changing the sea-water, and that the new form came from the added portion.' The second form has been shown to correspond very exactly with the larval stage of a prawn, and from this Bell weakly argues that the observation of Slabber was correct, although the first form, as Bell had reason to know, was the larval stage of a crab, and in this Slabber correctly observed the gradual dwindling of the horns of the carapace. The minuteness and transparency of these infants and the readiness with which they perish will account for the confusion in regard to the principal change into which he appears undoubtedly to have fallen, but it is remarkable that such an error should have been in close agreement with the real facts of the case, that a discovery apparently so full of interest should have been neglected for half a century, and that then, when at length it was placed upon a solid foundation, the facts should have been hotly and stoutly disputed for a long series of years. In 1837 Milne-Edwards was still undecided on many of the details of the question, but as to the statement made by Vaughan Thompson in 1835 that the great French naturalist had been deputed by the Academy of Science to investigate the development of the Crustacea, that he had passed a summer in the Isle of Ré for that purpose, and had come to the conclusion that the Crustacea are born in their permanent forms, in all that, MilneEdwards retorts, there is not a word of truth. He had never been in the Isle of Ré, he had never denied that some Crustacea underwent considerable changes, and he could only hope that Thompson was more careful in his observations than in his quotations. Notwithstanding this sharp denial, Bell in 1853 still sends Milne-Edwards to the Isle of Ré, and wonders that observations which he never made should have led him to conclusions which he did not entertain.

The larval stages of the American Cancer irroratus have been studied by Professor S. I. Smith. As might have been expected, they agree very nearly with those of the European Cancer pagurus. In its latest stage the Zoëa still has a frontal and a dorsal spine that are

very conspicuous, but at an earlier period it has much longer spines on the carapace, and as yet no rudiments of the legs of the trunk or pleon. After many months they attain the final Zoëa stage, in which the terminal segment of the pleon is very broad, and divided nearly to the base by a broad sinus, formed by long spiniform diverging processes, at the base of which the sinus is armed with six to eight spines on each side. Shortly before the change into the second or Megalopa form, they were not quite so active as previously, but still continued to swim about until they appeared to be seized by violent convulsions, and after a moment began to wriggle rapidly out of the old zoëa skin, and at once appeared in the full megalops form. The new integument seems to stiffen at once, for in a very few moments after freeing itself from the old skin the new megalops was swimming about as actively as the oldest individuals. In this megalops stage the animal begins to resemble the adult. The five pairs of cephalothoracic legs are much like those of the adult, and the mouth-organs have assumed nearly their final form. The eyes, however, are still enormous in size, the carapace is elongated and has a slender rostrum and a long spine projecting from the cardiac region far over the posterior border, and the abdomen is carried extended, and is furnished with powerful swimming legs as in the Macroura.' Professor Smith observed a few instances of the change from the megalops or Megalopa stage to the young crab. "The little crab worked himself out of the megalops skin quite slowly. For a short time after their appearance the young crabs were soft and inactive, but the integument very soon stiffened, and in the course of two or three hours they acquired all the pugnacity of the adult. They swam about with ease, and were constantly attacking each other and their companions in the earlier stages.'

Professor Smith has remarked that in 'The Crayfish,' fig. 74 represents the Zoëa and Megalopa stages of Carcinus manas, not, as stated by a misprint, those of Cancer

pagurus.

If the tiny young of the Crustacea attack and destroy one another, it is not for want of innumerable other enemies fitted to keep their numbers in check. As far as the timidity of human experience can decide, the Crustacea in general, though by no means particular as to the food they consume, invite rapacity by the agreeable quality of the food they supply. The enormous spines of the very young and the strong armature of the adults have probably been called into existence in consequence. Where these are wanting or inadequate, the life of the species has been protected by extreme fertility. In Geryon quinquedens, Smith, for example, it has been computed that one specimen was carrying no less than forty-seven thousand eggs, and there are other species reckoned to be at least twice as prolific.

To the extensive genus Xantho Bell assigns three British species, naming them florida, rivulosa, and tuberculata. But, Montagu's floridus having lapsed as a synonym, the first of the three should be named Xantho incisus, Leach. The second, on Bell's own showing, ought to be called Xantho hydrophilus (Herbst), and of this Couch's tuberculata is now held to be a variety.

