The work which forms the subject of the present memoir was carried on at the Zoological Station at Naples during the six months from October, 1891, till April, 1892.

My original intention was to study the development of the genital organs in an Echinoderm, following a suggestion of Mr. Sedgwick. On my arrival at Naples I found that Amphiura squamata was the only form available for my purpose, since this viviparous Ophiurid is the only Echinoderm which breeds in late autumn (as, indeed, it does at every other season of the year as well).

In tracing the origin of the primitive germinal cells I was led, as the sequel will show, to examine the development of the ovoid gland, the “Centralgeflecht des Blutsystems “of Ludwig, together with the sinuses in connection with it; and, as far as I can make out, a systematic account of the history of these problematical organs is presented in this paper for the first time.

Before proceeding to describe the methods employed, I shall say a word or two as to the suitability of Amphiura squamata as an object for this study. It has both advantages and disadvantages. The former are (1) that it breeds the whole year round ; (2) that it retains the young in its genital bursæ till a very late stage of development; and (3) that the adult is of such a small size as to render it a very favorable object for microscopic study. Hence it will be understood that one can always obtain an unbroken series of stages; and those phases frequently occur which in other Echinoderms are most difficult to obtain, viz. those immediately following on the close of the larval period. On the other hand, the great disadvantages are (1) that the Ophiurids are probably the most modified group of the Echinoderms; (2) that the development of Amphiura squamata, being more or less direct, is likely to be somewhat altered ; (3) that the embryos are excessively small, my earliest post-larval stages having a diameter of only about ·2 mm., whilst the unsegmented eggs of Antedon rosacea have a diameter of ‘3 mm.

With regard to the second disadvantage mentioned above, one must remember that a larval form can be recognised though much simplified in comparison with the ordinary Ophiurid Pluteus; and that the organs treated in this paper are less likely to be modified in their development, since they are all post-larval. The small size of the embryos forced me to give up the attempt to investigate the origin of some of the sinuses, such as the “radial perihsemal “and the “perihaemal ring sinus.” So also I can make no statements with certainty as to the origin of the subneural sinus; except this, that Cuenot (6) is most certainly wrong when he asserts that it is formed by an involution of the ventral ectoderm of the arm. As a matter of fact, the nerve-cord has attained its deep position from the time of the very first appearance of the arms, and the figure which Cuenot gives has no counterpart in nature. All the sinuses mentioned are at first exceedingly narrow slits, considerably less in diameter than the nuclei of the surroundiug tissue; and as the cell plasma of Echinoderm mesoderm is difficult to stain, one can realise how easy it is to be mistaken about the extent and communications of such cavities. Added to all this, the coelom in Amphiura squamata is represented at first by a mass of mesenchyme; and as this condition of things is certainly not primitive, I do not think that even if reliable results as to the development of cavities originating before the coelom clears were attainable, they would be of much phylogenetic importance. I hope to be able to find something more instructive when I study their history in a larger and more typical form.

I. Fixing

—The following is a list of the preserving fluids employed. Corrosive sublimate in distilled water; corrosive sublimate in sea water; corrosive sublimate three parts, glacial acetic one part; chromic acid; picric acid; Flemming’s fluid; glacial acetic; alcohol of 30 per cent.; hot alcohol of 70 per cent, with a few drops of corrosive sublimate; osmic acid from about to 1 per cent.; osmic acid followed by Müller’s fluid for eighteen to twenty hours.

THE RESULT OF MY EXPERIMENTS IS THIS

—The only liquid which gives reliable results is osmic acid, though its use is attended by disadvantages. It renders the animals very brittle, and has little penetrating power. When used in too strong solution, or for too long a time, it is followed by considerable shrinkage. I have found, however, that when it was followed by Müller’s fluid this shrinkage was entirely prevented, and the brittleness, though still considerable, was very much less. Alcohol I rejected on account of the imperfect preservation it gave. Corrosive sublimate would give very fair results if it were possible to render its solution quite neutral, but this is very difficult. In no case, however, could one obtain the beautiful differentiation of tissues as to their staining properties which is yielded by osmic acid.

All liquids which decalcified as well as fixed were of no use, on account of the cavities which the evolution of gas produced in the still soft tissues. The method I finally adopted was as follows :—The animals were killed in a solution of osmic acid of about per cent, allowed to act for ten minutes or more, of 1 per cent, acting for five minutes only; they were then rinsed with water and transferred to Müller’s fluid for eighteen to twenty hours, then put at once into alcohol of 30 per cent., and brought slowly up into alcohol of 90 per cent. In the latter they were hardened for a night; then two or three drops of nitric acid were added to some fresh alcohol of 90 per cent., so as to give roughly a solution of from to 1 per cent, of acid, and the animals were immersed in this till decalcification was complete, a process which occupied not more than twenty hours.

I found that Müller’s fluid itself has a slow decalcifying action, and in some cases the acid alcohol was unnecessary. For this reason, were I doing the work again, I should prefer to stop the action of the osmic acid with ammonium picro-car-minate instead of Müller’s fluid, so as to avoid even the slight and easily recognisable “artefacts” produced by the latter. Glacial acetic acid gave fair results, and I used it to confirm results obtained by osmic acid. Strange to say, it has little or no decalcifying effect, and the suddenness of its action and its penetrating power are unsurpassed. But it is apt to cause swelling of the tissues, and of course on transferring the animals to alcohol one gets a solution of acetic acid, which then decalcifies before the tissues are hardened.

II. STAINS

—Double staining was used in order to be certain about the boundaries of sinuses, since the ordinary plasma of Echinoderms stains with great difficulty.

