112 ANATOMY AND DEVELOPMENT OF THE ECHINODERMS 



longish sac, in which, as in the tube, there is a ciliary motion. The 

 sac Has behind, at the side of the oesophagus (Diag. IX. [Plate 12]). 



Soon after the appearance of these parts, a hyaline mass, in which 

 very small cells are imbedded, is seen lying like a mantle upon the 

 dorsal side of the stomach. 



The sac becomes developed into a rosette of five caeca, the first 

 foundation of the water-vascular system. 



The mantle-like mass curves over and covers in the stomach and 

 foundation of the tentacles like a cap, widely open below. The dorsal 

 pore becomes invested by it, and it extends round the anus ; but the 

 oesophagus remains outside it (Diag. XL [Plate 12]). 



A crest or elevation now appears on the mantle-like mass, and 

 runs obliquely over it in a curved line, whose ends become eventually 

 united. It then forms the margin of the starfish. 



What lies beneath this thickened margin belongs to the dorsum of 



o o 



the starfish, what lies above it to its ventral surface. 



The }'oung starfish now attains a diameter of ith of a line, becomes 

 slightly pentagonal, and retains only a narrow connexion with the 

 Bipinnaria. 



The digestive canal, and with it the rosette-like rudiment of the 

 water-vascular system, becomes turned so as to present the latter 

 towards the ventral surface of the starfish, at that point where its 

 mouth is subsequently formed. The tube which connected the rosette 

 with the pore, which is now imbedded in the dorsal surface of the star- 

 fish, receives a calcareous deposit and becomes the sand-canal, while 

 the " pore " is converted into the madreporic plate. 



The oesophagus of the larva is obliterated, whilst its rectum projects 

 as an anal tube subcentrally from the dorsal surface of the starfish 

 (Diag. XIII. [Plate 12]). 



The slightest touch now separates the starfish from the larva in 

 which it was developed ; the former sinks to the bottom and creeps b)- 

 the aid of its newly-developed feet ; the latter swims about as before 

 for some time, but eventually perishes. 



In the Echinidse the process is essentially the same. An in- 

 ternal diverticulum of the integument of the larva is formed, but 

 from a somewhat different spot,^ namely in front of the ciliated 



^ It is remarkable that in the Asterid-larvae, while the development of accessory ciliary 

 processes, iic. takes place in the pre-trochal dorsal region, the bud of the Echinoderm is 

 developed from the post-trochal region. In the Echinus-larvs we have just the reverse — 

 the bud is developed from the pre-trochal region ("below the lateral arch of the ciliated 

 band," Miiller), while the processes, &c. , as we have seen, are developed from the post- 

 trochal region. The Ophiurre appear to present the same relations as the Echinidre, though 

 Prof. Miiller has not been able to make out the point with certainty. 



