I have seen P. setosus only. P. planifrons is a perfectly distinct species, and is found abundantly in many places in the North Island of New Zealand. P. setosus is not known to occur in the North Island, but it is widely distributed in the South Island, being found in the River Avon, Christchurch, from which the specimens for this paper were obtained, and also in the rivers near Invercargill, at the south of the island. Thus P. planifrons appears to be confined to the North Island, and to be represented in the South Island by P. setostus.

P. zealandicus was described as belonging to New Zealand by White in 1847,^* but it does not appear to have been since recognized. Professor Hutton, who at the time when he described P. setosus in 1873^† had no opportunity of consulting White's description, tells me that he thinks it very probable that P. zealandicus is nothing more than a young specimen of P. setosus. From the comparison of the two descriptions in Miers' Catalogue, and from the figure in the Zoology of the Voyage of the Erebus and Terror, this appears very likely to be the case, and it also agrees well with the small size of P. zealandicus (3 inches) as compared with that of P. setosus (5 ⅖ inches).

With regard to P. zealandicus, Mr. Miers says:^‡ This species is certainly distinct from P. setosus, Hutton. In P. zealandicus, of which the type specimens are in the British Museum Collections, the hands are clothed externally with tufts of hair arranged in longitudinal series, and are armed with spines only upon the superior margins, and the sides of the carapace are smooth. In P. setosus there are spines arranged seriately upon the external surface as well as the upper margin of the hand, and the branchial and hepatic regions of the carapace are armed with numerous inequal conical spines.” The first point will certainly not serve to distinguish the two species, for there are tufts of hair on the hand of P. setosus, there having been a slight mistake in Professor Hutton's description (see below). With regard to the other points they are certainly subject to some variation in P. setosus, and it is quite possible that the spines on the hands and on the sides of the carapace may be developed only in the older specimens, but I have not been able to examine a sufficiently large number of specimens to give a decisive answer on this point. However, there are certainly only the two species, P. planifrons and P. setosus, known to New Zealand collectors, and this leads one to think either that P. zealandicus is not a distinct species or that the locality given is wrong, and that it belongs to Fiji, where a species of Paranephrops is found.^§

The most important result arising from the examination of Paranephrops setosus is that its affinity to Palinurus now seems to be placed beyond doubt. Paranephrops and the Parastacidœ generally resemble the Palinuridœ in that they have no appendages upon the first abdominal segment; in this they differ from the crayfishes of the Northern Hemisphere, and from Homarus and Nephrops. The Palinuridœ and the Parastacidœ also agree in having hooked setæ,^* while in the Potamobiidœ and in the lobsters the setæ are not hooked. Moreover the branchial formulæ of Palinurus and and Paranephrops are almost identical. Taking the presence or absence of the first abdominal appendages as the basis of his classification, Professor Huxley placed the Palinuridœ and the Parastacidœ together as the Astyla, while the Potamobiidœ Homaridœ, etc., together form the Stylophora. This classification is confirmed by the structure of the male reproductive organs in Paranephrops setosus, for these agree in every essential particular with those of Palinurus vulgaris as described by Brocchi.^† In both the testes consist of two long tubes united towards their anterior ends by a commissure; both have long convoluted vasa deferentia; and in both the extremities are greatly expanded. This would appear to be the oldest and most generalized form of the male reproductive organs of crayfishes, New Zealand having preserved the old form in this as in many other cases. In Astacus fluviatilis the testes are very different in shape; they are trilobed, two lobes being directed anteriorly, and one posteriorly: thus the two posterior portions, which in Paranephrops and Palinurus are quite distinct, appear in Astacus fluviatilis to have coalesced into a single mass. The vasa deferentia of Astacus are much convoluted, but their extremities are not expanded or at least only slightly so. Professor Huxley gives them without any expansion,^‡ but in Milne-Edwards' figure they are slightly expanded.^§ The male reproductive organs of Homarus vulgaris appear to be intermediate between those of Paranephrops setosus and of Astacus fluviatilis, for in them the two posterior portions are close together and apparently confluent, though not so perfectly coalescent as in Astacus; the vasa deferentia are curiously enough not convoluted, though their extremities are considerably expanded.^∥

