700 PROCEEDINGS OF THE ACADEMY OF [Oct., 



With the formation of 4d and 3b the segregation of the germ layers 

 now consists of 29 cells distributed as follows : 



{ First quartette, 12 cells. 



Ectoblast \ Second quartette 8 " 



[ Third quartette, '. 4 " 



Mesoblast . Fourth quartette (4d), 1 " 



Entoblast . Entomeres. 4 " 



Total, 29 " 



Thus in Dinophilus another example is already added to a long list 

 of forms in which the ectoblast arises from the first three quartettes 

 of micromeres, the mesoblast from the left posterior cell of the fourth 

 quartette, and the entoblast from the remaining cells. To this mode 

 of origin of the germ layers apparently the cephalopods alone form 

 an exception. Among the annelids it is at present known to obtain 



A B 



Fig. Ill, illustrating the reduction of the cleavage cavity. A, optical section of 

 stage of about 54 cells; B, 72 cells. Cleavage cavity stippled. 



only in the groups of the Polychaeta and Echiuridce. The fine paper 

 of Eisig (1898) on Capitella seems to indicate that this manner of 

 derivation of the germ layers does not extend to the Capitellidse, 

 since in Capitella Eisig derives the permanent mesoblast ("coelo- 

 mesoblast") from cells of the third quartette (3c^ and 3d^), while 

 4d gives rise to larval or secondary mesoblast ("paedomesoblast") 

 and ectoderm. However, in view of the probable tendency of the ova 

 of Capitella to abnormal development, it is, I think, permissible to 

 doubt these somewhat surprising results until they are verified in the 

 same or some allied form. These results seem all the more surprising 

 in view of the great resemblance which, in other respects, the cleavage 

 bears to that of the Polychaeta, Among the Echiuridse, Torrey (1902, 

 1903) has found that Thalassema in its cleavage and in its mode of 

 origin of the germ layers agrees closely with the Polychseta. 



