Chapter XVIII: Crustacea185 (3)

The general arguments in favour of a Zoæa ancestor with partially or completely aborted thoracic appendages having actually existed in the past appear to me very powerful. In all the Malacostracan groups in which the larva leaves the egg in an imperfect form a true Zoæa stage is found. That the forms of these Zoææ should differ considerably is only what might be expected, considering that they lead a free existence and are liable to be acted upon by natural selection, and it is probable that none of those at present existing closely resemble the ancestral form. The spines from their carapace, which vary so much, were probably originally developed, as suggested by Fritz Müller, as a means of defence. The simplicity of the heart--so different from that of Phyllopods--in most forms of Zoæa is a difficulty, but the reduction in the length of the heart may very probably be a secondary modification; the primitive condition being retained in the Squilla Zoæa. In any case this difficulty is not greater on the hypothesis of the Zoæa being an ancestral form, than on that of its being a purely larval one.

The points of agreement in the number and character of the appendages, form of the abdomen, etc. between the various types of Zoæa appear to me too striking to be explained in the manner attempted by Claus. It seems improbable that a peculiarity of form acquired by the larva of some ancestral Malacostracan should have been retained so permanently in so many groups[207]--more permanently indeed than undoubtedly ancestral forms like that of Mysis--and it would be still more remarkable that a Zoæa form should have been two or more times independently developed.

[207] A secondary larval form is less likely to be repeated in
development than an ancestral adult stage, because there is
always a strong tendency for the former, which is a secondarily
intercalated link in the chain, to drop out by the occurrence of
a reversion to the original type of development.

There are perhaps not sufficient materials to reconstruct the characters of the Zoæa ancestor, but it probably was provided with the anterior appendages up to the second pair of maxillipeds, and (?) with abdominal swimming feet. The heart may very likely have been many-chambered. Whether gill pouches were present on the maxillipeds and abdominal feet does not appear to me capable of being decided. The carapace and general shape were probably the same as in existing Zoæas. It must be left an open question whether the six hindermost thoracic appendages were absent or only very much reduced in size.

On the whole then it may be regarded as probable that the Malacostraca are descended from Protophyllopod forms, in which, on the adoption of swimming habits, six appendages of the middle region of the body were reduced or aborted, and a Zoæa form acquired, and that subsequently the lost appendages were redeveloped in the descendants of these forms, and have finally become the most typical appendages of the group.

The relationship of the various Malacostracan groups is too difficult a subject to be discussed here, but it seems to me most likely that in addition to the groups with a Zoæa stage the Edriophthalmata and Cumaceæ are also post-zoæal forms which have lost the Zoæa stage. Nebalia is however very probably to be regarded as a præ-zoæal form which has survived to the present day; and one might easily fancy that its eight thin thoracic segments with their small Phyllopod-like feet might become nearly aborted.

Copepoda. The Copepoda certainly appear to have diverged very early from the main stem, as is shewn by their simple biramous feet and the retention of the median eye as the sole organ of vision. It may be argued that they have lost the eye by retrogressive changes, and in favour of this view cases of the Pontellidæ and of Argulus may be cited. It is however more than doubtful whether the lateral eyes of the Pontellidæ are related to the compound Phyllopod eye, and the affinities of Argulus are still uncertain. It would moreover be a great paradox if in a large group of Crustacea the lateral eyes had been retained in a parasitic form only (Argulus), but lost in all the free forms.

Cirripedia. The Cirripedia are believed by Claus to belong to the same phylum as the Copepoda. This view does not appear to be completely borne out by their larval history. The Nauplius differs very markedly from that of the Copepoda, and this is still more true of the Cypris stage. The Copepod-like appendages of this stage are chiefly relied upon to support the above view, but this form of appendages was probably very primitive and general, and the number (without taking into consideration the doubtful case of Cryptophialus) does not correspond to that in Copepoda. On the other hand the paired eyes and the bivalve shell form great difficulties in the way of Claus' view. It is clear that the Cypris stage represents more or less closely an ancestral form of the Cirripedia, and that both the large bivalve shell and the compound eyes were ancestral characters. These characters would seem incompatible with Copepod affinities, but point to the independent derivation of the Cirripedia from some early bivalve Phyllopod form.

[FIG. 237. FIGURES ILLUSTRATING THE DEVELOPMENT OF ASTACUS. (From
Parker; after Reichenbach.)

A. Section through part of the ovum during segmentation. _n._
nuclei; _w.y._ white yolk; _y.p._ yolk pyramids; _c._ central
yolk mass.
B and C. Longitudinal sections during the gastrula stage. _a._
archenteron; _b._ blastopore; _ms._ mesoblast; _ec._ epiblast;
_en._ hypoblast distinguished from epiblast by shading.
D. Highly magnified view of the anterior lip of blastopore to shew
the origin of the primary mesoblast from the wall of the
archenteron. _p.ms._ primary mesoblast; _ec._ epiblast; _en._
hypoblast.
E. Two hypoblast cells to shew the amoeba-like absorption of yolk
spheres. _y._ yolk; _n._ nucleus; _p._ pseudopodial process.
F. Hypoblast cells giving rise endogenously to the secondary
mesoblast (_s.ms._). _n._ nuclei.]

Ostracoda. The independent origin of the Ostracoda from the main Crustacean stem seems probable. Claus points out that the Ostracoda present by no means a simple organisation, and concludes that they were not descended from a form with a more complex organisation and a larger number of appendages. Some simplifications have however undoubtedly taken place, as the loss of the heart, and of the compound eyes in many forms. These simplifications are probably to be explained (as is done by Claus) as adaptations due to the small size of body and its enclosure in a thick bivalve shell. Although Claus is strongly opposed to the view that the number of the appendages has been reduced, yet the very fact of the (in some respects) complex organisation of this group might seem to indicate that it cannot have diverged from the Phyllopod stem at so early a stage as (on Claus' view of the Nauplius) would seem to be implied by the very small number of appendages which is characteristic of it, and it therefore appears most probable that the present number may be smaller than that of the ancestral forms.

_The formation of the germinal layers._

The formation of the germinal layers has been more fully studied in various Malacostraca, more especially in the Decapoda, than in other groups.

