THE GASTEULA STAGE. 55 



of fluid may perhaps be one of the factors that determine or aid 

 the process of invagination. It seems more probable, however, 

 that invagination is due rather to inequality in the rates of 

 growth of the cells at different parts, than to direct pressure 

 from any cause on the surface of the embryo. 



The later stages in the development of the gastrula show some 

 features of importance. At its earliest formation, as shown in 

 Fig. 16, the axis of the gastrula coincides with that of the blastula; 

 and the blastopore'or gastrula mouth is circular in outline. 

 Later on, as shown by the careful observations of Hatschek, owing 

 to unequal rates of growth in different directions, the blastopore 

 becomes oval instead of circular in outline, and the shape of the 

 embryo changes (Fig. 18) in such way that the axis of the gastrula 

 no longer coincides with the original axis of the blastula, but 

 forms a considerable angle with this. 



At the stage shown in Fig. 18 there may be seen at the 

 lower lip of the blastopore, and placed one on each side of the 

 median plane, a pair of cells, PC, which differ from the other 

 hjrpoblast cells in their larger size and more rounded form, 

 and in having very large nuclei. T hese two cells give r ise 

 at a later stage to important portions of the middle germinal 

 la yer or mesoblas t : they may be named the p olar mesoblas t 

 cells. 



The further stages in the completion of the gastrula will be 

 understood from a comparison of Figs. 18, 19, and 20. The 

 embryo elongates, becoming ovoid or egg-shaped : at the same 

 time the blastopore becomes still further reduced in size ; the 

 n arrowing being effected, acc ording to Hatschek, e ntirely by 

 gro wth backwards of its anterior lip , t he posterior li p, indicated 

 by the pair of polar mesoblast cells, remaining quiescent through- 

 out, the process. 



In the fully formed gastrula (Figs. 19 and 20), the ends and 

 surfaces of the larva may be clearly recognised. The pola r 

 m esoblast cells, p C. m ark t he posterior end of the embryo ; the 

 blastopore, B P, now reduced to a small circular aperture, lies at 

 the hinder end of the embryo, and slightly on the dorsal surface- 

 The anterior end of the embryo is rounded and imperforate. 

 The dorsal surface is flattened, an d is f urther indicated by th e 

 blastogorej whil e the ventral surface is strongly conve x. 



If Hatschek is right in stating that the narrowing of the 