In accounting for the origin and present distribution of the crayfishes Professor Huxley says:—“Let it be supposed that, at some former period

of the earth's history, a Crustacean, similar to Paranephrops or Astacopsis in its general characters, but with the first pair of abdominal appendages fully formed, which we may call provisionally Protastacus, inhabited the ocean, and that it had as wide a distribution as Palœmon or Penœus at the present day. Let us suppose, further, that the northern form of the genus tended towards the assumption of the Potamobiine, and the southern towards that of the Parastacine type. Under these circumstances it is easy to understand how such rivers as were, or became, accessible in both hemispheres, and were not already too strongly tenanted by formidable competitors, might be peopled respectively by Potamobiine or Parastacine forms, which, acquiring their special characters in each great river-basin, would bring about the distribution we now witness. As time went on, the Protastacus stock might become extinct, or might be represented only by rare deepwater forms, as the Homaridœ are represented in the Indian Ocean only by Nephropsis.”^*

The comparison of the male reproductive organs in Palinurus, Paranephrops, Homarus, and Astacus, appears to lend every support to this hypothesis, and I have only to add that the Protastacus stock appears to have left Palinurus, which has lost the chelate limbs possessed by its ancestors, as its marine representative in the Southern Hemisphere.

Though Paranephrops is thus in all probability more nearly akin to Palinurus than it is to Astacus, yet it is curious to notice that in general appearance it resembles Astacus much more nearly than it does Palinurus. I am not referring to the absence of chelate limbs in Palinurus, for they must obviously have been lost after the Parastacidœ branched off, but to the size, the colour, the shape of the antennules and antennæ and their size relatively to that of the animal, the narrow thoracic sterna, the movability of the last thoracic segment and the shape of the abdominal appendages. Some of these points, such as the size and colour, are no doubt due to mere adaptation to surrounding circumstances, but it seems difficult to believe that the other resemblances to Astacus can be due to the same cause.

My observations with regard to the gastroliths or “crab's eyes” of Paranephrops do not agree with those given by Professor Huxley for Astacus in “The Crayfish,” and I therefore mention it here referring to the body of the paper for the details. He states that the Gastroliths “are found fully developed only in the latter part of the summer season, just before ecdysis sets in,” and that they “are cast off with the gastric armature in general.” I have, however, obtained specimens in September and October (i.e. in the Spring) with gastroliths present. Some caught in September had very small gastroliths, but one caught in October had them

very largely developed; and I have also had two specimens in which ecdysis had evidently taken place some little time before their capture and yet the gastroliths were very largely developed in both. The conclusion I have therefore to draw is that the development of the gastroliths in Paranephrops differs from that in Astacus.

Paranephrops setosus was first described by Professor Hutton in 1873.^*

The various parts mentioned in his description will be more minutely described in their proper places, but there is one small point that needs amending. This is with regard to the hairs on the great claws. These are described as “distant long stiff hairs, the tips of which are often split,” but these are in reality small tufts of hairs each containing from six to twelve separate hairs. In each tuft there are two kinds of hairs, one naked and jointed the other plumose and without joints (see under integument). The hairs in each tuft become closely united together in dried or spirit specimens, and thus appear very like single stiff hairs, and this is no doubt the cause of the mistake.

The thoracic sterna in Paranephrops setosus are quite narrow, and those corresponding to the first four pairs of ambulatory legs are firmly united together, while that of the last thoracic segment is separate, so that this segment is more or less movable. In this Paranephrops resembles Astacus and differs from Palinurus.

The inferior edge of the pleura of the third abdominal somite is rounded; the anterior portion is slightly more convex than the posterior part and is fringed with several plumose setæ; the posterior portion is almost or quite naked (pl. XXI., fig. 9).

The rostrum (pl. XIX., fig. 9) has been already well described by Professor Hutton. The end projects slightly upwards. On the under surface there are two large, median, rather blunt teeth. These teeth and likewise the four teeth on the two sides are subject to some variation, for in one specimen there were four teeth on one side and only three on the other, and in another specimen there was only one tooth underneath.