As a nuclear stain I used Dr. Mayer’s paracarmine. The great advantages about this are that it acts rapidly, and that all superfluous stain can be extracted by 70 per cent, of alcohol, which can be allowed to act for an indefinite time. I allowed my earlier stages about twenty minutes in the stain, adults and later stages an hour or two. An hour or two suffices to remove all surplus stain, provided the alcohol be changed several times. The plasma stain was applied on the slide. I used two, both of which gave good results, viz. solution of picric acid in turpentine and Dr. Mayer’s oxidised hæmatoxylin pr “hæmatëin,” The advantage of the former is that it can be used with the shellac method of mounting, and the danger of staining the mounting agent, as in the case of glycerine and albumen, is avoided. On the other hand, hsematein gives more differentiation. For embryos preserved in glacial acetic acid I used Dr. Mayer’s hsemalaun, which gives a blue nuclear stain, and at the same time colours much of the plasma a faint yellow.

III. CUTTING AND MOUNTING

—Embryos fixed and stained according to my standard method were embedded in paraffin, and cut into series of sections in a plane parallel to the line joining the madreporite with the mouth, and at the same perpendicular to the plane of the disc.

A word or two on their orientation may not be out of place here. The young were in all cases extracted from the mother, and the madreporite found by examination with the dissecting microscope. Then, if the arms were long, all were broken off except that which was opposite to the madreporite; but if they were short and comparatively stiff only this arm was broken off. In the case of my earliest stages, where the arms were undeveloped, it was necessary to break off one point of a pentagon, and often it occurred that the animal (if only about ·22 mm. in diameter) was destroyed or the breakage went obliquely, so as to mislead one afterwards. In general, out of three series of sections of this stage only one was satisfactorily orientated. This orientation had in all cases (except in the latest stages) to be performed on the living animal. When the animals had been embedded the paraffin was cut away, so as almost to lay them bare on one side, and the position of the madreporite was thus recognised, and-the orientation of the block effected. Fresh paraffin was then melted on to the exposed place.

I must here raise an emphatic protest against the method employed by Cuénot (6), who professes to give an account of the origin of the ovoid gland and its surrounding sinus in Amphiura squamata, and is naturally quite wrong. He cuts a number of adults containing embryos, and then founds conclusions on casual sections of the contained young. The preserving fluid in this case penetrates neither rapidly nor effectively, there is no control over the orientation, and it is impossible to get thin enough sections. I found it necessary to use for my youngest and most important stages sections thick (five teeth of the rocking microtome). Above a diameter of ·3 mm. my sections were μ thick (seven teeth), and only for adults and the oldest stages did I use sections 7 μ thick (ten teeth).

Sections perpendicular to the disc, but transverse to the axis of the stone canal, and horizontal sections were also used, but proved to be not nearly so instructive as those orientated as described above; they were useful, however, in determining some minor points.

My results are founded on perfect series of sections with finely differentiated stain and clear, sharp outlines. The sections themselves are clearer and more diagrammatic than the figures I have been able to make of them. In fact, I obtained preparations which for distinctness of outline and diagrammatic alternation of colour could hardly, I think, be surpassed by the use of osmic acid, Müller’s fluid, paracarmine, and hæmatin.

It will be most convenient to first of all detail my results, and then to examine and discuss the literature; and as the ovoid gland precedes in development the genital organs, it will be well to commence with it.

I. Structure

—This has been described by Cuenot (4), Hamann (7), Apostolides (1), Ludwig (11), and Kohler (9). They agree in describing a pyriform mass enclosed in a space limited by a strong membrane, and almost filling it. Most agree in finding a supporting network of fibres with meshes partly occupied by cells, partly forming lacunae. This mass sends out a prolongation, gradually becoming like ordinary connective tissue, towards the nerve-ring. On the dorsal surface of the latter it enters into connection with a structure called the haemal ring. At its lower end the gland is prolonged into a curious string of tissue enclosed in a prolongation of the sinus which surrounds the gland. This string (the genital rachis) is ventral in the interradii, but at the sides of the radii it mounts to the dorsal surface, thus passing over the genital bursæ, which are involutions of the integument close to each side of the arm. On the interradial side of these bursse the rachis bears the ovary; on the radial side it gives off a branch terminating in the testis, which runs into the short extension of the body-cavity into the arm, called by Russo (11) the testicular chamber.

Fig. 2 a is a longitudinal section of the disc passing through the ovoid gland of an adult. Fig. 2 b is another section from the same series, and fig. 2 c is a horizontal section of the gland of another individual.

In the centre of the organ one sees the stone canal (st. c.) with its characteristic columnar epithelium, which dorsally passes into the flattened epithelium of the water-vascular ring (w. v.,r.). A section of the pore canal (p. c.), leading to the madreporie pore, is seen in the body-wall. I find three spaces in connection with the ovoid gland, and, as they have all been confused together under the name of axial sinus, I have called them sinus a, b, and c respectively, though I do not thereby mean to imply that they belong to the same category.

