condition of the body-cavity, as described by Sedgwick (No. 19) and von Kennel (No. 12), during the development of Peripatus. Briefly stated, the course of events in the latter animal is as follows:—The hollow mesoblastic somites which form in the ventral bands of mesoderm divide into two portions, which are distinguished as dorsal and lateral. The lateral portions of the somites are transformed into nephridia, whilst the cavity of the lateral sinus of the body originates as a hollowing out in their thickened external walls. The dorsal portions of the somites enlarge upwards, but those of opposite sides do not meet in the dorsal middle line. A space is left between them, which becomes the cavity of the heart. At this stage the dorsal portions of the somites may be said to have attained their maximum state of development. By the diminution of their cavities they then cause the formation of spaces external to themselves, which become lined by wandering mesoderm-cells, and form the pericardial chamber and central body-cavity of the adult. These two cavities are separated by the pericardial septum, which is apparently formed in a more direct way from the walls of the diminishing somites. In certain regions of the body the remnants of the dorsal portions of the somites persist, according to Sedgwick, as the external membrane of the genital glands; whilst, according to v. Kennel, they give rise both to the internal and external cells of these organs. The condition in the adult Peripatus is described by Sedgwick (No. 19, pt. 4, p. 385) as follows:—” The hæmocoele is divided into five main chambers: (1) the central compartment of the body-cavity; (2) the heart; (3) the pericardial cavity; (4) the two lateral compartments or lateral sinuses (in which the nerve-cord and salivary glands lie). In addition to these there are the leg-cavities, which contain the nephridia and communicate with (4).”

We are now in a position to compare the body-cavities of Palæmonetes and Peripatus. Considering first the anterior region of the thorax of the former animal, as shown in figs. 6 and 9, we may compare it with the condition of things in Peripatus at the time when the dorsal portions of the mesoblastic somites have attained their maximum development. Bearing in mind that the dorsal sac of Palæmonetes has been formed by the union of two lateral cavities, which lay on either side of the aorta (cf. figs. 14, D, and 16, B), the differences between the two forms are very slight. The dorsal sac represents the two dorsal portions of the mesoblastic somites, whilst the central cavity, the lateral cavities, and the nephridia agree, with the one exception that the two lateral portions of the nerve-cord of Peripatus have united in the middle line in Palæmonetes, and in the process have passed out of the lateral cavities. The fact that the internal end of the nephridium lies in the lateral cavity, and not in the cavity of the leg, which, as already stated, contains the greater part of its tube, appears to agree with the condition of things figured by Sedgwick for certain parts of Peripatus (No. 19, pt. 4, pl. xxvii, fig. 11).

The agreement is so close that it appears to me to be fully justifiable to homologise the various parts. If this be so, it follows that the dorsal sac of Palæmonetes is homologous with the dorsal portions of the mesoblastic somites of Peripatus, and that its cavity is a true coelom (enterocœle). The central and lateral cavities, together with the cavities of the legs, will represent the pseudocoele, and, being filled with blood, may be termed with Lankester hæmocoele.

Passing now to the posterior part of the thorax, the region of the heart (figs. 18 and 19), we find that the different cavities correspond with those which persist in the adult Peripatus. Heart, pericardium, and pericardial septum of Palæmonetes present exactly thesame relations as in Peripatus, and are clearly homologous structures in the two animals. The central and lateral cavities only differ with regard to the relative position of the nervous system, and this difference has already been accounted for. In the leg-cavities of Palæmonetes in this region no nephridia have been developed.

Beneath the anterior end of the pericardial septum are found, as has been already stated, the genital organs (fig. 18, gen.), and here also the comparison with Peripatus may be instituted. We find a similar agreement to that which existed in the other regions compared, and we may with a considerable degree of probability again homologise corresponding parts of the two forms. The genital organs of Palæmonetes must then be regarded as the representatives in this region of the cœlom.

It is worthy also of note that the genital ducts, as seen in young adults of Palæmonetes, agree in all their relations with those of Peripatus, and the probability of their being derived from nephridia, as suggested by Lankester (No. 15), is very great.

