180 THE VOYAGE OF H.M.S. CHALLENGER. 



t 



the nature of this cavity having been as yet published, it is not possible to compare the 

 two modes of carrying the young in these two genera more closely, 



" In Spatangoids, with deeply sunken ambulacra, we find, nearly in all cases, that 

 from the sharp edge of the ambulacral groove, long spines extend, so as nearly to close 

 the opening of the cavity, entirely bridging it over, and completely conceahng from view 

 the ambulacral pores. This arrangement has usually been considered in Spatangoids 

 as a sort of filter to keep foreign particles from affecting the delicate water tubes, which 

 in the Spatangoids perform more or less the function of gills. This is undoubtedly the 

 case in several genera, but in the case of Hemiaster, and perhaps in other allied genera, 

 the sunken ambulacral area is used for an entirely different purpose, as was correctly 

 observed by Phdippi — that of sheltering the young. 



" That the many specimens (eight) found in the two posterior sunken ambulacral 

 areas are really the young of Hemiaster, is, of course, only probable, from the fact that the 

 genital o^^enings, which are unusually large, open directly into the upper part of their 

 sunken area ; so that the eggs (or more properly an imperfectly developed pluteus, like that 

 of Echinaster) on escaping from the genital openings would readdy find their way into 

 the artificial cavity formed by the spines which conceal the presence of the sunken areas. 



" Unlike many Echinids, the ovaries of this genus are small, consisting of compact 

 grape-like clusters of eggs, in very difi'erent stages of development, a few of the eggs only 

 attaining a considerable size (nearly 1 mm.), and apparently ready to escape into the 

 sunken area as soon as the place should be left unoccupied by the preceding brood. No 

 two of the small Echinids were in the same stage of development ; they varied in size from 

 2 mm. to 3 mm., the smaller specimens having a somewhat pentagonal outline, with 

 rounded angles ; the larger were more nearly elliptical and cylindrical in shape. In the 

 smaller specimens (PL XX.* fig. 7) the spines were short, straight ; the longest, and only 

 a few in each interambulacral area, about one-fifth the length of the axis, while the greater 

 number were mere tubercles, scarcely rising above the level of the test. In the largest 

 specimens (PI. XX.* fig. 6) many of the spines, nearly equalling the radius of the test, 

 had become curved and had assumed the characteristic appearance of Spatangoid spines. 

 Seen from below (PL XX." fig. 10) the large angular mouth, covered by a thick 

 membrane, was nearly central, somewhat anterior, the edge of the mouth on the level of 

 the test, and a few small indistinct pores arranged in parallel lines, showing the position 

 of the future actinal petal ; the ambulacral areas were occupied by coarse granulation, 

 while the tubercles of the interambulacral spaces were large with well-developed crenu- 

 lation, and already perforated. The interambulacral areas were already broad, leaving 

 but narrow ambulacral spaces, in which the short, club-shaped ambulacral tubes could 

 wh difiiculty be traced ; they were largest near the apex, and near the actinostome. 

 Seen from above (PL XX.* fig. 9), the most marked feature of all these young Echinids 

 was the broad fasciole, occuppng so large a part of the abactinal surface, the position of 