Sinus c is the thin-walled ampulla, into which both stone canal and pore canal open, and it has no connection whatever with any other cavity. Sinus a is the part of the aboral sinus which comes in contact with the ovoid gland, and is flattened out on its external surface. Sinus b is the axial sinus proper, which is principally developed round the distal swollen end of the gland, but also extends up on to its sides. These three spaces—sinus a, sinus b, and sinus c—are all quite independent of each other. I have traced them in every well-preserved series which I have examined. As to the structure of the gland itself, I find that it often presented the areolated texture referred to above. This I believe to be entirely due to imperfect preservation. When sufficiently strong osmic acid is used, and precautions taken to ensure its penetration (by opening the animal whilst living), one sees a uniformly staining plasma (figs. 2 b, 2 c) supported by a close network of interlacing fibres. Intermingled are cells whose bodies stain slightly differently from the general plasma of the organ. In fig. 2 d I have drawn some curious nuclei from the ovoid gland of another adult, which seem to me to point at direct cell division. In the extension of the organ towards the nerve-ring the plasma gradually disappears, aud it takes on the form of the ordinary connective tissue; but there seems to me to be no reason to regard the spaces in this as in any special sense lacunae. In the part which is enclosed by the sinus a some cells are seen with larger and better defined bodies than in other portions of the gland. These are Hamann’s “Urkeimzellen,” primitive germ-cells, of which the genital rachis is also composed. Their meaning will be understood when the genital rachis is treated of.

II. DEVELOPMENT

—Fig. 3 represents a section (in the standard direction) through an embryo about ·2 mm. in diameter. The stone canal is seen lying in the middle of a mass of cells which take the place later occupied by the coelom ; the thickening on the inner side of the canal is an indication of the presence already of the ampulla. Unfortunately the other sections of this series are somewhat broken in this place, and in my other series of the same stage the stain was not good enough to give a clear figure of this cavity. I have no doubt, however, that it exists at this stage.1

Russo (17) has mistaken the mass of cells filling the coelom for the rudiment of the ovoid gland, but the next stage (which I have obtained at least half a dozen times) clearly shows his error. Figs. 4 a, b, c, represent three consecutive sections . thick through a slightly older embryo (the size is not an exact guide to age). As these three sections comprise the whole stone canal and surrounding rudiment of the gland, the importance of dealing only with perfect series of very thin sections will at once be perceived.

A single layer of nuclei (ps.) lining the coelom, which therefore can be called peritoneal, can be seen in fig. 4 b covering the stone canal. This is the first rudiment of the ovoid gland. The coelom can be seen to run into a chink at the distal end of the stone canal; a single thread of plasma stretches across it. This chink is the commencement of the axial sinus (sinus b), which is thus a diverticulum of the coelom. The ampulla (sinus c) is seen to be quite distinct, and by comparing figs. 4 b and 4 c may be seen to communicate with m. p., the madreporic pore.

In fig. 5 a, and still more distinctly in fig. 5 b, we see an increase in the number of nuclei forming the rudiment of the ovoid gland. At the point where a mesenteric bridle goes off to the stomach they form more than one layer, and show the characteristic staining properties of the primitive germ-cells. The axial sinus is clearly seen, and though its lumen is the merest slit, its continuity with the cœlom can easily be traced by its walls.

In the next stage (figs. 6 a, b, c) the multiplication of the nuclei forming the ovoid gland is very marked, and they project as a lobe which encloses under it a nook, so to speak, of the coelom. This is the first rudiment of the aboral sinus, which, like the axial sinus, is an involution of the cœelom, but quite a distinct one, though of course both at this stage open freely into the coelom.

The next and most important stage is represented in figs. 7 a, b, c. Here the ovoid gland has increased in size, and has grown round the stone canal, so that in the upper part of the section it appears both external (i. e. towards the mouth) as well as internal to it. In the dorsal portion of the gland the nuclei are spaced widely, and the characteristic plasma is beginning to appear. In the ventral portion the nuclei are very lafge, and are, in fact, becoming the primitive genital cells, the “Urkeimzellen.” Sinus b is now distinctly closed off by a double membrane from sinus a, which opens freely into the cœlom. Figs. 7 b and 7 c show, however, that sinus a is prolonged laterally as an involution, and ends in a solid knob of cells. This is the beginning of the lateral outgrowth of the aboral sinus. It is remarkable that it takes place only on one side of the ovoid gland; and this side, as shown by comparison of vertical sections transverse to the axis of the stone canal, is the right, if we suppose the animal to be placed with its mouth down and the madreporite posterior.

That the interpretation given to these rudiments is the right one is convincingly shown by the next stage (figs. 9 a, b). Here both axial and aboral sinuses are shut off from the coelom and from one another. The Urkeimzellen are more distinct, and one sees also smaller interstitial cells amongst them. In this stage also I have been able to detect a trace of the characteristic fibres of the adult gland. In fig. 9 b one sees an unmistakable rachis (compare fig. 11) constricted off from the swollen ventral part of the ovoid gland; and yet, traced forward for a section or two, this ends in a similar knob of cells to that represented in fig. 8 c. This will, I hope, convince every one that the aboral sinus grows out from the involution at the base of the ovoid gland. These intermediate stages are, however, exceedingly difficult to find; and in the majority of cases when one seeks them one finds in embryos of the same size either that the rachis is not formed, or that it is formed the whole way round. I have, however, succeeded in getting about half a dozen specimens with a rudimentary rachis, though this was in all cases very short.

I have said that the aboral sinus extends laterally as a tubular involution, and of course the genital rachis is an outgrowth (as seen from figs. 9 a, b) of the similar part of the ovoid gland. If we now compare fig. 10 (which is a section in a plane somewhat oblique to the standard direction) we see the perfect continuity of the primitive germ-cells in the genital rachis with those in the ovoid gland. At the same time we notice a very interesting change in their form. From being spherical in the gland their bodies in the rachis become long and fusiform, and overlap each other, and even show irregularities of outline, which might be termed pseudopodia. Their nuclei at the same time change from a spherical to an oval shape.