If the homologies here suggested are valid, the body-cavity relations of the Crustaceans under consideration may be stated briefly thus:—Both enteroccele (true cœlom) and pseudoœle are present, the enterocœle consisting of the dorsal sac, the green gland and shell-gland, together with the genital organs and their ducts, whilst the pericardial septum may be regarded as equivalent to the fused walls of another portion of the same structure (cf. Korschelt u. Heider, No. 14, p. 901). The pseudocœle consists of the heart and arteries, the pericardial cavity, the central cavity of the thorax, with the lateral cavities and the cavities of the limbs, and the various sinuses of the abdomen. The whole of the pseudocoele is filled with blood, and hence can be termed a hsemocœle.

A few considerations may now be added with regard to that portion of the coelom which persists as the dorsal sac. The most striking feature of this cœlom is the late period of development at which it opens; but it may be pointed out that the cavity of the green gland (both of its tube and end-sac), which we also regard as cœlomic in nature, begins to appear at precisely the same time, the two cases differing, however, in all probability, in the fact that in the former the cells which subsequently form the walls of the cavity have been through a wandering stage (mesenchyme), whilst those of the latter, so far as we can judge, have not. This does not, however, I believe, alter the conclusions which have been arrived at by the comparison with Peripatus, and which are also supported by what is known of corresponding parts in Arachnids and insects.

The question suggests itself, what effect has the presence of this persistent coelom in the anterior part of the thorax upon our views as to the nature of the heart of these animals ? The view held by Claus, and which is generally accepted, is that the short heart of Decapods has been reduced from a heart similar to that of Branchipus, which extends throughout almost the whole length of the body. Claus (No. 3) further considers that the Decapod heart represents the anterior portion of the heart of Stomatopods, and the evidence which he brings forward in support of this view is very convincing. How are we to account for the fact that in the region of the body anterior to the heart of these Decapods a more primitive condition of things persists than is found even in the adult Peripatus? There appear to me to be two alternatives: either the heart and pericardium have at one time been differentiated along the whole length of the body, and there has been subsequent degeneration in the anterior portion to the more primitive condition, or we must assume that in the ancestor of the Crustacea the differentiation into heart and pericardium had never taken place at all in this anterior region. What evidence I have been able to gather appears on the whole to favour the latter view, although it is by no means conclusive, and the question can only be settled by further research. The following facts, however, have a bearing on the matter. In Branchipus the cephalic aorta is present, it has no ostia, and does not pulsate. It extends through the region of the maxillae to the second segment of the body (Claus, No. 6, p. 71). In the larva of Artemia the heart ends anteriorly at about the level of the shell-gland—that is to say, in the region of the second maxillae (ib., pl. iv, fig. 1). In Nebalia Claus figures a section which passes through the valves which separate the heart from the aorta, and this also passes through the point where the nerves of the second maxillae are given off from the nerve-cord (No. 8, pl. xi, fig. 11). Claus, however, states that this section is somewhat oblique, sloping downwards and backwards, so that we must place the anterior end of the heart at least as far forwards as the front end of the segment of Maxillae II.

In a Stomatopod larva (No. 2, pl. iv, fig. 8) the front end of the heart is figured as being slightly anterior to the shellgland.

In young larvæ of Palæmonetes, as I have already shown, the dorsal sac ends in the posterior part of the region of the second maxillae; but in all these cases it must be borne in mind that, owing to the moving forward of the appendages towards the mouth, the actual segment to which any part of the cephalic aorta or heart belongs can only be very approximately determined. Considering, however, the facts adduced, we may, I think, fairly say that in no Crustacean does the heart extend in front beyond the region of the second maxillae, and that there is therefore no evidence to show that heart and pericardium have ever been differentiated in segments anterior to this.

1. The green gland of Palæmonetes (and Palaemon) at the time of hatching of the larva has not developed a lumen, although the external opening can be detected. When the larva leaves the egg the lumen commences to open, and the gland consists of an end-sac and a U-shaped tube, of which the distal portion gives rise to the bladder. The bladder then enlarges greatly, growing at first inwards towards the middle ventral line, then upwards, within the oesophageal nerve-ring and anterior to the oesophagus, to the middle dorsal line, where it meets its fellow of the opposite side. The two bladders grow backwards over the stomach and beneath the dorsal sac, subsequently fusing together in the middle line to form the unpaired nephro-peritoneal sac.