Decapoda. To Bobretzky (No. 472) is due the credit of having been the pioneer in this line of investigation; and his researches have been followed up and enlarged by Haeckel, Reichenbach (No. 488), and Mayer (No. 482). The segmentation is centrolecithal and regular (fig. 237 A). At its close the blastoderm is formed of a single uniform layer of lens-shaped cells enclosing a central sphere of yolk, in which as a rule all trace of the division into columns, present during the earlier stages of segmentation, has disappeared; though in Palæmon the columns remain for a long period distinct. The cells of the blastoderm are at first uniform, but in Astacus, Eupagurus, and most Decapoda, soon become more columnar for a small area, and form a circular patch. The whole patch either becomes at once invaginated (Eupagurus, Palæmon, fig. 239 A) or else the edge of it is invaginated as a roughly speaking circular groove deeper anteriorly than posteriorly, within which the remainder of the patch forms a kind of central plug, which does not become invaginated till a somewhat later period (Astacus, fig. 237 B and C). After the invagination of the above patch the remainder of the blastoderm cells form the epiblast.

The invaginated sack appears to be the archenteron and its mouth the blastopore. The mouth finally becomes closed[208], and the sack itself then forms the mesenteron.

[208] Bobretzky first stated that the invagination remained open,
but subsequently corrected himself. _Zeit. f. Wiss. Zool._, Bd.
XXIV. p. 186.

[FIG. 238. TWO LONGITUDINAL SECTIONS OF THE EMBRYO OF ASTACUS. (From
Parker; after Bobretzky.)

A. Nauplius stage. B. Stage after the hypoblast cells have absorbed
the food-yolk. The ventral surface is turned upwards. _fg._
stomodæum; _hg._ proctodæum; _an._ anus; _m._ mouth; _mg._
mesenteron; _abd._ abdomen; _h._ heart.]

In Astacus the archenteron gradually grows forwards, its opening is at first wide, but becomes continuously narrowed and is finally obliterated. Very shortly after this occurrence there is formed, slightly in front of the point where the last trace of the blastopore was observable, a fresh epiblastic invagination, which gives rise to the proctodæum, and the opening of which remains as the definite anus. The proctodæum (fig. 238 A, _hg)_ is very soon placed in communication with the mesenteron (_mg_). The stomodæum (_fg_) is formed during the same stage as the proctodæum. It gives rise to the oesophagus and stomach. The hypoblast cells which form the wall of the archenteron grow with remarkable rapidity at the expense of the yolk; the spherules of which they absorb and digest in an amoeba-like fashion by means of their pseudopodia. They become longer and longer, and finally, after absorbing the whole yolk, acquire a form almost exactly similar to that of the yolk pyramids during segmentation (fig. 238 B). They enclose the cavity of the mesenteron, and their nuclei and protoplasm are situated externally. The cells of the mesenteron close to its junction with the proctodæum differ from those elsewhere in being nearly flat.

In Palæmon (Bobretzky) the primitive invagination (fig. 239 A) has far smaller dimensions than in Astacus, and appears before the blastoderm cells have separated from the yolk pyramids. The cells which are situated at the bottom of it pass into the yolk, increase in number, and absorb the whole yolk, forming a solid mass of hypoblast in which the outlines of the individual cells would seem at first not to be distinct. The blastopore in the meantime becomes closed. Some of the nuclei now pass to the periphery of the yolk mass; the cells appertaining to them gradually become distinct and assume a pyramidal form (fig. 239 B, _hy_), the inner ends of the cells losing themselves in a central mass of yolk, in the interior of which nuclei are at first present but soon disappear. The mesenteron thus becomes constituted of a layer of pyramidal cells which merge into a central mass of yolk. Some of the hypoblast cells adjoining the junction of the proctodæum and mesenteron become flattened, and in the neighbourhood of these cells a lumen first appears. The stomodæum and proctodæum are formed as in Astacus. Fig. 239 B shews the relative positions of the proctodæum, stomodæum, and mesenteron. Although the process of formation of the hypoblast and mesenteron is essentially the same in Astacus and Palæmon, yet the differences between these two forms are very interesting, in that the yolk is _external_ to the mesenteron in Astacus, but _enclosed within it_ in Palæmon. This difference in the position of the yolk is rendered possible by the fact that the invaginated hypoblast cells in Palæmon do not, at first, form a continuous layer enclosing a central cavity, while they do so in Astacus.

[FIG. 239. TWO STAGES IN THE DEVELOPMENT OF PALÆMON SEEN IN SECTION.
(After Bobretzky.)

A. Gastrula stage.
B. Longitudinal section through a late stage. _hy._ hypoblast;
_sg._ supra-oesophageal ganglion; _vg._ ventral nerve
cord; hd. proctodæum; _st._ stomodæum.]

The mesoblast appears to be formed of cells budded off from the anterior wall of the archenteron (Astacus, fig. 237 D), or from its lateral walls generally (Palæmon). They make their first appearance soon after the invagination of the hypoblast has commenced. The mesoblast cells are at first spherical, and gradually spread, especially in an anterior direction, from their point of origin.

According to Reichenbach there are formed in Astacus at the Nauplius stage a number of peculiar cells which he speaks of as 'secondary mesoblast cells.' His account is not very clear or satisfactory, but it appears that they originate (fig. 237 F) in the hypoblast cells by a kind of endogenous growth, and though they have at first certain peculiar characters they soon become indistinguishable from the remaining mesoblast cells.

Towards the end of the Nauplius period the secondary mesoblast cells aggregate themselves into a rod close to the epiblast in the median ventral line, and even bifurcate round the mouth and extend forwards to the extremity of the procephalic lobes. This rod of cells very soon vanishes, and the secondary mesoblast cells become indistinguishable from the primary. Reichenbach believes, on not very clear evidence, that these cells have to do with the formation of the blood.