Ozius, Milne-Edwards, 1834, was a genus established to receive certain species found in the Indian and Australian waters. The name had been given much earlier by Dr. Leach, but without published description. It presents a peculiarity by help of which the large family of the Cancridæ is divided into two sections. The space between the front margin of the buccal frame and the mouth itself was called by Milne-Edwards the prelabial space. By English writers it is called the endostome or palate. In Cancer, Xantho, and many other genera, this endostome is without distinct longitudinal ridges defining the apertures of the efferent branchial channels, whereas in Ozius, Pilumnus, Eriphia, and others, it has these ridges.

Pseudozius, Dana, 1851, is, as the name implies, a genus that might be mistaken for Ozius, but the crests of the endostome do not quite reach the upper margin of the buccal frame. In 1881 the species Pseudozius Mellissi

6

from St. Helena was carefully described by Mr. Miers, who pointed out its resemblance to and differences from Xantho Bouvieri, A. Milne-Edwards, a species from the Cape Verd Islands. In 1886. Mr. Miers re-described it, and gave a figure of it in his Challenger Report,' but he then placed it in a new sub-genus Euryozius, entitling it Pseudozius bouvieri, var. mellissiï,' in a hesitating manner identifying it with the species Xantho Bouvieri of A. Milne-Edwards. In 1888 Professor Th. Barrois, in his catalogue of the marine Crustacea of the Azores, once more describes this species, and gives a beautiful figure of it in its natural colour of bright orange-red, with black tips to the chelipeds. He and Mr. Miers are in exact agreement in their descriptions, as two such excellent naturalists were likely to be. But Professor Barrois calls the species Ozius Edwardsi, and explains that he had submitted it to the highly competent judgment of M. Alphonse Milne-Edwards, who pronounced it to be a new Ozius, of which he had himself obtained a specimen at the Canaries during the expedition of the Tulisman. It will be consoling to the beginner and the amateur, when involved in perplexity amid species that they cannot name or can only name at random, to find the past masters of the science thus entangled as it were in their own web. For it must not be forgotten that Alphonse Milne-Edwards is acknowledged to be the highest authority on the Brachyura,'' and yet he leads Barrois to make a new species of that which had been twice described and twice named by Miers, and which had probably been already named and described by Professor Milne-Edwards himself. The instance is significant of the stress, to which the highest powers must sometimes prove unequal, of keeping in mind each individual species of the vast multitude now known, and each individual chapter of the vast literature which records them.

Barrois mentions an interesting peculiarity in this elegant crab. The carapace along the antero-lateral margins is obliquely striated on the under side with fine parallel grooves, in correspondence with which the fifth joint or Miers, Challenger Report, p. 146.

wrist of the chelipeds has a long sharp crest, and the rapid rubbing of this crest against the striæ produces a shrill sort of stridulating noise such as a grasshopper makes by drawing the thighs of its hind legs over the salient nervures of its wing-cases.

Pilumnus, Leach, 1815, is represented in Great Britain by the single species Pilumnus hirtellus (Linn.), but for the world at large more than eighty forms have been described under separate specific names, and still await the discriminating criticism of some future monograph. In this genus, as at present defined, the antero-lateral margins are normally armed with spines instead of the usual teeth, and the pleon is seven-jointed in both sexes. But when the description and figures of Pilumnus xanthoides, Krauss, 1843, are examined, they exhibit not spines but rounded teeth or lobes on the antero-lateral border, and a five-jointed pleon in the male. Thus there is prima facie reason to suppose that this species ought to be removed to some other genus. Otherwise the boundaries of the existing genus must be enlarged, whereas for convenience they rather require to be narrowed.

Pirimela, Leach, 1815, like Pilumnus, is represented in Great Britain only by a single species, Pirimela denticulata (Montagu), which occurs also in the Mediterranean, but, unlike Pilumnus, it is not represented by any other species elsewhere. In this genus the pleon of the female is sevenjointed, but that of the male five-jointed, the three middle joints being coalesced into a single piece. It differs from all the rest of the Cyclometopa in the character of the third maxillipeds, for here the fourth joint receives the articulation of the fifth on its inner instead of on its apical margin.

Family 2.-Trapeziidæ.

'Carapace depressed and nearly quadrilateral, smooth, with the postero-lateral angles truncated, the dorsal regions not defined; the antero-lateral margins are straight, form a right angle with the front, and are entire or have but one tooth (the lateral epi-branchial tooth) developed. The