2. The shell-glands are the functional excretory organs at the time of hatching and during the latter part of the embryonal period. They open at the bases of the second maxillæ, and each consists of an end-sac and a Y-shaped renal tube, which have the typical structure of a crustacean nephridium.

3. A dorsal sac, which is completely enclosed by an epithelial lining, persists in adults of Palæmo n, Palæmonetes, and Crangon. This sac, which does not contain blood, lies upon the nephro-peritoneal sac and the front end of the ovary, being much enlarged at its posterior end. The cephalic aorta (ophthalmic artery) lies within the dorsal sac.

4. At its anterior end the dorsal sac is surrounded by a mass of tissue which appears to be producing blood-corpuscles.

5. The dorsal sac is formed as a hollowing out in masses of mesoderm-cells, which lie on either side of the cephalic aorta. Two lateral cavities are thus formed, which increase in size and unite below the aorta. Taking into account this mode of development, a comparison with Peripatus shows that the dorsal sac is homologous with the dorsal portions of the mesoblastic somites of that animal, and must therefore be regarded as a true cœlom.

6. The body-cavity of these Crustaceans varies in different regions.

(a) In the anterior part of the thorax it consists of a true coelom (the dorsal sac and nephridia) and a hæmocœle. The hæmocœle consists of (1) a central cavity, in which the stomach and intestine, the liver and the nerve-cord lie; (2) two lateral cavities, which contain the end-sac and proximal end of the tube of the shell-gland, and which communicate with the central cavity and with the cavities of the legs; and (3) these leg-cavities, which, in the second maxillae, contain the tube of the shell-gland.

(b) In the posterior part of the thorax the body-cavity is entirely a hæmocœle. It consists of (1) the pericardial cavity, in which lies (2) the heart, and which is separated by the pericardial septum from (3) the central cavity of the body, which contains the genital organs, liver, intestine, and nerve-cord; (4) the lateral cavities, which communicate with the central cavity and with (5) the cavities of the legs.

(c) In the abdomen the body-cavity is entirely a hæmocœle. It consists of a dorsal and a ventral sinus, which communicate with one another by lateral sinuses.

