232 s. GOTO. 



apposition and subsequent transverse separation of the two perito- 

 neal walls, as shown in figs. 3 & 8. In fig. 7 h we see the 

 primary mesentery on tlie ventral side still present, and at a 

 short distance from it on the right side, a new septum is being 

 formed. In Ästerias pallida I was not able to see these two 

 septa in a similar condition, but in the present species the process 

 of formation of the one as well as that of disappearance of the 

 otlier can be followed step by step in many sections, one of whicli 

 I have also reproduced in my preliminary paper on this subject 

 ['98 Ô]; and no doubt can, therefore, be entertained as to the exact 

 extent of the two posterior enterocœls. The cavity formed by the 

 union of the whole left, with a portion of the right, posterior 

 enterocœl I have called the secondary left posterior enterocœl, 

 and the cavity left at the postero-dorsal corner is the epigastric 

 enterocod; the latter being, as I have stated in my preliminary 

 paper, very nearly one-half of the original right posterior entero- 

 cœl. The formation and completion of the secondary mesentery 

 between the epigastric and the secondary left posterior enterocœl 

 proceeds from before backwards, as in Asterias pallida. 



We have then at this stage two cavities entirely separated 

 from oacli other ; one is the; comparatively small epigastric ente- 

 rocœl, and the other is the anterior enterocœl and the secondary 

 left posterior enterocœl freely communicating with each other. 

 In the next stage (E) the dorsal mesentery dividing thé two 

 portions of the secondary left posterior enterocœl is absorbed, 

 and both come to communicate with the anterior enterocœl {cf. 

 fig. 11 d& e and fig. 12 d & e). The secondary left posterior 

 enterocœl has therefore, at this stage, two passages of communi- 

 cation with the anterior enterocœl, one on tlio right, and the 

 other on the dorsal, side, 