_General form of the body._ The ventral thickening of epiblast or ventral plate, continuous with the invaginated patch already mentioned, forms the first indication of the embryo. It is at first oval, but soon becomes elongated and extended anteriorly into two lateral lobes--the procephalic lobes. Its bilateral symmetry is further indicated by a median longitudinal furrow. The posterior end of the ventral plate next becomes raised into a distinct lobe--the abdomen--which in Astacus at first lies _in front_ of the still open blastopore. This lobe rapidly grows in size, and at its extremity is placed the narrow anal opening. It soon forms a well-marked abdomen bent forwards over the region in front (figs. 239 B, and 240 A and B). Its early development as a distinct outgrowth causes it to be without yolk; and so to contrast very forcibly with the anterior thoracic and cephalic regions of the body. In most cases this process corresponds to the future abdomen, but in some cases (Loricata) it appears to include part of the thorax. Before it has reached a considerable development, three pairs of appendages spring from the region of the head, viz. two pairs of antennæ and the mandibles, and inaugurate a so-called Nauplius stage (fig. 240 A). These three appendages are formed nearly simultaneously, but the hindermost appears to become visible slightly before the two others (Bobretzky). The mouth lies slightly behind the anterior pair of antennæ, but distinctly in front of the posterior pair. The other appendages, the number of which at the time of hatching varies greatly in the different Decapods (_vide_ section on larval development), sprout in succession from before backwards (fig. 240 B). The food-yolk in the head and thoracic region gradually becomes reduced in quantity with the growth of the embryo, and by the time of hatching the disparity in size between the thorax and abdomen has ceased to exist.

Isopoda. The early embryonic phases of the Isopoda have been studied by means of sections by Bobretzky (No. 498) and Bullar (No. 499) and have been found to present considerable variations. When laid the egg is enclosed in a chorion, but shortly after the commencement of segmentation (Ed. van Beneden and Bullar) a second membrane appears, which is probably of the nature of a larval membrane.

[FIG. 240. TWO STAGES IN THE DEVELOPMENT OF PALÆMON.

A. Nauplius stage.
B. Stage with eight pairs of appendages. _op._ eyes; _at1._ and
_at2._ first and second antennæ; _md._ mandibles; _mx1_, _mx2._
first and second maxillæ; _mxp3._ third maxillipeds; _lb._ upper
lip.]

In all the forms the segmentation is followed by the formation of a blastoderm, completely enclosing the yolk, and thickened along an area which will become the ventral surface of the embryo. In this area the blastoderm is formed of at least two layers of cells--an external columnar epiblast, and an internal layer of scattered cells which form the mesoblast and probably in part also the hypoblast (Oniscus, _Bobretzky_; Cymothoa, _Bullar_).

In Asellus aquaticus there is a centrolecithal segmentation, ending in the formation of a blastoderm, which appears first on the ventral surface and subsequently extends to the dorsal.

In Oniscus murarius, and Cymothoa the segmentation is partial [for its peculiarities and relationship _vide_ p. 120] and a disc, formed of a single layer of cells, appears at a pole of the egg which corresponds to the future ventral surface (Bobretzky). This layer gradually grows round the yolk partly by division of its cells, though a formation of fresh cells from the yolk may also take place. Before it has extended far round the yolk, the central part of it becomes two or more layers deep, and the cells of the deeper layers rapidly increase in number, and are destined to give rise to the mesoblast and probably also to part or the whole of the hypoblast. In Cymothoa this layer does not at first undergo any important change, but in Oniscus it becomes very thick, and its innermost cells (Bobretzky) become imbedded in the yolk, which they rapidly absorb; and increasing in number first of all form a layer in the periphery of the yolk, and finally fill up the whole of the interior of the yolk (fig. 241 A), absorbing it in the process.

It appears possible that these cells do not, as Bobretzky believes, originate from the blastoderm, but from nuclei in the yolk which have escaped his observation. This mode of origin would be similar to that by which yolk cells originate in the eggs of the Insecta, etc. If Bobretzky's account is correct we must look to Palæmon, as he himself suggests, to find an explanation of the passage of the hypoblast cells into the yolk. The thickening of the primitive germinal disc would, according to this view, be equivalent to the invagination of the archenteron in Astacus, Palæmon, etc.

[FIG. 241. TWO LONGITUDINAL SECTIONS THROUGH THE EMBRYO OF ONISCUS
MURARIUS. (After Bobretzky.)

_st._ stomodæum; _pr._ proctodæum; _hy._ hypoblast formed of large
nucleated cells imbedded in the yolk; _m._ mesoblast; _vg._ ventral
nerve cord; _sg._ supra-oesophageal ganglion; _li._ liver; _do._
dorsal organ; _zp._ rudiment of masticatory apparatus; _ol._ upper
lip.]

Whatever may be the origin of the cells in the yolk they no doubt correspond to the hypoblast of other types. In Cymothoa nothing similar to them has been met with, but the hypoblast has a somewhat different origin being apparently formed from some of the indifferent cells below the epiblast, which collect as a solid mass on the ventral surface, and then divide into two masses which become hollow and give rise to the liver cæca. Their fate, as well as that of the hypoblast in Oniscus, is dealt with in connection with the alimentary tract. The completion of the enclosure of the yolk by the blastoderm takes place on the dorsal surface. In all the Isopods which have been carefully studied, there appears before any other organ a provisional structure formed from the epiblast and known as the dorsal organ. An account of it is given in connection with the development of the organs. The general external changes undergone by the larva in its development are as follows. The ventral thickened area of the blastoderm (ventral plate) shapes itself and girths nearly the whole circumference of the ovum in Oniscus (fig. 241 A) but is relatively much shorter in Cymothoa. Anteriorly it dilates into the two procephalic lobes. In Cymothoa it next becomes segmented; and the anterior segments are formed nearly simultaneously, and those of the abdomen somewhat later. At the same time a median depression appears dividing the blastoderm longitudinally into two halves. The appendages are formed later than their segments, and the whole of them are formed nearly simultaneously, with the exception of the last thoracic, which does not appear till comparatively late after the hatching of the embryo. The late development of the seventh thoracic segment and appendage is a feature common to the majority of the Isopoda (Fritz Müller). In Oniscus the limbs are formed in nearly the same way as in Cymothoa, but in Asellus they do not arise quite simultaneously. First of all, the two antennæ and mandibles (the future palp) appear, inaugurating a stage often spoken of as the Nauplius stage, which is supposed to correspond with the free Nauplius stage of Penæus and Euphausia. At this stage a cuticle is shed (Van Beneden) which remains as an envelope surrounding the larva till the time of hatching. Similar cuticular envelopes are formed in many Isopoda. Subsequently the appendages of the thorax appear, and finally those of the abdomen. Later than the appendages there arise behind the mouth two prominences which resemble appendages, but give rise to a bilobed lower lip (Dohrn).

