ON THE GASTRULATION IX PETROMYZON. 13 



approached tlie dorsal blastoporic lip, since in order to add such 

 a considerable extent to the embryonic shield, a great deal of 

 the opaque hemisphere must invaginate within the ovum. But as 

 seen in the figures referred to, this hemisphere in fact suffers but 

 little change of extent, except that it is more or less depressed on 

 the surface. We are thus led to conclude that the marked 

 translocation of the groove is mainly brought about by the exten- 

 sion of tlie embryonic shield produced by the bachward shiftinr/ 

 of the translucent animal layer. The opaque hemisphere has only 

 little, if at all, to do with the invagination or with the reduction 

 of the segmentation cavity. Thus, it is obvious that the marked 

 translocation of the groove is an apparent one : its real shifting 

 due to the actual invagination of the opaque hemisphere is probably 

 very little. As will be shown farther on, this translucent layer, 

 shifting backwards, partly contributes to the growth of the 

 blastoporic lip and is partly turned inwards. 



In this way, the translucent part containing the segmentation 

 cavity, is, in a still further advanced stage, reduced into a small 

 vesicle and comes to be situated somewhat in the ventral 

 aspect of the ovum {Figs. 9a) ; the boundary groove around 

 it is accordingly seen also on the ventral aspect. Finally with 

 tlie thorough obliteration of the segmentation ca\âty, the groove 

 is, of course, lost from sight {Fig. 10a). Now the ovum assumes 

 a new outline ; it is swollen in the anterior part and decreases 

 in bulk posteriorly, while, in a little younger stage, it is rather 

 bigger in the posterior part with the ventral part of the yolk-hemi- 

 sphere still bulging out {Fig. 9a). The ovum is now pear-shaped 

 {Fig. 10a). 



Standing thus in intimate connection with the gastrula 

 invagination, the boundary groove marks precisely the extent in 