These facts, I think, strongly support Hamann’s idea that the primitive germ-cells actually migrate, and I should be inclined to suggest that their place of proliferation continues to be where it is undoubtedly at first, viz. in the base of the ovoid gland. This wandering of the Urkeimzellen is also suggested by the observation which I have made, that when the rachis is complete there is great variation in the position of the primitive germ-cells. Sometimes there are a large number close to the ovoid gland and sometimes very few, and once just close to it the aboral sinus was quite empty.

The last stage in the development of the ovoid gland is shown in fig. 11. Here we see how the axial sinus, sinus b, extends laterally so as to enwrap the gland. This is a condition of things which seems to occur sometimes sooner and sometimes later.

To demonstrate this, clear and accurate longitudinal sections are required, but indeed this holds for all these sinuses. Very often, in spite of all precautions, one finds their lumen quite obliterated, but when one succeeds in preventing this their relations are very evident.

Before leaving the discussion of fig. 11 I wish to call attention to the curious space sk. This I have called a shrinkage space, because it is due to the different behaviour of the great interradial muscle, mi., to the alcohol and fixing reagents from that of the neighbouring tissues. I mention it specially because I have reason to believe that it has been mistaken for part of the axial sinus by Hamann (7), and given rise to his idea that the axial sinus communicates with the sinus lying dorsal to the nerve-cord, a communication of which I can find no trace in any of my sections. Of course it may be different in the species which Hamann examined. This shrinkage space is quite easy to distinguish from the various sinuses described, as it has no endothelial membrane for its boundary as they all have.

Fig. 8 is a section from an individual slightly younger than the one from which fig. 11 is taken; it is interesting as showing the opening of the ampulla, both into the pore canal and into the stone canal; it also shows very distinctly the strong double membrane which separates the aboral from the axial sinus.

HÆMAL SYSTEM

Before proceeding to describe the development of the genital organs it will be well to detail my observations on the haemal system, as the composition of the genital rachis is connected with this subject. The view (accepted from Ludwig [11] and Hamann [7]) which is commonly taught in text-books is that there is a true haemal system of vessels filled in preserved specimens with a deeply staining clot. The aboral sinus is said to enclose one of them, which in its turn surrounds the genital rachis. Another ring-shaped blood-vessel is said to lie close to the dorsal surface of the nerve-ring, and to send branches along the mid-dorsal line of the radial nerves. Finally the ovoid gland has been described as a plexus of vessels connecting these two rings.

IN AMPHIURA SQUAMATA CONDITIONS ARE AS FOLLOW

—A section of the genital rachis shows no trace of a blood-vessel. In the centre one sees the characteristic primitive germ-cells with large oval nucleus, and well marked body which also stains well. Usually not more than two of these are seen in one section of the rachis. Associated with them are a number of smaller interstitial cells. These form a kind of epithelium over the surface of the rachis which is turned towards the sinus. The outer wall of the latter (i. e. the wall projecting into the cœlom) is covered with the ordinary peritoneal epithelium. The formation of these interstitial cells can be seen in figs. 7 a and 9 a proceeding contemporaneously with that of the primitive germ-cells. There is no particle of the section which is not accounted for by the nucleus or plasma of some cell.

That the ovoid gland is no network of lacunae I have already pointed out. In figs. 7 a, 9 a, and 11 this organ can be seen to send out a prolongation which extends on to the upper and outer surface of the nerve-ring. This prolongation becomes thinner and thinner as it proceeds, but there is no trace of any swelling of this to form a blood-vessel, and it can only be traced in the few sections of the nerve-ring which also contain the ovoid gland.

As to radial blood-vessels—in many sections of the radial nerve-cord—there does appear at first sight to be a small round structure, in the mid-dorsal line, between two masses of ganglion-cells. But if one uses careful double staining and examines successive sections, one sees that it is often absent— as, for instance, in fig. 21. There the ganglion-cells with nuclei and plasma cover the whole dorsal surface of the cord. Further, when one looks closely at the sections where there is an appearance of a blood-vessel (cf. fig. 22) one sees that the nuclei of its wall are exactly like the ganglion nuclei; that its clot stains almost if not quite the same as the plasma of the surrounding cells. It is limited by no cell layer from the nerve-fibres below, and its boundary towards them is often jagged and uneven, and the angles of this outline run out into vertical fibres, so as to leave no doubt in my mind that the so-called blood-vessel is merely composed of the cell plasma of two or three rather larger dorsal ganglion-cells which are prolonged into these vertical fibres. A remarkable confirmation of this view is obtained by the study of specimens which have remained too long in osmic acid (vi. d., fig. 23). Here the mass of nerve-fibres is strongly shrunk and its shape altered, so that its dorsal outline is far less convex. At the sides it is more or less wrenched away from the dorsal ganglion-cells, leaving an artificial space; but in the middle one or two adhere, and produce the appearance of a blood-vessel. As to the so-called branches of the haemal system which go to the alimentary canal, these seem to me to be nothing more than mesenteric bridles; one very marked one is inserted just at the base of the ovoid gland, and in some young specimens there appears to be a prolongation of the peculiar tissue of the gland along it. In adults its relations are obscured by the great extension of the genital bursæ.

GENITAL ORGANS

The formation of the genital organs need not detain us long, as it is very simple. Amphiura squamata is hermaphrodite, and the general disposition of the genitalia has been described when speaking of the adult ovoid gland.

