62 STR C. WrVFLLR THOMSON ON THE MODE OF 



young in their early stages are exposed by removing the plates 

 (fig. 3). At first,whenthere are only morales orvery young embryos 

 in the crypts, the marsupium is barely raised above the general sur- 

 face of the perisom, and the plates of the marsupium fit accurately 

 to one another ; but as the embryos increase in size, the marsupium 

 projects more and more, and at length the joints between the plates 

 begin to open (fig. 2), and finally they open sufficiently to allow 

 the escape of the young. The young in one marsupium seem to 

 be all nearly of an age. In P. epJiippifer the marsupium occupies 

 the greater part of the dorsal surface, and its passages run close 

 up to the edge of the mouth, so that the eggs pass into them at 

 once from the ovarial opening without exposure. 



In the male there is, of course, no regular marsupium ; but the 

 plates are arranged in the middle of the back somewhat as they 

 are in the female, except that they are not raised upon peduncles, 

 so that it is not easy at once to distinguish a male from an infe- 

 cund female. 



Although we have taten species of Psolus sometimes in great 

 abundance in various parts of the world, particularly in high lati- 

 tudes, southern and northern, I have never observed this peculiar 

 form of the reproductive process except on this one occasion. 



. II. EOHINOIDEA DESMOSTICHA. 



Among the marine animals which we dredged from the steam- 

 pinnace on the 19th of January, 1874, at depths of from 50 to 70 

 fathoms in Balfour Bay (a fine recess of one of the many channels 

 which separate the forelands and islands at the head of Eoyal 

 Sound, Kerguelen Land), there were several examples of a small 

 Oidaris, which I will name provisionally O. nutrix. As, how- 

 ever, in the case of Psolus epMppifer, I do not feel by any means 

 certain that this is a distinct species. It comes certainly very 

 near to some of the smaller varieties of C.papillata ; but as it pre- 

 sents differences which serve at once to distinguish it, and as its 

 peculiar mode of reproduction may perhaps be regarded as in itself 

 a character of specific value, I give it in the mean time the benefit 

 of the doubt. 



0. nutrioo (fig. 4) resembles G. papillata in the general form 

 and arrangement of the plates of the corona, in the form and ar- 

 rangement of the primary tubercles of the interambulacral areas 

 and of the secondary tubercles over the general surface of the 

 test, in the form of the plates of the apical disk and of the im- 