The telson (pl. XX., fig. 4) consists of a single piece, there being no transverse hinge. On each side about one-third of its length from the end there is a strong single spine. The extremity is rounded and fringed with plumose setæ.

Integument.—The integument is completely calcified throughout, excepting such portions as must necessarily remain soft and flexible to allow of the movements of the various parts of the body. The telson and the appendages of the sixth abdominal somite are hard and not semi-membranaceous as in Palinurus. The hardest parts are the anterior portion of the

carapace and the great claws. In both of these, but more especially in the latter, the integument is very thick and hard and is beset with many strong sharp spines.

By making sections of parts of the integument it was found to be much the same as that of Astacus figured on page 191 of “The Crayfish.” It showed the epiostracum (pl. XXI., fig. 3 a), the ectostracum (b), and the endostracum (c), the last being much thicker than the other two together. In the inner part of the ectostracum there were numerous rectangular projections extending from the endostracum about half-way into the ectostracum, but these were very irregular, no two sections being alike as far as these projections were concerned.

Setœ.—There are two kinds of setæ found abundantly on various parts of the body. The first (pl. XXI., fig. 8) consists of a central stem which bears numerous filaments on it, so that it is densely plumose. The stem is all in one piece and is not jointed. In the other kind (pl. XXI., fig. 7) the stem has a kind of joint about the middle. The basal half is quite naked but the terminal half is minutely serrate towards the end, which is usually slightly curved. Both kinds are lodged in a sort of socket in the integument at their base.

The plumose setæ are found on almost all parts of the body, but they are thickest on the telson, the abdominal appendages, the chelæ and the antennæ. The jointed setæ are found chiefly on the chelæ and the antennæ. On the chelæ both kinds are found together forming little clusters each containing about five or six plumose setæ and two or three jointed setæ. Modified forms of these setæ are found in the respiratory organs, in the stomach, and on the antennæ, and are described in their proper places.

Appendages.—The antennule (pl. XIX., fig. 1) has the peduncle formed of three joints. The first or basal joint is somewhat longer than the other two together and is broader at its base than at its distal end. On the inner side towards the distal end there is a single, strong, sharp spine. The outer edge is thickly beset with plumose cilia. The aperture of the auditory sac is on the upper side of this basal joint. The next two joints are slightly narrower than the first, but they are much broader in comparison than the two corresponding joints in Astacus. Both bear numerous plumose setæ on either side. The third joint supports the two flagella which represent the exopodite and endopodite. The outer one, exopodite (ex), is larger and longer than the inner one, endopodite (en). In both of the flagella each joint bears some of the two kinds of setæ, the plumose setæ being the more numerous on each joint. In addition to these most of the joints of the endopodite bear a tuft-of olfactory filaments (see olfactory organs).

The protopodite of the antenna (pl. XIX., fig. 2) is two-jointed. The basal joint is small and bears the conical opening of the green gland (gg) and just above it a strong spine. The other joint of the protopodite is large and broad and is divided into two parts more or less movable upon one another. There is a strong spine on the outer side at the base of the scale and another on the inside; and there are three or four other spines on the under surface of the protopodite. The exopodite (ex) is represented by the squame or scale, which reaches beyond the basal joints of the endopodite. Its inner edge is curved and fringed with plumose setæ. The base of the endopodite is composed of two joints of about equal size, each of which bears setæ abundantly on the inner side. Each joint of the flagellum bears some of both kinds of setæ.

Mandible (pl. XIX., fig. 3).—The protopodite appears to be essentially the same as in Astacus. The palp is 3-jointed but the first two joints are not movable upon one another and the joint between them is somewhat indistinct. The third joint is freely movable upon the second, it is slightly bent near its proximal end, its extremity is rounded and provided with many setæ.

In the first maxilla (pl. XIX., fig. 4) the endopodite (en) is better developed than in Astacus; it consists of a basal joint followed by a short indistinctly segmented flagellum. The basipodite (bp) is pear-shaped while the coxopodite (cxp) is comparatively small and narrow. Both have their extremities fringed with short styliform setæ.

