THE EARLY DEVELOPMENT OF THE FROG 109 



region just in front of it, show that the rudiments of the chorda, 

 mesoderm, and dorsal endoderm are for a time not distinguish- 

 able (Fig. 33, E). Sections through the narrowed blastopore, 

 while it is still filled with the yolk plug, show that the rim of the 

 blastopore is composed of a thick undifferentiated mass of cells, 

 representing a part of the contracted germ ring. Farther 

 laterally the ectoderm is separated by a line or narrow space 

 which is formed in gastrulation ; the thin endoderm is separated 

 from the middle layer by a line or space which results from 

 delamination as described above. In front of the blastopore, 

 in the region formed by confluence, the arrangement of the 

 cells and germ layers is much the same, except that they are not 

 interrupted in the mid-line. The pigmentation of the inner 

 surface of the mass is an indication of the derivation of these 

 cells from the outer layer through invagination. 



At a later stage, when the yolk plug has withdrawn from the 

 surface and the blastopore has become slit-like, transverse 

 sections show several important changes in this axial mass. 

 In front of the blastopore the separation between ectoderm and 

 mesoderm has extended, by delamination, toward the mid- 

 line, and just before reaching this, turns sharply downward 

 toward the line of delamination between the endoderm and 

 mesoderm, not, however, reaching quite to this, thus leaving in 

 the mid-line a narrow vertical ridge of cells. In the regions 

 where the endoderm and mesoderm remain continuous, a pair 

 of slight depressions appear as shallow grooves out of the 

 archenteron; these are continued into the cell mass a short 

 distance as virtual grooves, indicated only by the arrangement 

 of the pigmented cells. Further forward (Fig. 36) the lower 

 margins of these grooves become better marked, as low lip-like 

 structures approaching the mid-line, and the mesoderm in 

 these regions is more extensively separated from the lining of 

 the archenteron. Finally, still farther forward, the grooves 

 disappear and the extension vertically of the spaces separating 

 the mesoderm from the endoderm and ectoderm completely 

 delimits the pair of mesoderm masses. The cells left in the 

 mid-line between the proximal ends of the mesoderm sheets 



