234 s. GOTO. 



posterior eiiterocœl the two remain very distinct, and there is, 

 on the other hand, no ambiguity as to the connection of the 

 sac with the left posterior enterocœh This series and several 

 more, which it is not necessary to reproduce, decides the ques- 

 tion, as it seems to me, beyond doubt. The error of MacBride 

 on tliis point is perhaps to be attributed to the fact that he 

 studied principally by means of frontal sections and also that 

 his specimens were fixed with osmic acid. The former cut the 

 plane of separation of the two structures under consideration 

 obliquely, and the latter makes, as every one knows, subsequent 

 staining a matter of difficulty. The cases of abnormal larvae 

 adduced by MacBride in su^^port of his opinion can not, it 

 seems to me, be set oft' against the observation of normal larvie. 



The subsequent change which the dorsal sac undergoes is, 

 as stated above, different in Asterias and Asterina. In the 

 former it unites with the anterior enterocœl ; in the pi-esent 

 species, however, the sac remains, as already mentioned, entirely 

 distinct from other cavities. It undergoes enlargement in subse- 

 quent stages, and i(s wall becomes consequently very thin. Its 

 definitive position in the star is exactly the same as in Asterias 

 2)all{da. 



In my former paper ['98, p. 25Ö] I criticised MacBride's 

 statement to the effect that the "dorsal sac" was observed by 

 Cuenot in the adult star directly under the madreporite. I 

 then had the impression that the sac marked sm'. ax. in my 

 figures might be something entirely different from the ''dorsal 

 sac "; but my o1)servations on Asterina (jibbosa lead me to 

 conclude that the two structures must be identical. 



The gradual change of position of the epigastric and the 

 secondary left posterior enterocœl takes place in exactly the 