The second maxilla (pl. XIX., fig. 5) closely resembles that of Astacus in general appearance, but the scaphognathite (sg) is small compared with the rest of the appendage. The endopodite (en) is small and undivided. The coxopodite (cxp) and basipodite (bp) are lamellar and are subdivided by deep fissures into four portions which are much narrower than the corresponding parts in Astacus. Their extremities are fringed with setæ.

In the first maxillipede (pl. XIX., fig. 6) the coxopodite (cxp) and the basipodite (bp) are broad thin plates with setose edges; the basipodite is considerably larger than the coxopodite. The endopodite (en) is small and consists of a short basal joint and an indistinctly segmented flagellum slightly longer than the basal joint. The basal joint of the exopodite (ex) is long and is followed by a flagellum about as long as itself. The epipodite (ep) is a soft membranous plate rounded at the extremity and bearing numerous branchial filaments.

Second maxillipede (pl. XIX., fig. 7).—The exopodite (ex) is large compared with the rest of the limb; the basal joint is long, thin and fringed with setæ, the flagellum is very well developed, being as long as the basal joint. The coxopodite (cxp) is large and broad and bears the podo-

branchia (pb) which has no lamella. The anterior arthrobranchia (ab) (attached to the membrane uniting the base of the limb to the part of the thorax to which it is articulated), also often comes away with the limb. The basipodite (bp) and ischiopodite (ip) are both short and of nearly equal length, the latter is broader than long. The meropodite (mp) is the longest joint in the endopodite, it is about twice as long as broad. The inner edge of these three last-mentioned joints is abundantly supplied with setæ. The carpopodite (cp) is small and narrow. The protopodite (pp) is subtriangular, expanding towards the distal end, the dactylopodite (dp) is small and rounded. The last two joints are fringed with setæ on both sides.

Third maxillipede (pl. XIX., fig. 8).—In this appendage the exopodite (ex) is small compared with the rest of the limb, it reaches to nearly the end of the ischiopodite of the endopodite. It is composed of two parts, the undivided basal joint and the flagellum, the former being slightly the longer. The coxopodite (cxp) bears the podobranchia (pb) and also the corresponding anterior arthrobranchia (ab); at the base of the podobranchia is a small tuft of coxopoditic setæ (cxs) [see under Respiratory Organs]. The basipodite (bp) is indistinctly divided from the ischiopodite (ip) which is the longest joint of the endopodite. There are three or four spines on its inner edge, and the outer edge is produced distally into a long sharp spine. The meropodite (mp) is little more than half as long as the ischiopodite, on its inner edge are three spines. The carpopodite (cp) is narrow at its proximal end, but expands considerably at the distal end, the inner corner of which bears a sharp spine. The propodite (pp) is about as long as the ischiopodite, it narrows towards the distal end; the dactylopodite (dp) is narrow, being about three times as long as broad. The inner edges of the basipodite and succeeding joints are fringed with setæ.

The great claws each bear a large podobranchia and a small tuft of coxopoditic setæ. The various joints of the limb are abundantly supplied with strong spines arranged as follows:—The ischiopodite has two strong sharp spines on the inside, and two short blunt ones on the outside; the meropodite has on the inside two rows containing six and four spines respectively, and at the distal end a single spine placed between these two rows, on the outside is a row of four spines, and there are one or two other spines at the distal end; the carpopodite has three large spines on the inside, one large one on the outside, and smaller ones scattered over the joint; on the under side of the propodite there are two rows of three and four spines respectively and one spine irregularly placed. On the side on which the movable finger is there are five large spines in a row, and one large one more to the outside, on the other side are two rows extending right up on to the fixed

finger, each row containing about twelve spines. On the outside there is a central row, and also several other spines more or less irregularly placed. On the outer edge the dactylopodite bears two rows of spines, six in one and four in the other, and there are two small spines at the tip. Each finger ends in a strong spine pointing towards each other, and on the inside of the fingers are three or four rounded prominences. There are numerous tufts of setæ on most of the joints, but most abundantly on the propodite.