In Asellus and Oniscus the ventral plate moulds itself to the shape of the egg, and covers the greater part of the dorsal as well as of the ventral side (fig. 241 A). As a result of this the ventral surface of the embryo is throughout convex; and in Asellus a deep fold appears on the back of the embryo, so that the embryo appears coiled up within the egg with its ventral side outwards and its head and tail in contact. In Oniscus the ventral surface is convex, but the dorsal surface is never bent in as in Asellus. In Cymothoa the egg is very big and the ventral plate does not extend nearly so far round to the dorsal side as in Asellus, in consequence of which the ventral surface is not nearly so convex as in other Isopoda. At the same time the telson is early formed, and is bent forwards so as to lie on the under side of the part of the blastoderm in front. In having this ventral curvature of the telson Cymothoa forms an exception amongst Isopods; and in this respect is intermediate between the embryos of Asellus and those of the Amphipoda.

Amphipoda. Amongst the Amphipoda the segmentation is usually centrolecithal. In the case of Gammarus locusta (Ed. van Beneden and Bessels, No. 503) it commences with an unequal but total segmentation like that of the Frog (_vide_ p. 97), and the separation of a central yolk mass is a late occurrence; and it is noticeable that the part of the egg with the small segments eventually becomes the ventral surface. In the fresh-water species of Gammarus (G. pulex and fluviatilis) the segmentation is more like that of Insects; the blastoderm cells being formed nearly simultaneously over a large part of the surface of the egg.

Both forms of segmentation give rise to a blastoderm covering the whole egg, which soon becomes thickened on the ventral surface. There is formed, as in the Isopoda, a larval membrane at about the time when the blastoderm is completed. Very soon after this the egg loses its spherical shape, and becomes produced into a pointed extremity--the future abdomen--which is immediately bent over the ventral surface of the part in front. The ventral curvature of the hinder part of the embryo at so early an age stands in marked contrast to the usual condition of Isopod embryos, and is only approached in this group, so far as is known, in the case of Cymothoa.

At the formation of the first larval membrane the blastoderm cells separate themselves from it, except at one part on the dorsal surface. The patch of cells adherent at this part gives rise to a dorsal organ, comparable with that in Oniscus, connecting the embryo and its first larval skin. A perforation appears in it at a later period.

The segments and limbs of the Amphipoda are all formed before the larva leaves the egg.

Cladocera. The segmentation (Grobben, No. 455) takes place on the normal centrolecithal type, but is somewhat unequal. Before the close of the segmentation there may be seen at the apex of the vegetative pole one cell marked off from the remainder by its granular aspect. It gives rise to the generative organs. One of the cells adjoining it gives rise to the hypoblast, and the other cells which surround it form the commencement of the mesoblast. The remaining cells of the ovum form the epiblast. By a later stage the hypoblast cell is divided into thirty-two cells and the genital cell into four, while the mesoblast forms a circle of twelve cells round the genital mass.

The hypoblast soon becomes involuted; the blastopore probably closes, and the hypoblast forms a solid cord of cells which eventually becomes the mesenteron. The stomodæum is said to be formed at the point of closure of the blastopore. The mesoblast passes inwards and forms a mass adjoining the hypoblast, and somewhat later the genital mass also becomes covered by the epiblast. The proctodæum appears to be formed later than the stomodæum.

The embryo as first shewn by Dohrn passes through a Nauplius stage in the brood-pouch, but is hatched, except in the case of the winter eggs of Leptodora, in a form closely resembling the adult.

Copepoda. Amongst the free Copepoda the segmentation and formation of the layers have recently been investigated by Hoek (No. 512). He finds that there is, in both the fresh-water and marine forms studied by him, a centrolecithal segmentation similar to that of Palæmon and Pagurus (_vide_ p. 112), which might from the surface be supposed to be complete and nearly regular. After the formation of the blastoderm an invagination of some of its cells takes place and is completed in about a quarter of an hour. The opening becomes closed. This invagination is compared by Hoek to the invagination in Astacus, and is believed by him to give rise to the mesenteron. Its point of closing corresponds with the hind end of the embryo. On the ventral surface there appear two transverse furrows dividing the embryo into three segments, and a median longitudinal furrow which does not extend to the front end of the foremost segment. The three pairs of Nauplius appendages and upper lip become subsequently formed as outgrowths from the sides of the ventral blastodermic thickening.

Amongst the parasitic Copepoda there are found two distinct types of segmentation, analogous to those in the Isopoda. In the case of Condracanthus the segmentation is somewhat irregular, but on the type of Eupagurus, etc. (_vide_ p. 112). In the other group (Anchorella, Clavella, Congericola, Caligus, Lerneopoda) the segmentation nearly resembles the ordinary meroblastic type (_vide_ p. 120), and is to be explained in the same manner as in the cases of Oniscus and Cymothoa. The first blastodermic cells sometimes appear in a position corresponding with the head end of the embryo (Anchorella), at other times at the hind end (Clavella), and sometimes in the middle of the ventral surface. The dorsal surface of the yolk is always the latest to be inclosed by the blastoderm cells. A larval cuticle similar to that of the Isopoda is formed at the same time as the blastoderm. At the sides of the ventral thickening of the blastoderm there grow out the Nauplius appendages, of which only the first two appear in Anchorella. In Anchorella and Lerneopoda the embryos are not hatched at the Nauplius stage, but after the Nauplius appendages have been formed a fresh cuticle--the Nauplius cuticle--is shed, and within it the embryo develops till it reaches the so-called Cyclops stage (_vide_ p. 490). The embryo within the egg has its abdomen curved dorsalwards as amongst the Isopoda.

