185 



very faint or quite indistinct muscular impressions. The radiais are 

 perforated for the nerves. The cartilaginous ring is well developed 

 around the calcareous ring, but very faint and translucent in its 

 posterior part. The perforations through the posterior margin are 

 so large, that there are, as said by Thé el in the description of 

 Synapta picta ("Challenger" Rep. Holoth. II, pag. 10), twelve large 

 "tentacle-canals" which reach from the circular canal to the 

 calcareous ring. The stone-canal is single. In the specimens from 

 Tobago, the small specimens from Bermuda as well as in 0rsted's 

 specimens, there are 2 — 4 polian vesicles. In the large specimens 

 from Bermuda there are 7 — 10 polian vesicles, and in Clark's 

 specimens from Jamaica there are twelve. 



The alimentary canal has usually a large loop on the intestine, 

 but in a single specimen the loop is wanting. In this specimen 

 the alimentary canal is in its whole length fastened by the mid- 

 dorsal mesentery, and in the posteriormost end also by the faintly 

 developed, right, ventral mesentery. The well developed, left, dorsal 

 mesentery is quite free of the intestine. All the three mesenteries 

 are supplied with ciliated funnels of the usual size and shape. 



The gonads are in the specimens from Tobago short and club- 

 shaped and enclose large eggs. The specimens have also embryos 

 in the body-cavity. In most of the other specimens the gonads are 

 long and thin but unbranched, and in the largest specimens there 

 are 1 — 5 short side-branches. 



The shape and size of the calcareous deposits in the specimens 

 from Tobago differ from those of the specimens from Bermuda 

 and Jamaica. In all the specimens at hand the deposits from the 

 anterior end of body are of the same size and shape as the normal 

 ones from the posterior end, but there are here found some anchors 

 and plates which are much smaller than the others. An examination 

 of the embryos in the specimens from Tobago shows that the 

 embryonal anchors measure 80 — 90 ju and the plates ca. 60 \i 

 in length. This agrees excedingly well with the size of the small 

 anchors and plates in the posterior end of the mature specimens, 

 and I suppose that these small anchors and plates really are the 

 embryonal deposits, which have not disappeared. In the same way 

 the small anchors and plates which are found in the posterior end 

 of many of the other species may probably be regarded as embryonal. 