The four posterior pairs of ambulatory legs are somewhat slender; in all except the last the coxopodite bears a podobranchia and a small tuft of coxopoditic setæ.

The first pair of abdominal appendages are entirely absent both in the male and female.

The second, third, fourth and fifth abdominal appendages are all alike, and are rather simpler than those of Astacus. The coxopodite (pl. XX., fig. 1 cxp) is very short, and is followed by the long cylindrical basipodite (bp) which supports the exopodite (ex) and the endopodite (en). In the male these are of about equal length, and are imperfectly articulated through their whole length, neither of them having an undivided basal joint as in Astacus. Their edges are fringed with long plumose setæ. In a female with the eggs attached under the abdomen the appendages (pl. XX., fig. 2) were found to differ somewhat from those of the male. They were much slenderer and the exoskeleton was much softer; the endopodite (en) was considerably longer than the exopodite (ex), and in both the articulations were very indistinct; the setæ were long and did not appear to be plumose. In a young female, however, in which the eggs were still in the ovary, the abdominal appendages were much more like those of the male, and were supplied with plumose setæ.

The appendages of the sixth abdominal somite have the coxopodite (pl. XX., fig. 3, cxp) broad and indistinctly divided into two or three parts. The exopodite (ex) is in the form of a broad oval plate divided into two parts by a transverse hinge, the basal part ends distally in a row of short spines of which the outside one is the longest. The terminal portion is rounded and fringed with plumose setæ. A median ridge runs from the coxopodite through the whole length of the exopodite. The endopodite (en) is of similar shape, but consists of one piece only, the median ridge ends in a sharp spine at some distance from the edge.

Respiratory organs.—These differ considerably from those of Astacus, and closely resemble those of Astacopsis.^* The epipodite of the first maxillipede is in the form of a broad more or less oval-shaped lamina, the end of which bears numerous branchial filaments similar to the filaments of the

true branchiæ. Hence this epipodite must be looked upon as forming part of the respiratory organs. The next six appendages, viz., the second and third maxillipedes and the first four pairs of ambulatory legs, each bear a podobranchia. These are larger than any of the other branchiæ, but they have no lamina or epipodite corresponding to that of Astacus, and many though not all of the branchial filaments have hooked apices (pl. XXI., fig. 5). Each of these six appendages also bears on its interarticular membrane an arthrobranchia. These correspond to the anterior arthrobranchia of Huxley. They are all well developed, but are considerably smaller than the podobranchiæ. To all these appendages except the first, the second maxilla, there is also the corresponding posterior arthrobranchia. These are all of small size, the largest being composed of comparatively few filaments, and they become smaller from before backwards. In fact the last one is almost if not quite rudimentary, though evidently subject to some variation, for, in one specimen, on the left side it was composed of a single short filament, while on the right the filament was longer and bore a short filament branching from it. Four pleurobranchiæ are attached to the epimera of the four hindmost thoracic somites. They are all of moderate size.

These facts may therefore be tabulated thus:—

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

The coxopoditic setæ which are found on the coxopodites of the four pairs of ambulatory limbs, and also of the third pair of maxillipedes, differ considerably from those of Astacus. They are few in number and much shorter and straighter than those of Astacus. The setæ of which each tuft

is composed appear to be modifications of the naked jointed setæ already described. There is a joint about the middle (pl. XXI., fig. 4 a). The basal part is quite naked, the distal half is naked until some little distance past the joint, but is then thickly covered with short simple filaments. These filaments extend nearly to the end of the seta, which is hooked (b). In the concave portion of the hook there is often a slight projection (c).

The inner surface of each branchiostegite is thickly covered with jointed setæ. In these the filaments on the distal half are less conspicuous than in the coxopoditic setæ; in this point they are intermediate between the coxopoditic setæ and the ordinary jointed setæ found on the chelæ, etc., but they resemble the coxopoditic setæ in having their extremities hooked. On the inferior edge of the branchiostegite there is a row of setæ hanging downwards. These are similar to those found on the inner surface of the branchiostegite except that the extremities are not hooked. I am quite ignorant of the function of all the hooked setæ that I have described; they appear to have something to do with the respiratory organs, as it is only those in immediate connection with the respiratory organs that are hooked.