The ovaries appear first as small swellings of the genital rachis (compare figs. 16 a and b). The rachis here lies in the sinus in such a way that it touches the wall of the bursa from the first. The latter is here very thin, and one can detect a small involution towards the ovary. Here of course, when fully developed, is the point of discharge of the organ. The further history of the ovary merely consists in the increase of the number of the primitive germ-cells and of the subsidiary cells. Finally, one or two primitive germ-cells get the upper hand and form the fully developed ova. Of these there is usually only one in an adult ovary; when it is discharged one of the undeveloped primitive germ-cells takes its place.

Fig. 17 shows the passage of the young ovary into the rachis, and fig. 18 shows the ovary of a small adult.

When the ovary commences to form, the branch of the rachis leading round the bursa into the radial testicular chamber is undeveloped. Its formation is shown in fig. 12.

The first stage in the development of the testis (fig. 13) is indistinguishable from that of the ovary, viz. a solid mass of cells, except that these appear to consist more exclusively of Urkeimzellen. Afterwards a lumen forms as the rudiment increases in size (fig. 14), as sometimes occurs in old ovaries when more than one egg comes to maturity at a time. The Urkeimzellen are finally confined to one or two peripheral layers, and the centre is occupied by the products of their division. In neither the ovary nor testis did I see the least trace of a matrix of plasma and fibres, such as Cuénot (4) has described. Into the maturation of the sexual products it is not my purpose to enter. I only call attention to the following fact. Cuénot (4) states that in the testis the central Urkeimzellen are destroyed. I have seen no trace of this; the lumen appears to me to come in consequence of the increasing periphery of the organ. It is possible, however, that some of the internal primitive germ-cells divide completely to form spermatozoa, and leave no part to regenerate the mothercell. Young testes are often quite a solid mass of spermatozoa and sperm mother-cells, and no trace of degenerate cells is ever visible.

The genital bursæ originate late as invaginations of ectoderm (Russo, 16). There is, however, a very curious point about their development hitherto unnoticed. The first sign of their appearance is an increase in size and number of the peritoneal cells (fig. 19). Afterwards the ectodermal invagination is apparent (fig. 20). This first rudiment of the bursa is often very like a portion of the genital rachis; and as it is found at a stage when this is imperfectly or not at all formed, it is apt to lead to confusion. The best methods of distinction are of course the absence of a sinus, and its non-persistence on succeeding sections. When looking for specimens with a rudimentary rachis, one must carefully distinguish this rudiment of the bursa.

I. OVOID GLAND

—” Hæmal and Perihæmal Systems.”

In discussing this subject I think it neither necessary nor profitable to go back beyond the time of Ludwig, as before his time the ovoid gland was either not recognised or confounded with a Polian vesicle; whilst any space that happened to be seen in cross-section was called indiscriminately a bloodvessel. Ludwig (10, 11) not only distinguished the ovoid gland and homologised it with the organ occupying a similar position in other Echinoderms, but discovered the ampulla and distinguished it from the axial sinus; found the aboral sinus with its contained blood-vessel and rachis, and the relation of these to the genital organs. He also described an oral blood-ring and radial blood-vessels. The ovoid gland he regarded as a plexus of vessels uniting the oral and aboral blood-rings.

Like every student of Echinoderms, I cherish a great respect for the judgment of Ludwig; at the same time I must express regret at what seems to me to be the roughness of his methods. His figures are most diagrammatic; his distinction of the ampulla from the axial sinus is, by his own confession, mainly founded on one lucky section; and his figures of the blood-vessels are such as to render it impossible to conclude anything as to the histological nature of what he saw. Of course, at the time when the work was done the modern technique was undeveloped.

The work of Christo-Apostolides (1) was conducted with such crude methods, and contains results so contradictory of every one else, that its discussion would, I think, be mere waste of space. Hamann (7) has given a valuable paper on the Ophiurideae; he follows Ludwig’s interpretations, but gives far better figures. But he himself thinks it necessary to apologise for the imperfect preservation of many of his specimens. His results are chiefly founded on Ophioglypha albida; in venturing to criticise them I do so with some reserve, as I have only studied Amphiura squamata, but I cannot help thinking that my results throw some light on the appearances described by Hamann and others. Hamann figures the ovoid gland enclosed in a great sinus, which communicates with the perihsemal ring canal. I suspect that he has confused shrinkage spaces with part of the axial sinus. The ovoid gland he found to be lacunar when he used ½ per cent, osmic acid. I have already pointed out the great care which must be taken to ensure the due penetration of this fluid; when I used weak osmic, and left the animals only a short time in it, I got the ovoid gland very “maschig,” the plasma being, in fact, macerated out from the supporting fibres. As to the blood-vessels, I call attention especially to the following passages:—”Die Wandung ist eine diinne Membran “Dieser Membran liegen aussen ovale Zellkerner auf “Irgend welche Zellsubstanz ist nicht Vorhanden” Die Blutflüssigkeit stellt eine gerönne Masse dar welche sich mit neutraler Karminlösung helbrosa färbt.” Now it seems possible to me that the missing cell plasma of the oval nuclei which are on the wall of the blood-vesselis to be found in the blood fluid; for this latter stains brightly, and always quite fills the vessel. The aboral ring has, according to Hamann, a different constitution. Its wall consists of well-marked cells, and there are scattered cells in the “fluid?’ I believe the suggestion has been made, though I cannot find the reference, that the aboral blood-ring results from a degeneration of the rachis. I have myself seen a structure like what Hamann describes in the large non-viviparous species, A mphiura Chiajii; which has, like most Ophiurids, a special breeding season. Amphiura squamata, on the other hand, breeds all the year round; and hence one can understand why there should be no degeneration in its rachis. Why, however, the degenerate part of a rachis should be called a blood-vessel I fail to see.

