DEVELOPMENT OF THE TWO PRIMARY GERM-LAYERS. 



95 



a 

 *1 

 | | 

 M : ~ 



IS 



especially behind (hi} the crescentic groove, free nuclei are constantly 

 to be found keeping up the supplementary cleavage. 



Owing to the appearance of the new fissure (subgerminal cavity) 

 (fig. 53 ud), the cleavage- cavity (fig. 51 fh) is almost completely 

 obliterated. The two cell-layers of the 

 blastula-stage (fig. 51 dw, vw), described as 

 lying one above and one below the cleavage- 

 cavity, have come close together (figs. 53 

 and 54), being separated from each other 

 by only a narrow fissure. In the upper 

 layer (ak) the cells have assumed a cubical, 

 and at a somewhat later stage a cylindrical, 

 form, and constitute a compact epithelial 

 membrane. The lower layer (ik) is composed 

 of larger roundish and loosely arranged cells 

 in several layers. The former is the primary 

 outer germ -layer, the latter the inner layer. 

 In the region of the posterior marginal 

 ridge (vl), where the cells are at the same 

 time engaged in more active proliferation, 

 the two layers are continuous with each 

 other. 



The highly important processes, by means 

 of which are produced the conditions repre- 

 sented in figs. 53 and 54, present many points 

 of comparison with the gastrulation of the 

 Selachians and Amphibia. We can conceive 

 that the newly appearing fissure has arisen, 

 as in the case of the germ-disc of Pristiurus 

 (fig. 50), by an infolding, in such a way that, 

 as in the former case, cells grow inward from 

 the posterior marginal ridge ; and that at 

 the same time, at the de< -p part of the iu- 

 vagination, the cells which are originally 

 continuous with the yolk (fig. 53 dk} detach 

 themselves from the latter, and are employed for the increase of the 

 inner germ layer. 



If this explanation is correct, the fissure (ltd) which now exists be- 

 tween the inner germ-layer and the floor of the yolk corresponds to 

 the coelenteron, as GOETTE and RAUBER have already remarked, and 

 as DUVAL has for the first time demonstrated ; moreover, the cres- 



<*- 