Circulatory system.—The circulatory system, as far as could be seen without injection, does not appear to differ in any important particular from that of Astacus. The heart is of similar shape and lies behind the stomach and above the intestine and reproductive organs. The abdominal artery was readily seen running along the dorsal surface of the abdomen and giving off branches in each somite. The sternal artery was also seen passing vertically downwards to the ventral surface of the animal, where it divides into an anterior and a posterior branch. The arteries arising from the anterior portion of the heart are smaller and are not so readily seen.

Alimentary system.—The general course of the alimentary canal is, as might have been expected, very much like that of Astacus. The œsophagus is large in section and expands almost immediately into the capacious stomach. The stomach consists of two parts, the cardiac and the pyloric, the former of which contains a gastric armature, which is fully as complicated as that of Astacus. It is formed on the same type, so that the same names can be used in describing the various parts. The anterior edge of the cardiac ossicle (pl. XX., figs. 5 and 6 c) is much more convex than in Astacus; and the remaining part is divided into four portions, as shown in fig. 5. The urocardiac process (uc) is more or less oblong, not quite twice as long as broad, with the sides slightly concave, at the posterior end are two rounded prominences. The median tooth (mt) is dense and hard; the end curves forwards and is bifurcated at the summit. The urocardiac process and the median tooth are united by the prepyloric ossicle (pp) to the pyloric ossicle (p) in the same way as in Astacus. The pterocardiac

ossicles (pc) have much the same shape and position as in Astacus. The lateral teeth (lt) are large and reddish-yellow in colour. Anteriorly the teeth are large and distinct and there is one large tooth placed on one side of the row; posteriorly the teeth decrease greatly in size, but there are two or three rows so that they form an efficient grinding apparatus. Beneath the lateral teeth on each side there is a small, single, sharp tooth at the end of a long, thin, calcified bar. In a side view of the stomach (fig. 6) this tooth is seen to be also supported by a broad plate, the inner surface of which is thickly covered with short, stiff, plumose setæ. On the raised edge of this plate, projecting from either side into the interior of the stomach, there is a row of setæ similar to the others except in length. These stretch across and meet in the centre and appear to be for the purpose of stopping the food and forcing it to pass through the gastric armature already described. They (pl. XXI., fig. 6) are very long and slender, of the same size throughout almost the whole of their length, often slightly curved towards the end. The stem is unjointed and is covered with filaments, which are not much longer than the diameter of the stem itself.

The pyloric portion of the stomach seems to be essentially the same as in Astacus. The cœcum (pl. XX., fig. 6 cœ) appears to be variable, for though I have seen it quite distinctly in some specimens, I have been unable after careful search to find it in others. The cardio-pyloric valve (cpu) is present as in Astacus, and the transverse section of the pyloric region is so very much like that of Astacus that I have not given a figure of it.^* At the opening of the pyloric sac into the intestine there are valves, one median, one above, and apparently only one on each side.

From the pyloric end of the stomach the intestine passes direct to the anus on the ventral surface of the telson. There is no cœcum in connection with the rectum as there is in the lobster, Homarus vulgaris.^†.

At the sides of the stomach gastroliths were found in some specimens. These evidently differ much in shape according to their state of development, and when fully developed they differ considerably from those of Astacus. The side turned towards the stomach is either flat or slightly concave. The part which forms the convex side is doubled over so as to join with the flat or concave side, the junction between the two forms a wellmarked nearly circular indentation. The flat or concave portion inside this ring is quite smooth. The convex side is more convex than the corresponding part in the gastroliths of Astacus, and it also differs in being quite

smooth while that of Astacus being “rough with irregular prominences, is something like brainstone coral.” The gastroliths of Paranephrops are usually pale blue in colour.