Köhler (9) has examined Ophioglypha texturata, and has arrived at very curious results. I must, however, protest against the practice of painting in colours the blood-vessels in figures professing to give histological details. This mixing of diagram and figure seems to me quite to beg the question. Köhler fails to find the aboral sinus, but sees two prolongations of the ovoid gland attaching themselves to the sides of the madreporic plate. This I believe to be due to a misconception of a section like my fig. 11, in which the gland is cut tangentially, and the boundaries of the axial sinus might simulate prolongations such as he describes. The enormous shrinkage cavity one sees in his figure, and his failure to detect the proper relations of the aboral part of the gland, point, in my mind, again to imperfect preservation. The oral blood-ring he finds to be constituted of a mass applied to the dorsal and outer wall of the perihæmal canal, consisting of an irregular meshwork of fibres with cells and brown granules. This is certainly a curious structure to call a blood-vessel, especially when it is remembered that, excepting the brown granules, this is exactly the appearance of the ordinary decalcified tissue of an Ophiurid. The only reason Köhler has for singling out this as a blood-vessel would appear to be that when he forced an injection into the mass of the ovoid gland he got the colouring matter here. When one thinks of the thin and non-resistent nature of the boundary walls of sinuses, especially in cases of poor preservation, and the little firm tissue (other than lime) there is in the body-wall of Ophiurids, one will not be inclined to place much dependence on results obtained by injection. The radial blood-vessels Köhler finds, consist of fibres, cells, and pigment; and what strong distinguishing mark there is between them and the dorsal ganglioncells of the nerve-cord appears neither from Köhler’s figures nor description.

Cuénot, in his earlier work (4), denies the existence of bloodvessels, and traces cavities lined with cubical epithelium in the gut-wall communicating with the axial sinus. He has also failed to find the Urkeimzellen in the pseudo-heart, and concludes that these, when they occur in the rachis, are due to special formation in situ. The spaces he has found are very possibly mucus-cells cut obliquely; such are greatly developed in the ventral wall of the stomach of Ampbiura squamata. His failure to trace the primitive germ-cells into the ovoid gland is due to want of study of young specimens.

In his later work on general Echinoderm morphology (6) he finds one large axial sinus, into which stone canal and pore canal open, confusing thus together, from want of study of good series, ampulla and axial and aboral sinuses. He now accepts Köhler’s account of the haemal system.

As to the development of the ovoid gland in Echinoderms in general, most of the notices I have been able to collect are very scanty, and evidently based on one or two observations only.

Thus in Asterina gibbosa Ludwig (12) finds that it forms as a split in the mesentery containing the stone canal; in Asterias hyadesi Perrier (13) finds that it arises as a conical process of the peritoneal lining of the stomach, which grows down alongside the wall of the axial sinus, and becomes later invaginated into its interior. It contains a number of “corps vitellins,” which afterwards become the corpuscles of the bodycavity.

In Echinoids Prouho (15) describes the rudiment of the gland as an “amas cellule allongé” at the side of the stone canal. Cuénot (6), in Amphiura squamata, finds it arise as a thickening of the wall of the axial sinus when the embryo has attained a diameter of 700 μ! Russo (16), in the same form, as a cumulus of cells when the embryo has attained its pentagonal form ; and Apostolides (1) as a small swelling of the stone canal.

All these statements appear to me to be based on a few casual observations ; but I cannot raise too strong a protest against the superficial investigations on which the last three writers have based a history of this organ. Since Cuénot knows nothing of the origin of the sinuses, it is evident how entirely wrong he must be.

The case is very different with the interesting observation of Perrier (14) and Bury (2) on the dorsal organ of Crinoids. This organ, in spite of Cuénot’s objections, I regard as the homologue of the ovoid gland. Cuénofs first objection is that the ampulla of the primitive stone canal in Crinoids disappears, whereas in Ophiurids it surrounds the gland. As a matter of fact it has precisely similar relations in both groups. His second objection is that in Crinoids the genital rachis comes off from the oral and not the aboral end of the gland, as in Ophiurids. Taking Asterids for a moment, where the madreporite has a dorsal position, we find that the genital rachis comes off from not quite the aboral end, but there is a portion of the gland beyond it, and this may correspond to the main mass of the organ in Crinoids. The same thing is true of Ophiurids, only here the madreporite is brought round to the ventral side.

As, however, to the general homology of the two organs the observations of Bury and Perrier leave us in no doubt whatever. According to the first observer the dorsal organ arises as a cord of elongated cells ; since these lie in the concavity of a prolongation of the right cœlomic sac, which is crescentic in cross-section, they may be regarded as peritoneal. This prolongation becomes divided longitudinally into five cavities, and these form the beginning of the chambered organ. Later the axial organ grows up under the mesentery of the stomach with an axial cord of Urkeimzellen. Bury finds that here, as in Ophiurids (3), the ampulla of the stone canal represents an anterior cœlomic pouch.

Perrier finds that the dorsal organ is a thickening of one of the layers of the right peritoneal sac. When the dome of the calyx opens it is ovoid, and formed of large pyriform cells with a fibrous membrane. According to Perrier, diverticula from secondary madreporic pores (entonnoirs vibratiles) form vessels around the dorsal organ. If we accept Bury’s view, and make allowance for the great divergence of type between Crinoids and Ophiurids, we might regard the sinuses of Ophiurids as being equivalent to the chambered organ of Crinoids, both being diverticula of the cœlom.

