102 V/YVILLE THOMSOXj ON ASTEllACANTHlOX VIOLACEUS. 



Impregnation seems to take place before the eggs ai'c 

 placed in the pouch. 



I placed a specimen in Avhose marsupium a goodly mass of 

 cggSj sixty to eighty in number, were least advanced, in a 

 separate jar of water, and examined the embryos at first 

 daily, and afterwards at intervals of one or two days, 

 checking my observations on this brood by tlie examination 

 of many other individuals in all stages, the progeny of two 

 or three other mothers of the same species which were 

 bringing up a numerous family in another vessel. 



Segmentation appears to take place in this species in the 

 way usual in the class, and to involve equally the whole yelk. 

 I had no opportunity of observing the earlier stages or of 

 determining the presence of the so-called " directing cells ; 

 these, however, probably exist, as they are very evident in 

 some other Echinoderms. After segmentation the embryonic 

 mass is at first spherical, finely granular, and still invested 

 by the vitelline membrane. The membrane soon disappears, 

 and within a few hours the embryo seems perfectly homo- 

 geneous, regularly oval, and of a delicate flesh colour. I 

 could not detect the slightest trace of cilia on the surface. 

 Four or five hours later the oval form is still more marked, 

 one end has become slightly dilated, and towards this end 

 there is an accumulation of the denser part of the granulai* 

 substance. The whole embryo is now invested by a delicate, 

 structureless, gelatinous layer, which is thinner and less ap- 

 parent towards the narrower and more transparent end of the 

 oval ; at the broader end it invests a dark, consistent gra- 

 nular layer, of consideral)le thickness^ formed of oil-globules 

 and compound granular masses and cells, which lines a central 

 cavity filled with a clearer granular semi-liquid, in which 

 there are traces of molecular or ciliary motion. 



The embryo now becomes club-shaped, and there is a 

 decided aggregation of the great mass of the granular matter 

 to the thick end of the club, whose transparent investing 

 membrane becomes still more distinct, and the internal 

 granular layer thicker. 



The transparent investment of the narrow end protrudes 

 one and then two more tubercles, which rapidly declare 

 themselves as three transparent tubular processes, two turned 

 in one direction, narrow, four or five times longer than their 

 width — the otlier turning in an opposite direction, shorter and 

 thicker, and probably, from its form and its tendency to 

 divide at the extremity, representing a second pair. The 

 investing membrane of tlie tubes is transparent, delicate, 

 transversely wrinkled, and highly contractile. Each tube is 



