No. 3.] THE CELL-LINEAGE OF NEREIS. 433 



of the first somatoblast, from the left posterior macromere. 

 With this, according to his statements, the separation of the 

 germ-layers is completed. The first four micromeres give rise 

 to the entire ectoblast, the somatoblast contains the entire mes- 

 oblast, while the four macromeres form the entoblast. I have 

 given, at p. 371, reasons for regarding this account as erroneous. 

 A careful comparison of Gotte's figures with those of Salensky, 

 v. Wistinghausen, and my own, renders it practically certain 

 that Gotte entirely overlooked the formation of the second 

 and third sets of micromeres (the first somatoblast of course 

 excepted), and of the second somatoblast or primary mesoblast, 

 for which he mistook the first somatoblast. It is impossible to 

 believe that the discrepancy is owing to the difference of spe- 

 cies, for reasons already given. 



As regards the later stages, I believe Gotte's interpretation of 

 the axial relations to be essentially correct. His account of the 

 shifting of the axes, of the origin of the dorsal trunk-region, and 

 of the relations of the upper hemisphere of the larva to the adult 

 head, agree nearly with my own. The one important difference 

 relates to the mesoblast-bands, which Gotte describes and fig- 

 ures as horizontal, the mesoblast-axis coinciding with the neural 

 axis and shifting with it. On this point, again, Gotte is almost 

 certainly in error, as shown by the united testimony of Salensky, 

 v. Wistinghausen, and myself. 



Salensky (A^ ciiltrifera, suppressed trochophore) failed to 

 observe the early stages of cleavage. He showed, however, 

 that the ectoblast does not arise simply from the first four 

 micromeres, but is increased by the separation of additional 

 micromeres from the macromeres. He figures with perfect 

 accuracy the two somatoblasts (Fig. 3 ;^., PI. XXHI, "^" = A', 

 "<5'" = M) in their relations to each other and to the blastopore, 

 and they agree precisely with the American species. In the 

 succeeding stage (Fig. 3 n.c), the fission of the somatoblasts 

 takes place exactly ai in our species, the second somatoblast 

 dividing first. The agreement is so precise and detailed that 

 even the cell x^ is shown ("t-" in Salensky's figures) in the same 

 position and dividing at the same time as in N. limbata and A'". 

 megalops. Nevertheless, Salensky did not succeed in establish- 

 ing the connection between these stages and the later ones, and 

 hence, misled by Gotte's statements, he is curiously wide of the 



