PROPAGATION OF CEETAIN ECHINODEEMS. 77 



1874, a large number of specimens of a species of the genus 

 Ophiocoma of which I can find no description ; and I accordingly 

 name it provisionally 0. didelphys (fig. 13). "We afterwards met 

 with the same species in other localities among the fiords and bays 

 of Kerguelen. 



The disk is about 20 millims. in diameter ; and the arms are 

 four times the diameter of the disk in length. The disk is uni- 

 formly coarsely granulated ; the arm-shields, which are well-de- 

 fined through the membrane, are rounded in form and roughly 

 granulated like the remainder of the disk. The character which 

 at once distinguishes this species from all the others of the genus 

 is, that the normal number of the arms is seven instead of five, 

 whicli is almost universal in the class. The number of arms is 

 subject to certain variation. I have seen from six to nine, but 

 never fewer than six. The arm-spines are numerous and long. 

 Tlie general colour of the disk and arms is a dull greenish 

 brown. 



A large proportion of the mature females, if not all of them, 

 had a group of from three to ten or twelve young ones clinging to 

 the upper surface of the disk by their arms : the largest of these 

 were about a quarter the size of their mother ; and they graduated 

 down in size until the smallest had a diameter of less than 1'5 

 millim. across the disk. The largest and oldest of the progeny 

 were always uppermost, furthest from that dist, the series de- 

 creasing in size downwards, and the supply evidently coming 

 from the genital clefts beneath. In several specimens which I 

 examined, although by no means in all, there were groups of eggs 

 and of young in still earlier stages, free in the body-cavity in the 

 interbrachial spaces. 



It thus seems that in this case the true " marsupium " is a por- 

 tion of the body-cavity, and that the protection afforded by it is 

 supplemented by the attachment of the young to the surface of the 

 disk, maintained for some time after their extrusion or escape. 



The process of propagation in Ojphiocoma didelphys differs from 

 most of the other cases described in the eggs being successively 

 hatched, and the young being found consequently in a regularly 

 graduated series of stages of growth. Although I had not an op- 

 portunity of working the matter out with the care and complete- 

 ness I could have wished, I feel satisfied, from the examination of 

 several of the young at a very early period, that in this case no 