1.
Allen
,
E. J.
—“
On the Minute Structure of the Gills of Palæmonetes varians
,”
‘Quart. Journ. Mier. Sei.,’
vol.
xxxiv
,
1892
.
2.
Claus
,
C.
‘Untersuchungen zur Erforschung der Genealogischen Grundlage des Crustaceen-Systems,’ Wien
,
1876
.
3.
Claus
,
C.
—“
Die Kreislauforgane und Blutbewegung der Stomato-poden
,”
‘Arb. Zoöl. Inst. Wien,’ Bd
.
v
,
1884
.
4.
Claus
,
C.
—“
Zur Kenntniss der Kreislauforgane der Schizopoden und Decapoden
,”
‘Arb. Zool. Inst. Wien,’ Bd
.
v
,
1884
.
5.
Claus
,
C.
—“
Neue Beiträge zur Morphologie der Crustaceen
,”
‘Arb. Zool. Inst. Wien,’ Bd
.
vi
,
1886
.
6.
Claus
,
C.
-—
“Untersuchungen über die Organisation und Entwickelung von Branchipus und Artemia,“
‘Arb. Zool. Inst. Wien,’ Bd
.
vi
,
1886
.
7.
Claus
,
C.
“Ueber Apsendes Latreillii, Edw., und die Tanaiden,” II
,
‘Arb. Zool. Inst. Wien,’ Bd
.
vii
,
1888
.
8.
Claus
,
C.
—“
Ueber den Organismus der Nebaliden und die syste matische Stellung der Leptostraken
,”
‘Arb. Zool. Inst. Wien,’ Bd
.
viii
,
1889
.
9.
Milne
Edwards
. —
‘Histoire naturelle des Crustacés,’ Paris
,
1834
.
10.
Gkobben
,
C.
—“
Die Antennandrüse der Crustaceen
,”
‘Arb. Zool. Inst. Wien,’ Bd
.
iii
,
1881
.
11.
Grobben
,
C.
—“
Die Entwickelungsgeschichte von Cetochilus septentrionalis
,”
‘Arb. Zool. Inst. Wien,’ Bd
.
iii
,
1881
.
12.
Kennel
,
J. von
. —
“Entwickelungsgeschichte von Peripatus Edwardsii und P. torquatus,” I und II
,
‘Arb. Zool. Inst. Würzburg,’ Bde. vii and viii, 1885
,
1886
.
13.
Kingsley
,
J. S.
—“
The Development of Crangon vulgaris
,”
‘Bull. Essex Inst.,’ 1887
,
1889
.
14.
Kohschelt
,
E.
,
und Heider
,
K.
‘Lehrbuch der Vergleichenden Entwickelungsgeschichte der Wirbellosen Thiere,1 Jena
,
1891
.
15.
Lankester
,
E. R.
—“
Note on Memoir by Gulland, ‘Evidence in Favour of the View that the Coxal Gland of Limulus and of other Arachnid" is a Modified Nephridium,’”
‘Quart. Journ. Mier. Sei.,’
xxv
,
1885
.
16.
Lebedinski
,
J.
—“
Einige Untersuchungen über die Entwickelungs geschichte der Seekrabben
,”
‘Biol. Centralbl.,1 Bd
,
x
,
1890
.
17.
Marchal
,
P.
—“
Recherches anatomiques et physiologiques sur l’appareil excréteur des Crustacés Décapodes
,”
‘Arch. Zool. expér.,’ t
.
x
,
1892
.
18.
Reichbnbach
,
H.
‘Studien zur Entwickelungsgeschichte des Fluss krebses,’Frankfurt
,
1886
.
19.
Sedgwick
,
A.
“The Development of the Cape Species of Peripatus,” parts 1—4, ‘
Quart. Journ. Mier. Sei.,1
vols,
xxv—xxviii
,
1885
1888
.
20.
Weldon
,
W. F. R.
—“
The Renal Organs of certain Decapod Crus tacea
,”
‘Quart. Journ. Mier. Sei.,’
vol.
xxxii
,
1891
.

Illustrating Mr. Edgar J. Allen’s paper on “Nephridia and Body-cavity of some Decapod Crustacea.”

Ant. a. Antennary artery. Ao. c. Cephalic aorta.. b. c. Central bodycavity. bl. Bladder of green gland, b. lat. Lateral cavity of body. c. Dorsal sac (cœlom), ear. Carapace, e. m. Cephalic ‘muscle, con. t. Connective tissue, cor. Blood-corpuscle, cut. Cuticular layer of renal tube. E. Point at which bladder enlarges, ect. Ectoderm, e. s. End-sac. ex. o. External opening, gen. Genital organs, ht. Heart, int. Intestine. I. g. Tissue where blood-corpuscles are formed (lymphatic gland), li. Liver lobe. mes. Mass of cells from which dorsal sac develops, mus. Muscle. Mr. II. Maxilla II. n. c. Nerve-cord. 0, Point where bladder of green gland joins ureter and opens, o. es. Opening of end-sac into tube. per. Pericardium.p. sep. Pericardial septum, rt. gl. Reticulate gland, s. Septum dividing dorsal sac. sal.gl. Salivary gland, sh. gl. Shell-gland. "/.Stomach, st. a. Sternal artery. Cu. Renal tube. u. I. Upper lip.

All the figures were drawn with the aid of the camera lucida.

FIG. 1.

Horizontal longitudinal section through the green gland of a larva of Palæmonetes varians three or four days old. × 420.

FIG. 2.—Transverse section of green gland of Palæmonetes at the time of hatching, × 770.

FIG. 3.—Ditto, a few hours after hatching. × 770.

FIG. 4.—Somewhat diagrammatic transverse section through-green gland of old larva of Palæmonetes. × 240.

FIG. 5.—Transverse section of young adult Palæmonetes, just behind mouth, × 240.

FIG, 6.—Transverse section of Palæmonetes larva, ten days old, through the region of Maxilla II. × 240.