II. GENITAL ORGANS

Hamann (8) describes a genital rachis with Urkeimzellen in Asterids, Ophiurids, Echinoids, and Crinoids. He proves that the genital tubes are outgrowths of the rachis in Crinoids, Asterids, and Echinoids, and suspects it in Ophiurids. Cuénot (5) has also asserted this origin for the gonads in Asterids. He says a septum forms which closes the sinus around the genital bud from the aboral sinus, and in this the external opening is formed. Of the development of the gonads in Ophiurids he merely asserts that it is precisely similar; but I can assert that no such septum is formed in them. Cuénot also asserts that the genital rachis is an outgrowth of the pseudo-heart in Asterids; but this is a mere guess, based on the similarity of the cells in both organs. Between a stage where there is neither rachis nor aboral sinus, and one in which both are completely formed, he finds no intermediate phase. His assertion that the ordinary cells of the ovoid gland mature into Urkeimzellen finds no support from my observations, which show that the portion of the gland destined to form the primitive germ-cells is specialised from the very first.

Prouho (15) finds the genital rachis arises as a bud close to but independently of the pseudo-heart. Its further history is similar to that in Ophiurids. I think his statement of the independent origin of the rachis and ovoid gland requires confirmation, Russo (17) has made the strange error of saying that it results from Ludwig’s and Hamann’s researches that the genital organs, like the genital bursae, are of ectodermic origin.

Summary and General Conclusions,

The most important results detailed in this paper are as follows:

  1. The primitive germinal cells are peritoneal.

  2. The ovoid gland is a solid organ, which develops from a portion of the rudiment which gives rise to the primitive germinal cells.

  3. The axial and aboral sinuses are involutions of the coelom, and have no connection with the ampulla of the stone canal or each other.

  4. The genital rachis is an outgrowth from the ovoid gland into the aboral sinus, the formation of sinus and rachis proceeding together.

  5. There is no evidence of the existence of Ludwig’s haemal system in Amphiura squamata.

  6. Both kinds of cells, germinal and interstitial, which are found in the genital rachis, are formed in the ovoid gland.

  7. The germinal cells are formed from peritoneal cells directly; there is no evidence of the transformation of the special cells of the ovoid gland into Urkei mzellen.

The conclusions which I draw are these: that Echinoderms agree with other Cœlomata in the origin of their genital cells. These latter have at first an unsymmetrical position in Echinoderms, and afterwards take on a radially symmetrical disposition in correspondence with the secondarily acquired radial form of the body. The origin of these cells adjacent to the stone canal suggests a comparison of the origin of the genital cells near the nephridia in many Annelids, but the homology of the stone canal with a nephridium has yet to be proved.

In conclusion I have to express my warmest thanks to Dr. H. Eisig and Dr. P. Mayer for many valuable suggestions, especially to the latter for his invaluable assistance with technique.

1.
Apostolides
,
Nicolas Christo-.
—“
Anatomie et Développement des Ophiurides
,”
‘Archives de Zoologie Expérimentale,’
x
,
1882
,
2.
Bury
,
H.
—“
Early Stages in the Development of Antedon Rosacea
,”
‘ Phil. Trans. Royal Soc.,’
London
, vol.
clxxix
.
3.
Bürt
,
H.
—“
Studies in the Embryology of the Echinoderms
,”
‘Quarterly Journal of Microscopical Science,’
April
,
1889
.
4.
Cuénot
,
L.
—“
Études anatomiques et morphologiques sur les Ophiures
,”.
‘Archives de Zoologie Expérimentale,’
tome
vi
,
1888
.
5.
CuÉnot
,
L.
—“
Contribution à l’Étude anatomique des Astérides
,”
‘Archives de Zoologie Expérimentale,’
tome
v
bis, supplementale.
6.
CuÉnot
,
L.
—“
Études morphologiques sur les Échinoderms
,”
‘Archives de Biologie,’
tome
xi
, fascicles 1 and 2.
7.
Hamann
,
Otto von.
—“
Anatomie der Ophiuriden und Crinoiden
,”
‘ Jenaischen Zeitschrift Nat. Wiss.,’
Bd.
xxiii
.
8.
Hamann
,
Otto von
. — “
Die Wandernden Urkeimzellen und ihre Reifungs - statte bei den Echinodermen
,”
‘ Zeitschrift für wiss. Zoologie,’
6 d.,
46
,
1880
.
9.
Köhler
. —“
Recherches sur l’appareil circulatoire des Ophiures
,”
‘Annales des Sciences naturelles—Zoologie,’
1887
, H. 7.
10.
Ludwig
,
H.
—“
Beiträge zur Anatomie der Ophiuriden
,”
‘ Zeitschrift für wiss. Zoologie,’
Bd.
xxxi
,
1878
.
11.
Ludwig
,
H.
—“
Neue Beiträge zur Anatomie der Ophiuriden
,”
‘Zeit schrift für wiss. Zoologie,’
Bd.
xxxiv
,
1880
.
12.
Ludwig
,
H.
—“
Entwickelungsgeschichte von A s t e r i n a g ib b o s a
,”
‘ Zeit schrift für wiss. Zoologie,’
Bd.
xxxvii
,
1882
,
13.
Perrier
,
E.
— “
Recherches sur l’organisation des Étoiles de Mer
,”
‘Comptes Rendus,’
tome
cii
,
1886
, p.
1146
.
14.
Perrier
,
E.
—“
Mémoire sur l’organisation et le développement de la Comatula de la Méditerranean
,”
‘ Archives de Zool. Expérimentale,’
1886
.
15.
ProVho
,
M.
—“
Sur le développement de l’appareil génital des Oursins
,”
‘ Comptes Rendus,’
tome
civ
,
1887
.
16.
Russo
,
A.
—“
Embryologia d’Amphiura squamata
,”
‘Rendiconti delle Società Reale di Napoli,’
tome
v
, 2nd séries.
17.
Russo
,
A.
—“
Recerche citologiche sugli elementi seminale delle Ophiureæ
,”
‘ Internationale Monats. fùr Anatomie und Physiologie,’
voi.
vii
,
1890
.