Cirripedia. The segmentation of Balanus and Lepas commences by the segregation of the constituents of the egg into a more protoplasmic portion, and a portion formed mainly of food material. The former separates from the latter as a distinct segment, and then divides into two not quite equal portions. The division of the protoplasmic part of the embryo continues, and the resulting segments grow round the single yolk segment. The point where they finally enclose it is situated on the ventral surface (Lang) at about the position of the mouth (?).

After being enclosed by the protoplasmic cells the yolk divides, and gives rise to a number of cells, which probably supply the material for the walls of the mesenteron. The external layer of protoplasm forms the so-called blastoderm, and soon (Arnold, Lang) becomes thickened on the dorsal surface.

The embryo is next divided by two constrictions into three segments; and there are formed the three appendages corresponding to these, which are at first simple. The two posterior soon become biramous. The larva leaves the egg before any further appendages become formed.

_Comparative development of the organs._

Central nervous system. The ventral nerve cord of the Crustacea develops as a thickening of the epiblast along the median ventral line; the differentiation of which commences in front, and thence extends backwards. The ventral cord is at first unsegmented. The supra-oesophageal ganglia originate as thickenings of the epiblast of the procephalic lobes.

The details of the above processes are still in most cases very imperfectly known. The fullest account we have is that of Reichenbach (No. 488) for Astacus. He finds that the supra-oesophageal ganglia and ventral cord arise as a continuous formation, and not independently as would seem to be the case in Chætopoda. The supra-oesophageal ganglia are formed from the procephalic lobes. The first trace of them is visible in the form of a pair of pits, one on each side of the middle line. These pits become in the Nauplius stage very deep, and their walls are then continued into two ridges where the epiblast is several cells deep, which pass backwards one on each side of the mouth. The walls of the pits are believed by Reichenbach to give rise to the optic portions of the supra-oesophageal ganglia, and the epiblastic ridges to the remainder of the ganglia and to the circum-oesophageal commissures. At a much later stage, when the ambulatory feet have become formed, a median involution of epiblast in front of the mouth and between the two epiblast ridges gives rise to a central part of the supra-oesophageal ganglia. Five elements are thus believed by Reichenbach to be concerned in the formation of these ganglia, viz. two epiblast pits, two epiblast ridges, and an involution of epiblast between the latter. It should be noted however that the fate neither of the pair of pits, nor of the median involution, appears to have been satisfactorily worked out. The two epiblast ridges, which pass back from the supra-oesophageal ganglia on each side of the mouth, are continued as a pair of thickenings of the epiblast along the sides of a median ventral groove. This groove is deep in front and shallows out posteriorly. The thickenings on the sides of this groove no doubt give rise to the lateral halves of the ventral cord, and the cells of the groove itself are believed by Reichenbach, but it appears to me without sufficient evidence, to become invaginated also and to assist in forming the ventral cord. When the ventral cord becomes separated from the epiblast the two halves of it are united in the middle line, but it is markedly bilobed in section.

In the Isopoda it would appear both from Bobretzky's and Bullar's observations that the ventral nerve cord arises as an unpaired thickening of the epiblast _in which there is no trace of anything like a median involution_. After this thickening has become separated from the epiblast a slight median furrow indicates its constitution out of two lateral cords. The supra-oesophageal ganglia are stated to be developed quite simply as a pair of thickenings of the procephalic lobes, but whether they are from the first continuous with the ventral cord does not appear to have been determined.

The later stages in the differentiation of the ventral cord are, so far as is known, very similar throughout the Crustacea. The ventral cord is, as has been stated, at first unsegmented (fig. 241 A, _vg_), but soon becomes divided by a series of constrictions into as many ganglia as there are pairs of appendages or segments (fig. 241 B, _vg_).

There appears either on the ventral side (Oniscus) or in the centre (Astacus, Palæmon) of the two halves of each segment or ganglion a space filled with finely punctuated material, which is the commencement of the commissural portion of the cords. The commissural tissue soon becomes continuous through the length of the ventral cord, and is also prolonged into the supra-oesophageal ganglia.

After the formation of the commissural tissue the remaining cells of the cord form the true ganglion cells. A gradual separation of the ganglia next takes place, and the cells become confined to the ganglia, which are finally only connected by a double band of commissural tissue. The commissural tissue not only gives rise to the longitudinal cords connecting the successive ganglia, but also to the transverse commissures which unite the two halves of the individual ganglia.

The ganglia usually, if not always, appear at first to correspond in number with the segments, and the smaller number so often present in the adult is due to the coalescence of originally distinct ganglia.

Organs of special sense. Comparatively little is known on this head. The compound eyes are developed from the coalescence of two structures, both however epiblastic, viz. (1) part of the superficial epiblast of the procephalic lobes; (2) part of the supra-oesophageal ganglia. The former gives rise to the corneal lenses, the crystalline cones, and the pigment surrounding them; the latter to the rhabdoms and the cells which encircle them. Between these two parts a mesoblastic pigment is interposed.

Of the development of the auditory and olfactory organs almost nothing is known.

Dorsal organ. In a considerable number of the Malacostraca and Branchiopoda a peculiar organ is developed from the epiblast in the anterior dorsal region. This organ has been called the dorsal organ. It appears to be of a glandular nature, and is usually very large in the embryo or larva and disappears in the adult; but in some Branchiopoda it persists through life. In most cases it is unpaired, but in some instances a paired organ appears to take its place.

Various views as to its nature have been put forward. There is but little doubt of its being glandular, and it is possible that it is a provisional renal organ, though so far as I know concretions have not yet been found in it.

Its development has been most fully studied in the Isopoda.

[FIG. 242. DIAGRAMMATIC SECTION OF CYMOTHOA SHEWING THE DORSAL
ORGAN. (From Bullar.)]

In Cymothoa (Bullar, No. 499) there appears on the dorsal surface, in the region which afterwards becomes the first thoracic segment, an unpaired linear thickening of the blastoderm. This soon becomes a circular patch, the central part of which is invaginated so as to communicate with the exterior by a narrow opening only (fig. 242). It becomes at the same time attached to the inner egg membrane. It retains this condition till the close of larval life.

