242 K. MITSÜKUKI AND C. ISHIKA^YA 



and 29). This fact will, we think, answer Kolliker's objection, bated 

 upon the shape of cells in the mesoblast, against the epithelial 

 origin of the mesoblast. (We have not access to Kolliker's original 

 paper but take his views as given in Hetrwig's paper, No. G, p. 105). 

 Kölliker is no doubt correct in supposing that such forms are due 

 to very rapid proliferation. 



As to the formation of the chorda, it is only necessary to com- 

 pare our figs. 25 — 28 with Hertwig's figs. 3 — 6 (Taf. iii) of Triton, 

 and figs. 8 — 11 (Taf. viii) of Rana, in order to be convinced of the 

 similarity of the process in Reptilia and Amphibia. Our figs. 25 

 and 29 correspond with figs. 1 or 2 (Taf. iii, Hei'twig) of Triton. 

 Our fig. 26 with fig. 4 (Taf. iii) of Triton, our fig. 27 with fig. 5 

 (Triton), and, finally, our fig. 28 with fig. 6 (Triton). 



As to the contribution to the mesoblast from the germinal wall, 

 there is of course no equivalent in the holoblastic egg of Amphioxus 

 or Amphibia. It seems to us that phylogenetically this source is 

 not of much significance and is brought about wholly by adaptation. 

 Sarasin's (No. 15) researches on the Reptilian egg have brought out 

 the fact that new cells are added on from the yolk to the blastoderm 

 by a process very similar to budding. Why could we not suppose 

 that this process goes on until considerably later, and that the addi- 

 tion of cells to the mesoblast from the germinal wall is but the 

 continuation of this process ? 



We should like to add another suggestion. In Trionyx the 

 primitive streak is continuous with the lateral edges of the blasto- 

 pore, enclosing the yolk-plug (see fig. 6). Have we not here a case 

 where a part of the original blastopore lips has met in the median 

 line and formed the primitive streak, while the rest of the edge of 

 the blastopore has retained its original condition? 



We think we have succeeded in showing that the development 



