398 MARGARET ROBINSON. 



cells in the caudal thickening are many of them much vacuo- 

 lated. 



Putting together the facts concerning these two stages B 

 and B' one is forced to the conclusion that the cells which are 

 iuvaginated at the blastopore give rise to the endoderm. I 

 cannot help thinking that they give rise to that only, and 

 that the mesoderm arises quite independently of this invagi- 

 nation. I think that fig. 6 points strongly to this. That 

 the vitellophags are budded off from the blastoderm figs. 3 

 and 4 show, but unfortunately I have not been able to find 

 in any of the embryos at this stage a blastoderm cell in 

 process of division to form mesoderm, except in fig. 3, div., 

 and the results of these two divisions may be vitellophags 

 after all, 



Bergh (1893) in his account of the formation of the germ 

 layers in My sis describes two stages which almost corre- 

 spond as to age with Stage B described above. They seem 

 both to be a little younger than it, and also a third stage 

 which is perhaps a little younger than Stage B'. He is care- 

 ful to state that in none of these does he find any vitello- 

 phags. In the proliferation in the region of the blastopore 

 he finds two kinds of cells, large and small, the large being 

 mesoderm and the small endoderm. 



In a stage a little older than the oldest of these three 

 Bergh finds some vitellophags for the first time; and in 

 summing up the results of this part of his work he states 

 that the cell mass resulting from the invagination at the 

 blastopore differentiates itself further into three regions of 

 germination (Anlagen). 



(1) The vitellophags, 



(2) The entoderm disc, 



(3) The eight original cells of the mesoderm. 



The likeness between these stages in My sis and those in 

 Nebalia described above is very great; but there are 

 differences. 



(1) I find vitellophags in both Stage B and Stage B', and 

 indeed in an earlier stage than B, i.e. in Stage A, where the 



