Diptychus, A. Milne-Edwards, 1880, the name Diptychus being pre-occupied. Of the species Uroptychus insignis, Henderson, and Uroptychus gracilimănus, Henderson, figures are given on Plate VII. In this genus the second antennæ have on the second (the first free) joint of the peduncle an acicle, thus differing from all the rest of the family, except Eumunida, in which one is also present, though of very small size. Dr. Henderson says: In those species which I have examined, the fifth arthrobranchia, counting from before backwards, is not of larger size than the others, whereas in most of the Galathodea it is distinctly enlarged.' According to M. Jules Bonnier, in the species Uroptychus rubrovittatus (A. Milne-Edwards) there are no arthrobranchiæ, their places being taken by a corresponding number of pleuro-branchia-that is, by branchiæ inserted on the pleura or sides of the segments, instead of being placed on the articulating membranes that unite the appendages to the segments. In this genus, and in Ptychogaster, A. Milne-Edwards, 1880, and in Eumunida, there is a comparative weakness of the swimming-fan, which probably for that reason is twice folded on itself. The members of these genera being sometimes found in the branches of Gorgoniæ, it is conjectured that they lead a sedentary life, that the swimming-fan is in consequence losing its importance, and an advance is thus being made towards the brachyuran type. Ptychogaster MilneEdwardsi, Henderson (see Plate VII.), from Patagonia, has the pleon, except the telson and uropods, covered with rows of short stout spines.

The larval development in species of the genera Lithodes, Eupagurus, Anapagurus, Munidopsis, Galathea, Munida, and Porcellana, has been carefully studied by G. O. Sars, and his results confirm from this point of view the close union of the legions to which these genera respectively belong, as well as the propriety of including the whole group among the Macrura. From Sars' work on this subject have been borrowed the figures grouped together on Plate VIII., representing the end of the pleon in the last larval stage respectively of Lithodes maia

b

Pl. viii.

End of Pleon in the last larval stage a. Lithodes maia (Linn.). b. Eupagurus Bernhardus (Linn.). c. Porcellana longicornis (Linn.). d. Galathea intermedia, Lilljeborg

e. Munida rugosa (Fabricius)

PERILS OF BABY-FARMING

179

(Linn.), Eupagurus Bernhardus (Linn.), Porcellana longicornis (Linn.), Galathea intermedia, Lilljeborg, and Munida rugosa (Fabricius).

It will be remembered that the larval forms of Crustacea often show not the smallest resemblance to the adults, and also that it is extremely difficult to breed in confinement those which pass through many transformations from the egg to the perfect condition. In this interval at every moult, the shedding of the skin is attended. with danger. The movement, the saltness, the temperature, of the water in which they are kept should correspond with the conditions they would have experienced in their proper marine home. When all this has been suitably cared for, the supply of appropriate food must be considered, and this will probably not be the same for all the stages. When several specimens occupy a common nursery the more advanced are very apt to destroy the less forward, and the creatures being as a rule very small there are risks of confusions, the larvæ of distinct species being perhaps mistaken for the stages of one and the same species. In calm weather, and, at least, in the latitude of Great Britain, especially, though not exclusively, after dark in the months of August and September, it is easy in many parts of the sea by means of a surface-net to obtain an abundance of larval forms, but there is a great difficulty in determining the species and genera to which they severally belong Being minded to connect together the several stages of a crustacean's life with some approach to certainty, Professor Sars made a practice of carefully drawing and dissecting the forms he met with, and then by comparison of a long series he was able in numerous instances to assign them in proper sequence to species of which the adults were already known. Considering the vast numbers of the Crustacea, it is evident that there is valuable work to be done on these lines, enough to occupy a crowd of zoologists for many years to come, and the plan is available for many who have no access to the constantly improving resources of the modern scientific aquarium or marine biological station.

rostrum.

CHAPTER XII

TRIBE II.--THALASSINIDEA

THE carapace is short and compressed, with little or no The last segment of the trunk is articulated with the preceding. The eye-stalks are small. Both pairs of antennæ have long peduncles. Of the trunk-legs the first pair are perfectly or imperfectly chelate, the last pair are short, more or less abnormal, directed backwards. The pleon has the segments not overlapping, with the side-plates feebly developed and having their hinder angles generally rounded. The pleopods are long, biramous, variable; the swimming fan is strong. The branchiæ vary in number and form.

It

In this tribe are included four families, the Thalassinidæ, Callianassidæ, Axiidæ, and Thaumastochelida. The division of the Macrura adopted by the late Mr. Spence Bate, into Trichobranchiata, in which the branchial plumes are made up of long cylindrical filaments, Phyllobranchiata, in which the plumes are formed by a series of foliaceous plates, and Dendrobranchiata, in which the branches of the various plumes divide and subdivide in an arborescent manner, does not seem practically very convenient. has been already seen that two families so intimately allied as the Pagurida and Parapagurida would have to be placed, the former in the Phyllobranchiate, the latter in the Trichobranchiate, division. But also in the present tribe Spence Bate himself points out a weakness in the arrangement, for of the genus Callianassa he says:- The structure of the branchiæ of Callianassa is so intermediate in character that it may be claimed by anatomists as be