FIG. 7.—Transverse section through shell-gland of a recently hatched Palæmonetes larva. (The figure is combined from two sections.) × 240.

FIG. 8, A, B, o.—Three sagittal sections of ditto, of which A is the most internal, B the most external. × 240.

FIG. 9.—Transverse section of Palæmonetes larva, about twenty-four hours old, through the region of Maxilla I. × 240.

FIG. 10.—Palsemon serratus dissected from the dorsal surface.

FIG. 11.—Sagittal section of larva of Palæmonetes, ten days old. × 100.

FIG. 12.—Transverse section through anterior end of dorsal sac of young adult Palæmonetes, showing the tissue from which blood-corpuscles are budded off. × 240.

FIG. 13, A, B.—Two sections through the region of cephalic aorta of a Palæmonetes embryo, A represents the condition at anterior end, whilst B represents the condition further back, n’ = nucleus of internal layer, n′ = nucleus of external layer, × 240.

FIG. 14.—Transverse sections through cephalic aorta of Palæmonetes larva, within a few hours of hatching. × 240.

A. About level of mouth.

B. Region of paragnaths.

c. Ditto.

D. Region of Maxilla I.

E. Region of Maxilla II.

FIG. 15.—Horizontal section through anterior end of cephalic aorta of larva slightly older than that of Fig. 14. × 240.

FIG. 16, A, B.—Transverse sections of dorsal sac and aorta of Paltemonetes larva, a few hours old. × 240.

FIG. 17.—Transverse section through posterior portion of dorsal sac of Paltemonetes larva, three weeks old. × 240.

FIG. 18.—Transverse section of just-hatched larva of Palæmonetes, through front end of pericardium, × 240.

FIG. 19.—Ditto through middle of pericardium, × 240.

FIG. 1.

Horizontal longitudinal section through the green gland of a larva of Palæmonetes varians three or four days old. × 420.

FIG. 2.—Transverse section of green gland of Palæmonetes at the time of hatching, × 770.

FIG. 3.—Ditto, a few hours after hatching. × 770.

FIG. 4.—Somewhat diagrammatic transverse section through-green gland of old larva of Palæmonetes. × 240.

FIG. 5.—Transverse section of young adult Palæmonetes, just behind mouth, × 240.

FIG, 6.—Transverse section of Palæmonetes larva, ten days old, through the region of Maxilla II. × 240.

FIG. 7.—Transverse section through shell-gland of a recently hatched Palæmonetes larva. (The figure is combined from two sections.) × 240.

FIG. 8, A, B, o.—Three sagittal sections of ditto, of which A is the most internal, B the most external. × 240.

FIG. 9.—Transverse section of Palæmonetes larva, about twenty-four hours old, through the region of Maxilla I. × 240.

FIG. 10.—Palsemon serratus dissected from the dorsal surface.

FIG. 11.—Sagittal section of larva of Palæmonetes, ten days old. × 100.

FIG. 12.—Transverse section through anterior end of dorsal sac of young adult Palæmonetes, showing the tissue from which blood-corpuscles are budded off. × 240.

FIG. 13, A, B.—Two sections through the region of cephalic aorta of a Palæmonetes embryo, A represents the condition at anterior end, whilst B represents the condition further back, n’ = nucleus of internal layer, n′ = nucleus of external layer, × 240.

FIG. 14.—Transverse sections through cephalic aorta of Palæmonetes larva, within a few hours of hatching. × 240.

A. About level of mouth.

B. Region of paragnaths.

c. Ditto.

D. Region of Maxilla I.

E. Region of Maxilla II.

FIG. 15.—Horizontal section through anterior end of cephalic aorta of larva slightly older than that of Fig. 14. × 240.

FIG. 16, A, B.—Transverse sections of dorsal sac and aorta of Paltemonetes larva, a few hours old. × 240.

FIG. 17.—Transverse section through posterior portion of dorsal sac of Paltemonetes larva, three weeks old. × 240.

FIG. 18.—Transverse section of just-hatched larva of Palæmonetes, through front end of pericardium, × 240.

FIG. 19.—Ditto through middle of pericardium, × 240.