Illustrating Mr. E. W. MacBride’s paper “On the Development of the Genital Organs, &c., of Amphiura squamata.”

List of Abbreviations employed

C. Ccelom. gen. b. Genital bursa, gen. r. Genital rachis, h. r. Köhler’s “hæmal “ring. m.p. Madreporic pore. msch. Mesenchyme-cells filling the ccelom. muse. bucc. External muscles of mouth, muse. int. luterradial muscle of disc. muse. r. Muscles of the arm. on. g. Ovoid gland, nen. rg. Nerve-ring. p. c. Pore canal, rad. nerv. Radial nerve, sinus a. Aboral sinus, sinus b, Axial sinus, sinus c. Ampulla of the stone canal, sinus epin. Epineural sinus, sinus rad. perih. Radial perihæmal sinus, sinus perist. Peristomal sinus, sk. Shrinkage space, st. c. stone canal, w. v. r. Water vascular ring.

Figs. 2 E and 6 D are diagrams. The outlines of all the rest have been drawn with the camera lucida, and except where otherwise stated are magnified 340 diameters, and taken from sections parallel to the axis of the stone canal of specimens preserved in osmic acid followed by Mü ller’s fluid.

FIG. 1.—From an individual ‘86 mm. in diameter. Magnified 130 diameters. Osmic alone, sinus rad. perih. Radial perihæmal sinus, sinus ep. n. Epineural sinus, sinus perist. Circumoral sinus.

FIG. 2 A.—Section of ovoid gland and surrounding structures of an adultgen. r. Portion of genital rachis where it passes into ovoid gland, sk. Shrink age space due to action of preserving fluids on great interradial muscle, h. r. Pedicle of ovoid gland (= Kohler’s hæmal ring ?).

FIG. 2B.—Another section of ovoid gland from the same series, st. c. Proximal parts of stone canal.

FIG. 2 C.—Horizontal section of ovoid gland of another adult, sinus b. Transverse section of the canal sinus.

FIG. 2D.—Nuclei suggesting direct division from another adult organ.

FIG. 2E.—Diagram of relation of gland and surrounding sinuses in the adult, m. p. Madreporic pore.

FIG. 3.—From an individual about -2 mm. in diameter, msch. Mesenchyme filling ventral portion of the cœlom.

FIGS. 4 A, B, C.—Three consecutive sections from a somewhat older individual than Fig. 3. m. p. Madreporic pore. ov. g. First rudiment of the ovoid gland.

FIGS. 5 A, B.—Two sections from a series through an individual about ·25 mm. in diameter, gen. r. First appearance of primitive germ-cells, sinus b. A nook of the cœlom, afterwards becoming the axial sinus.

FIGS. 6A, B, C.—Three sections from a series through an individual -3 mm. in diameter, sinus b. Axial sinus still communicating with cœlom though its lumen is obliterated, sinus a. A nook of the cœlom, afterwards becoming aboral sinus.

FIG. 6D.—Diagram of relations of the rudiments of axial and aboral sinuses to each other and to the incipient ovoid gland.

FIGS. 7 A, B, C.—Three sections from a series through an individual ‘48 mm. in diameter, gen. r. Solid knob in which incipient aboral sinus and genital rachis end. h. r. Prolongation of the ovoid gland over the dorsal surface of the nerve-ring.

FIG. 8.—Section of an individual ‘83 mm. in diameter. Osmic alone. sinus a, sinus b. Sharply separated axial and aboral sinuses.

FIGS. 9A, B.—Two sections from a series through an individual ‘58 mm. in diameter, gen. r. Genital rachis constricted off from ovoid gland.

FIG. 10.—Oblique section of an individual ‘59 mm. in diameter.

FIG. 11.—Section of an individual ‘9 mm. in diameter, x. Curious branched cell in the axial sinus.

FIG. 12.—Another section from same series as Fig. 11, showing growing branch of rachis, t. Position in which testis will be formed.

FIG. 13.—Young testis from an individual about -97 mm. in diameter.

FIG. 14.—More advanced stage of testis. From an individual ‘9 mm. in diameter.

FIG. 15.—Adult testis.

FIGS. 16 A, B.—Two sections from same series as Fig. 1, but magniGed 340 diameters. 16A is through the genital rachis; 16B through the incipient ovary.

FIG. 17.—Shows young ovary passing into the genital rachis, from same series as Fig. 14.

FIG. 18.—Adult ovary.

FIG. 19.—First rudiment of genital bursa. From an individual -62 mm. in diameter. Osmic alone.

FIG. 20.—More advanced stage of the bursa, w. v. b. B.adial water vessel.

FIG. 21.—Section of radial nerve-cord of young adult, w. v. b. Radial water vessel.

FIG. 22.—Section of radial nerve-cord of another adult, w. v. b. Radial water vessel.

FIG. 23.—Section of radial nerve-cord of adult treated with osmic alone. w. v. b. Radial water vessel.

1

Since writing the above I have seen the Ampulla not only in embryos of this age, but also in bilateral larvæ.