In Oniscus (Dohrn, No. 500; Bobretzky, No. 498) there appears very early a dorsal patch of thickened cells. These cells become attached at their edge to the inner egg membrane and gradually separated from the embryo, with which they finally only remain in connection by a hollow column of cells (fig. 241 A, _do_). The original patch now gradually spreads over the inner egg membrane, and forms a transverse saddle-shaped band of flattened cells which engirths the embryo on all but the ventral side.

In the Amphipods the epiblast cells remain attached for a small area on the dorsal surface to the first larval skin, when this is formed. This patch of cells, often spoken of as a micropyle apparatus, forms a dorsal organ equivalent to that in Oniscus. A perforation is formed in it at a later period. A perhaps homologous structure is found in the embryos of Euphausia, Cuma, etc.

[FIG. 243. DIAGRAMMATIC SECTION OF AN EMBRYO OF ASELLUS AQUATICUS TO
SHEW THE PAIRED DORSAL ORGAN. (From Bullar; after E. van Beneden.)]

In many Branchiopoda a dorsal organ is found. Its development has been studied by Grobben in Moina. It persists in the adult in Branchipus, Limnadia, Estherea, etc.

In the Copepoda a dorsal organ is sometimes found in the embryo; Grobben at any rate believes that he has detected an organ of this nature in the embryo of Cyclops serrulatus.

A paired organ which appears to be of the same nature has been found in Asellus and Mysis.

In Asellus (Rathke (No. 501), Dohrn (No. 500), Van Beneden (No. 497)) this organ originates as two cellular masses at the sides of the body just behind the region of the procephalic lobes. Each of them becomes trifoliate and bends towards the ventral surface. In each of their lobes a cavity arises and finally the three cavities unite, forming a trilobed cavity open to the yolk. This organ eventually becomes so large that it breaks through the egg membranes and projects at the sides of the embryo (fig. 243). Though formed before the appendages it does not attain its full development till considerably after the latter have become well established.

In Mysis it appears during the Nauplius stage as a pair of cavities lined by columnar cells, which atrophy very early.

Various attempts have been made to identify organs in other Arthropod embryos with the dorsal organ of the Crustacea, but the only organ at all similar which has so far been described is one found in the embryo of Linguatula (_vide_ Chapter XIX.), but there is no reason to think that this organ is really homologous with the dorsal organ of the Crustacea.

The mesoblast. The mesoblast in the types so far investigated arises from the same cells as the hypoblast, and appears as a somewhat irregular layer between the epiblast and the hypoblast. It gives rise to the same parts as in other forms, but it is remarkable that it does not, in most Decapods and Isopods (and so far we do not know about other forms), become divided into somites, at any rate with the same distinctness that is usual in Annelids and Arthropods. Not only so, but there is at first no marked division into a somatic and splanchnic layer with an intervening body cavity. Some of the cells become differentiated into the muscles of the body wall and limbs; and other cells, usually in the form of a very thin layer, into the muscles of the alimentary tract. In the tail of _Palæmon_ Bobretzky noticed that the cells about to form the muscles of the body were imperfectly divided into cubical masses corresponding with the segments; which however, in the absence of a central cavity, differed from typical mesoblastic somites. In _Mysis_ Metschnikoff states that the mesoblast becomes broken up into distinct somites. Further investigations on this subject are required. The body cavity has the form of irregular blood sinuses amongst the internal organs.

Heart. The origin and development of the heart and vascular system are but very imperfectly known.

In Phyllopods (Branchipus) Claus (No. 454) has shewn that the heart is formed by the coalescence of the lateral parts of the mesoblast of the ventral plates. The chambers are formed successively as the segments to which they belong are established, and the anterior chambers are in full activity while the posterior are not yet formed.

In Astacus and Palæmon, Bobretzky finds that at the stage before the heart definitely appears there may be seen a solid mass of mesoblast cells in the position which it eventually occupies[209]; and considers it probable that the heart originates from this mass. At the time when the heart can first be made out and before it has begun to beat, it has the form of an oval sack with delicate walls separated from the mesenteron by a layer of splanchnic mesoblast. Its cavity is filled with a peculiar plasma which also fills up the various cavities in the mesoblast. Around it a pericardial sack is soon formed, and the walls of the heart become greatly thickened. Four bands pass off from the heart, two dorsalwards which become fixed to the integument, and two ventralwards. There is also a median band of cells connecting the heart with the dorsal integument. The main arteries arise as direct prolongations of the heart. Dohrn's observations on Asellus greatly strengthen the view that the heart originates from a solid mesoblastic mass, in that he was able to observe the hollowing out of the mass in the living embryo (cf. the development of the heart in Spiders). Some of the central cells (nuclei, Dohrn) become blood corpuscles. The formation of these is not, according to Dohrn, confined to the heart, but takes place _in situ_ in all the parts of the body (antennæ, appendages, etc.). The corpuscles are formed as free nuclei and are primarily derived from the yolk, which at first freely communicates with the cavities of the appendages.

[209] Reichenbach describes these cells, and states that there is
a thickening of the epiblast adjoining them. In one place he
states that the heart arises from this thickening of epiblast,
and in another that it arises from the mesoblast. An epiblastic
origin of the heart is extremely improbable.

Alimentary tract. In Astacus the formation of the mesenteron by invagination, and the absorption of the yolk by the hypoblast cells, have already been described. On the absorption of the yolk the mesenteron has the form of a sack, the walls of which are formed of immensely long cells--the yolk pyramids--at the base of which the nucleus is placed (fig. 238 B). This sack gives rise both to the portion of the alimentary canal between the abdomen and the stomach and to the liver. The epithelial wall of both of these parts is formed by the outermost portions of the pyramids with the nuclei and protoplasm becoming separated off from the yolk as a layer of flat epithelial cells. The yolk then breaks up and forms a mass of nutritive material filling up the cavity of the mesenteron.

