36 K. MITSUKURI ; 



Woodcut VI. 





Eeconstrncted surface-view of the posterior portion of the embryonic shield. 

 [Enlargement in length is slightly less than enlargement in breadth]. 



epiblast. AVoodciit VI. is such a reconstruction. The heavy dark 

 lines are intended to show the epiblast. It will l3e seen that the mass 

 emliraced within the horse-shoe shaped bl;i.stopore is entirely free from 

 the epiblast. As we go backwards, the surface layer gradually acquires 

 an arrangement like columnar epithelium (see r/) and the thus ar- 

 ranged layer little by little separates itself from the subjacent mesoblast, 

 first in the median line and then by degrees laterally until there is an 

 independent epiblast layer over the entire breadth. Short transverse 

 lines in the posterior part of the yolk-plug in \yoodcut Vl. indicate 

 the extent of the separation of the topmost layer from the mesoblast 

 below. It is natural that the fusion of the upper layer with the meso- 

 blast goes backwards furthest along the groove or lips of the blasto- 

 pore, as the mesoblast is being proliferated here. 



Stages similar to, or preceding, this in Cleiumjis, Trioni/x, and 

 Ghelonia have already been described in Contributions I, HI, and IV. 

 In them, the mesoblast formation will also be found described. 



Series II. 



(PI. IV.) 



Triotiyx japon ieus. 



An enihnjo with 2-3 luesohlastic somites ; taken out 4 days after 

 deposition; surface vieiv given in Fig. 18, Fl. III. {\mß). 



