22 



AKT. 11. — S. HATTA. 



other macromeric parts occupying the largest ventral portion of 

 the ovum (see Fig. 23). While these characteristics disapj^ear 

 rather abruptly in the micromeres forming the columnar 

 epithelium, there is ventrally, in this respect, never any distinc- 

 tion between this mass of cells and the remaining macromeric 

 part. In addition to this, there occur active cell multiplication, 

 — a fact which shows that the cells are not derived from other 

 parts, but are formed here. Compare Fig. 14 with Fig. 15. 

 Notwithstanding the fact that the eminence has grown much 

 more conspicuously in the latter section than in the former, 

 there can be detected no indication of downward shifting of any 

 cellular elements which may contribute to the growth of the 

 eminence. After all, it is plain that the cells forming the 

 conical eminence are the macromeres which have here multiplied 

 in situ, not '' scheinbar " but in reality. 



Having made out the early steps of the gastrula invagina- 

 tion and the relations to the parts with which it stands in 

 connection, I will turn to the explanation of further changes 

 going on in different parts : the changes in reference to a) the 

 difïërentiation of the micromeric elements forming the roof of 

 the segmentation cavity, b) the invagination of the macromeric 

 and micromeric parts, and c) the translocation of the boundary 

 groove and the obliteration of the segmentation cavity in conse- 

 quence of the progressing invagination. 



a) The epithelial structure of the micromeric layer, which 

 is in early stages confined to a small area, is by stages extend- 

 ed farther and farther {Figs. 10-22). Already at a stage when 

 tlie segmentation cavity is reduced into a small vesicle {Figs. Sa 

 and 85), the differentiation almost reajches, at least on the dorsal 

 median part, the entire roof of the segmentation cavity {Fig. Vô). 