The differentiation both of the liver and alimentary tract proper first takes place on the ventral side, and commences close to the point where the proctodæum ends, and extends forward from this point. A layer of epithelial cells is thus formed on the ventral side of the mesenteron which very soon becomes raised into a series of longitudinal folds, one of which in the middle line is very conspicuous. The median fold eventually, by uniting with a corresponding fold on the dorsal side, gives rise to the true mesenteron; while the lateral folds form parallel hepatic cylinders, which in front are not constricted off from the alimentary tract. The lateral parts of the dorsal side of the mesenteron similarly give rise to hepatic cylinders. The yolk pyramids of the anterior part of the mesenteron, which projects forwards as a pair of diverticula on each side to the level of the stomach, are not converted into hepatic cylinders till after the larva is hatched.

The proctodæum very early opens into the mesenteron, but the stomodæum remains closed till the differentiation of the mid-gut is nearly completed. The proctodæum gives rise to the abdominal part of the intestine, and the stomodæum to the oesophagus and stomach. The commencement of the masticatory apparatus in the latter appears very early as a dorsal thickening of the epithelium.

The primitive mesenteron in Palæmon differentiates itself into the permanent mid-gut and liver in a manner generally similar to that in Astacus, though the process is considerably less complicated. A distinct layer of cells separates itself from the outer part of the yolk pyramids, and gives rise to the glandular lining both of the mid-gut and of the liver. The differentiation of this layer commences behind, and the mid-gut very soon communicates freely with the proctodæum. The lateral parts of the primitive mesenteron become constricted into four wings, two directed forwards and two backwards; these, after the yolk in them has become absorbed, constitute the liver. The median part simply becomes the mesenteron. The stomachic end of the stomodæum lies in contact with the mesenteron close to the point where it is continued into the hepatic diverticula, and, though the partition wall between the two becomes early very thin, a free communication is not established till the yolk has been completely absorbed.

The alimentary tract in the Isopoda is mainly if not entirely formed from the proctodæum and stomodæum, both of which arise before any other part of the alimentary system as epiblastic invaginations, and gradually grow inwards (fig. 244). In Oniscus the liver is formed as two discs at the surface of the yolk on each side of the anterior part of the body. Their walls are composed of cubical cells derived from the yolk cells, the origin of which was spoken of on p. 516. These two discs gradually take the form of sacks (fig. 244 B, _li._) freely open on their inner side to the yolk. As these sacks continue to grow the stomodæum and proctodæum do not remain passive. The stomodæum, which gives rise to the oesophagus and stomach of the adult, soon exhibits a posterior dilatation destined to become the stomach, on the dorsal wall of which a well-marked prominence--the earliest trace of the future armature--is soon formed (fig. 244 B, _zp_). The proctodæum (_pr_) grows with much greater rapidity than the stomodæum, and its end adjoining the yolk becomes extremely thin or even broken through. In the earliest stages it was surrounded by the yolk cells, but in its later growth the yolk cells become gradually reduced in number and appear to recede before it--so much so that one is led to conclude that the later growth of the proctodæum takes place at the expense of the yolk cells.

[FIG. 244. TWO LONGITUDINAL SECTIONS THROUGH THE EMBRYO OF ONISCUS
MURARIUS. (After Bobretzky.)

_st._ stomodæum; _pr._ proctodæum; _hy._ hypoblast formed of large
nucleated cells imbedded in yolk; _m._ mesoblast; _vg._ ventral
nerve cord; _sg._ supra-oesophageal ganglion; _li._ liver; _do._
dorsal organ; _zp._ rudiment of masticatory apparatus.]

The liver sacks become filled with a granular material without a trace of cells; their posterior wall is continuous with the yolk cells, and their anterior lies close behind the stomach. The proctodæum continually grows forwards till it approaches close to the stomodæum, and the two liver sacks, now united into one at their base, become directly continuous with the proctodæum. By the stage when this junction is effected the yolk cells have completely disappeared. It seems then that in Oniscus the yolk cells (hypoblast) are mainly employed in giving rise to the walls of the liver; but that they probably also supply the material for the later growth of the apparent proctodæum. It becomes therefore necessary to conclude that the latter, which might seem, together with the stomodæum, to form the whole alimentary tract, does in reality correspond to the proctodæum and mesenteron together, though the digestive fluids are no doubt mainly secreted not in the mesenteron but in the hepatic diverticula. The proctodæum and stomodæum at first meet each other without communicating, but before long the partition between the two is broken through.

In Cymothoa (Bullar, No. 499) the proctodæum and stomodæum develop in the same manner as in Oniscus, but the hypoblast has quite a different form. The main mass of the yolk, which is much greater than in Oniscus, is not contained in definite yolk cells, but the hypoblast is represented by (1) two solid masses of cells, derived apparently from the inner layer of blastoderm cells, which give rise to the liver; and (2) by a membrane enclosing the yolk in which nuclei are present.

The two hepatic masses lie on the surface of the yolk, and each of them becomes divided into three short cæcal tubes freely open to the yolk. The stomodæum soon reaches its full length, but the proctodæum grows forwards above the yolk till it meets the stomodæum. By the time this takes place the liver cæca have grown into three large tubes filled with fluid, and provided with a muscular wall. They now lie above the yolk, and no longer communicate directly with the cavity of the yolk-sack, but open together with the yolk-sack into the point of junction of the proctodæum and stomodæum. The yolk-sack of Cymothoa no doubt represents part of the mesenteron, but there is no evidence in favour of any part of the apparent proctodæum representing it also, though it is quite possible that it may do so. The relations of the yolk-sack and hepatic diverticula in Cymothoa appear to hold good for Asellus and probably for most Isopoda.

The differences between the Decapods and Isopods in the development of the mesenteron are not inconsiderable, but they are probably to be explained by the relatively larger amount of food-yolk in the latter forms. The solid yolk in the Isopods on this view represents the primitive mesenteron of Decapods after the yolk has been absorbed by the hypoblast cells. Starting from this standpoint we find that in both groups the lateral parts of the mesenteron become the liver. In Decapods the middle part becomes directly converted into the mid-gut, the differentiation of it commencing behind and proceeding forwards. In the Isopods, owing to the mesenteron not having a distinct cavity, the differentiation of it, which proceeds forwards as in Decapods, appears simply like a prolongation forwards of the proctodæum, the cells for the prolongation being probably supplied from the yolk. In Cymothoa the food-yolk is so bulky that a special yolk-sack is developed for its retention, which is not completely absorbed till some time after the alimentary canal has the form of a continuous tube. The walls of this yolk-sack are morphologically a specially developed part of the mesenteron.