The facts with regard to the occurrence of the gastroliths in the specimens of Paranephrops setosus that I have examined are as follows:—First specimen, male, caught October, 1880 (i.e. in the Spring), ecdysis had taken place shortly before this animal was caught, for the shed gastric armature was found in the stomach, the exoskeleton, however, was moderately hardened—the gastroliths were very large: second specimen, male, caught about the middle of April, 1881, gastroliths rather small: third specimen, female, caught shortly afterwards, gastroliths absent: fourth specimen, female, caught May, 1881; in this specimen ecdysis had taken place shortly before its capture, the integument was thin and scarcely hardened and the shed gastric armature and membrane were found in the stomach, gastroliths very large: fifth specimen caught later on in May, no gastroliths. Three other specimens were obtained in September, 1881 (i.e. early in Spring), of these in two the gastroliths were present though very small, in the third there were no gastroliths. It is therefore evident that the development of the gastroliths in Paranephrops setosus must differ from that in Astacus fluviatilis.

The liver is large and yellow in colour, but it does not appear to differ essentially from that of Astacus.

The green gland as in Astacus consists of two portions, first the gland which is green in colour and communicates with the second part, the sac, which has extremely delicate whitish walls and opens at the base of the antenna (pl. XIX., fig. 2, gg). When examined microscopically the gland is seen to contain granular cells, but I did not make out the “much convoluted tube” of Lezdig.^*.

Nervous system and sense organs.—The main portion of the nervous system, viz., the chain of ganglia on the ventral aspect of the body, appears to closely resemble that of Astacus.

Eye.—As in Astacus the cornea of the eye is “divided into a great number of minute usually square facets, by faint lines, which cross it from side to side nearly at right-angles with one another.”^† I have not studied the internal structure of the eye.

The auditory sacs, which are situated in the basal joint of the antennules, closely resemble those of Astacus, and the auditory hairs from them are exceedingly like the one figured by professor Huxley.^‡

Olfactory organs.—These are borne on the exopodite of the antennule. The various joints in the flagellum differ in shape from those of Astacus, for the distal end of each is considerably enlarged; this enlargement is chiefly on the under side, for, while the upper edge is nearly straight, the under edge curves so as to form a rounded protuberance towards the distal end (pl. XXI., fig. 1). On these protuberances are situated the tufts of olfactory setæ (c), thus there is only one tuft of olfactory setæ on each joint, wbile in Astacus there are two on each joint. Each tuft arises from a slight cavity in the joint and consists of usually five or six olfactory setæ.

The olfactory setæ are similar in shape to those of Astacus. Each consists of two parts (pl. XXI., fig. 2), which at first sight appear to be divided by a joint, but on more careful observation it appears that this is not really so, but that the walls of the two parts are continuous, and that the appearance of a joint is caused by the basal part being filled with opaque granular matter while the distal part is clear and transparent. This opaque granular matter extends up the sides further than it does in the middle of the seta.

Reproductive organs.—Male (pl. XX., fig. 7). The testes (t) consist of two long tubes united towards their anterior ends by a transverse portion or commissure. The two parts in front of this commissure lie between the heart and the posterior portion of the stomach, and are somewhat directed upwards towards the dorsal surface of the body. Immediately in front of the commissure they become considerably narrower. The portions of the tubes behind the commissures are narrow at first but they soon widen and then soon contract again at the origin of the vasa deferentia. After this they again widen out and at their posterior ends are more closely approximated. Thus a little behind the commissure a considerable space is left between the two tubes, and in this space the heart rests. The vas deferens (vd) arises as a very fine tube. The first part differs from the remainder in being smaller and less boldly curved. The remainder is exceedingly convoluted and increases only very slightly in size until it comes to the portion which proceeds directly downwards to the aperture on the bases of the last pair of ambulatory legs. This (a) is enormously expanded and is not quite cylindrical, being somewhat laterally compressed. In the figure the convolutions have been separated. I have not been able to observe the spermatozoa, but I have seen in some specimens the other substance which Professor Huxley mentions as filling the vasa deferentia together with the spermatozoa. As in Astacus it was of a viscid material and gave “the secretion of the testis the form and consistency of threads of vermicelli,”