6 E. W. MacBKIDE and J. 0. SIMPSON. 



rudimentary." They " attached themselves by the tentacular feet of the trivial 

 ambulacra, which are early and fully developed." He says, however, that they " were 

 too late at the Falklands (January 23rd) to see the process of attachment of the 

 young in their nursery .... There can be little doubt that, according to the analogy 

 of the class, the eggs are impregnated either in the ovarial tube or immediately after 

 their extrusion, that the first developmental stages are run through rapidly, and the 

 young are passed back from the ovarial opening, which is at the side of the mouth, 

 along the dorsal ambulacra, and arranged in their places by the automatic action of 

 the ambulacral tentacles themselves." (Vol. II., p. 215 et seq.) 



Now the young brought back by the ' Discovery,' with two exceptions, all 

 measured from 2| mm. to 3 mm. in length, and had been born in the interval 

 between capture and preservation ; it seems, therefore, that we had here a key to the 

 solution of the manner in which the early developmental stages are passed over, and 

 that we might thus get a more complete history of the " brood-care " of this species. 



A careful examination of the adult female specimen, made with this object in view, 

 showed two prominent openings, each about 3 mm. in diameter, situated in the dorsal 

 interambulacrum, a short distance behind the peristome (Fig. 4). Looking into these 

 openings, one could see a number of embryos of about the same size as the new-born 

 specimens of the collection. At first sight these openings seemed to lead into the body- 

 cavity, but a careful dissection revealed the fact that they were in reality the openings 

 of two large brood-pouches situated in the dorsal body-wall (Fig. 5). Owing to the 

 extreme contraction of the specimen, the extent of these pouches can be made out only 

 approximately. The right, which is by far the larger of the two, extends from the 

 mid-dorsal line downwards to the right ventral interambulacrum, and from a point just 

 behind the peristome backwards for about one-third of the body length. The left sac, 

 however, is only about half as large. Both pouches are divided up by a number of 

 pillars representing the podia, which thus seem to traverse the sacs. In reality, 

 however, the podia are probably only surrounded by the pouches as they develop. 

 Some idea of the extent of these spaces may be gained from the figure, which is drawn 

 of the natural size. No communication was found between the two sacs, nor was any 

 opening into the body-cavity or genital ducts made out. 



In these two pouches 140 embryos were found — 110 in the right and 30 in the 

 left. These, however, are all of the same size, and of a comparatively late stage of 

 development, with the exception of two specimens which are much less developed than 

 the rest. 



Through the kindness of Professor F. Jeffrey Bell we have also been able to examine 

 three other specimens of C. crocea belonging to the British Museum. It was hoped 

 that they might show some trace of these brood-pouches, but the most careful search 

 revealed nothing. 



If we now try, from the facts before us, to get a connected idea of the " maternal 

 care " of this species, we shall probably find that its course is somewhat as follows : — 