BIBLIOGRAPHY.

_General Works._

(447) C. Spence Bate. "Report on the present state of our knowledge of the Crustacea." _Report of the British Association for 1878._

(448) C. Claus. _Untersuchungen zur Erforschung der genealogischen Grundlage des Crustaceen-Systems._ Wien, 1876.

(449) A. Dohrn. "Geschichte des Krebsstammes." _Jenaische Zeitschrift_, Vol. VI. 1871.

(450) A. Gerstaecker. Bronn's _Thierreich_, Bd. V. _Arthropoda_, 1866.

(451) Th. H. Huxley. _The Anatomy of Invertebrated Animals._ London, 1877.

(452) Fritz Müller. _Für Darwin_, 1864. Translation, _Facts for Darwin._ London, 1869.

_Branchiopoda._

(453) Brauer. "Vorläufige Mittheilung über die Entwicklung u. Lebensweise des Lepidurus (Apus) productus." _Sitz. der Ak. d. Wiss. Wien_, Vol. LXIX., 1874.

(454) C. Claus. "Zur Kenntniss d. Baues u. d. Entwicklung von Branchipus stagnalis u. Apus cancriformis." _Abh. d. könig. Gesell. der Wiss. Göttingen_, Vol. XVIII. 1873.

(455) C. Grobben. "Zur Entwicklungsgeschichte d. Moina rectirostris." _Arbeit. a. d. zoologisch. Institute Wien_, Vol. II., 1879.

(456) E. Grube. "Bemerkungen über die Phyllopoden nebst einer Uebersicht etc." _Archiv f. Naturgeschichte_, Vol. XIX., 1853.

(457) N. Joly. "Histoire d'un petit Crustacé (Artemia salina, _Leach_) etc." _Annales d. Sciences Natur._, 2nd ser., Vol. XIII., 1840.

(458) N. Joly. "Recherches zoologiques anatomiques et physiologiques sur l'Isaura cycladoides (=Estheria) nouveau genre, etc." _Annales d. Sciences Nat._, 2nd ser., Vol. XVII., 1842.

(459) Lereboullet. "Observations sur la génération et le développement de la Limnadia de Hermann." _Annales d. Sciences Natur._, 5th ser., Vol. V., 1866.

(460) F. Leydig. "Ueber Artemia salina u. Branchipus stagnalis." _Zeit. f. wiss. Zool._, Vol. III., 1851.

(461) G. O. Sars. "Om en dimorph Udvikling samt Generationsvexel hos Leptodora." _Vidensk. Selskab. Forhand_, 1873.

(462) G. Zaddach. _De apodis cancreformis Schaeff. anatome et historia evolutionis. Dissertatio inauguralis zootomica._ Bonnæ, 1841.

_Nebaliadæ._

(463) C. Claus. "Ueber den Bau u. die systematische Stellung von Nebalia." _Zeit. f. wiss. Zool._, Bd. XXII. 1872.

(464) E. Metschnikoff. _Development of Nebalia_ (Russian), 1868.

_Schizopoda._

(465) E. van Beneden. "Recherches sur l'Embryogénie des Crustacés. II. Développement des Mysis." _Bullet. de l'Académie roy. de Belgique_, second series, Tom. XXVIII. 1869.

(466) C. Claus. "Ueber einige Schizopoden u. niedere Malakostraken." _Zeit. f. wiss. Zoologie_, Bd. XIII., 1863.

(467) A. Dohrn. "Untersuchungen üb. Bau u. Entwicklung d. Arthropoden." _Zeit. f. wiss. Zool._, Bd. XXI., 1871, p. 375. Peneus zoæa (larva of Euphausia).

(468) E. Metschnikoff. "Ueber ein Larvenstadium von Euphausia." _Zeit. für wiss. Zool._, Bd. XIX., 1869.

(469) E. Metschnikoff. "Ueber den Naupliuszustand von Euphausia." _Zeit. für wiss. Zool._, Bd. XXI., 1871.

_Decapoda._

(470) Spence Bate. "On the development of Decapod Crustacea." _Phil. Trans._, 1858.

(471) Spence Bate. "On the development of Pagurus." _Ann. and Mag. Nat. History_, Series 4, Vol. II., 1868.

(472) N. Bobretzky. _Development of Astacus and Palæmon._ Kiew, 1873. (Russian.)

(473) C. Claus. "Zur Kenntniss d. Malakostrakenlarven." _Würzb. naturw. Zeitschrift_, 1861.

(474) R. Q. Couch. "On the Metamorphosis of the Decapod Crustaceans." _Report Cornwall Polyt. Society._ 1848.

(475) Du Cane. "On the Metamorphosis of Crustacea." _Ann. and Mag. of Nat. History_, 1839.

(476) Walter Faxon. "On the development of Palæmonetes vulgaris." _Bull. of the Mus. of Comp. Anat. Harvard, Cambridge, Mass._, Vol. V., 1879.

(477) A. Dohrn. "Untersuchungen üb. Bau u. Entwicklung d. Arthropoden." "Zur Entwicklungsgeschichte der Panzerkrebse. _Scyllarus Palinurus._" _Zeit. f. wiss. Zool_., Bd. XX., 1870.

(478) A. Dohrn. "Untersuchungen üb. Bau u. Entwicklung d. Arthropoden. Erster Beitrag z. Kenntniss d. Malacostraken u. ihrer Larven Amphion Reynaudi, Lophogaster, Portunus, Porcellanus, Elaphocaris." _Zeit. f. wiss. Zool._, Bd. XX., 1870.

(479) A. Dohrn. "Untersuchungen üb. Bau u. Entwicklung d. Arthropoden. Zweiter Beitrag, etc." _Zeit. f. wiss. Zool._, Bd. XXI., 1871.

(480) N. Joly. "Sur la Caridina Desmarestii." _Ann. Scien. Nat._, Tom. XIX., 1843.