80 



SEITARO GOTO. 



Right Anterior Ent'c'l.... Right Ant. Ent'c'l. "| 



\ Ant. Ent'c'l. = Axial Sinus. 

 T a a * • v m J Left Ant. Ent'c'l. j 

 Left Anterior Ent c 1 j Hydw»c«I. 



ii* ii t» j. t. f Epigastric Ent'c'l. 



H.ght Posterior Ent c'l.... | j » Right Pogt EntVI . 



>• Secondary Left ") 



f Left Post. Ent'c'l J Post. Ent'c'l. V Hypogast rlc En t'c'l. 



Left Posterior Ent'c'l -j Periœsophageal Ent'c'l J 



I. Dorsal Sac. 



As an explanation of this diagram the following may be added. The two 

 anterior enterocœls unite at a very early stage, and the hydrocœl is formed from 

 the left enterocœl at about the same stage. The united anterior cavities remain 

 in the adult animal as the axial sinus. The two posterior enterocœls unite on the 

 ventral side, but very soon about one-half of the l ight posterior enterocoel is cut 

 oh" from the rest by the formation of a secondary septum; this is the epigastric 

 enterocoel. The remaining single cavity is what I have called the secondary left 

 posterior enterocoel, and is destined to form a large portion of the hypogastric 

 enterocoel. The periœsophageal enterocœl is budded out from the left posterior 

 enterocoel after its union with the right. The dorsal sac also arises from the left 

 posterior enterocœl, but at a very early stage. The cavity whose names are 

 printed in full-face are those that persist in the adult star. 



Turning now to Asterina gibbosa we see that the epigastric enterocœl is form- 

 ed by a precisely similar process as in Asteria*. MacBride (2) calls it " right 

 posterior" enterocœl, as does Bury for Bipinnaria aderigem (ß) ; but careful ob- 

 servation shows that it is something quite different from the right posterior entero- 

 cœl of the larva. It is true that on the right side there is no natural boundary 

 line between the anterior and posterior enterocœls, but I think there is n > valid 

 objection to the criterion I adopted in my paper on Asteria", viz. to regard the 

 transverse plane passing through the pore-canal as the boundary line betwee n the 

 two enterocœls, the anterior and posterior. With this criterion in mind, if we look 

 on the right side of a larva of Asterina gibbosa in about stage D of MacBride it is 

 very plain that the mesentery separating the " right posterior " enterocœl of Mac- 

 Bride runs very obliquely from the postero-ventral corner of the larva anteriorly 

 towards the pore-canal. This obliquity of the course of the mesentery was ob- 

 served long ago by Ludwig, and is correctly represented in several of his figures 

 (4, figs. 31 and 32). Again, the accompanying cut, which represents the vent- 

 ral portion of a transverse section through the region under consideration of a 



