TflE DEVELOPMENT OP VERTBBRATA. 235 



the myomeres, while the ventral portions retained their 

 cavities, but lost their metameric character, a continuous 

 ccelom being formed by end-to-end and right-and-left union 

 of their cavities. In the frog a similar condition is arrived 

 at, but by different stages. The mesoblast of each side 

 first splits as a whole ; then it becomes divided into a 

 dorsal portion called the vertebral plate, and a more 

 ventral portion the lateral plate. The lateral plate re- 

 tains its cavity, which presently unites with its fellow of 

 the other side, and is never metamerically segmented; 

 while the vertebral plate becomes metamerically segmented 

 into a series of cubical masses, whose cavities soon disappear. 

 These masses were formerly called protovertebrae, but are 

 now generally known as mesoblastic somites. Neither 

 name satisfactorily expresses their nature " protover- 

 tebrse " is misleading, because, although the vertebras are 

 formed from them, a great deal besides is also formed from 

 them ; while " mesoblastic somite " suggests that the whole 

 of the mesoblast is divided into somites, whereas only the 

 vertebral plate is so divided. 



It must be added that in the head-region (i.e. cor- 

 responding to the first two cerebral vesicles) the mesoblast 

 shows no metamerism, and its splitting is much less complete 

 than behind. 



10. Further Development of Mesenteron. The embryo 

 now begins to lose its spherical shape, and to elongate at its 

 posterior end, though, being still within the vitelline mem- 

 brane, it has to double on itself as it does so. This elongation 

 takes place by growth and division of cells, of course at the 

 expense of the yolk the yolk-cells continually dwindling 

 in size. The most obvious sign of growth, proving that it 

 takes place at the posterior end, is the formation of addi- 

 tional mesoblastic somites behind those first formed ; but of 

 course all the other layers must be increasing equally, 

 though, not being metameric, their growth is less obvious. 



The mesenteron, as it extends forwards, becomes wider 

 and deeper in the anterior region than it is behind, where 

 the yolk-cells remain accumulated on its floor (fig. 121) 

 Thus it is soon divided into two regions an anterior 



