xii DEVELOPMENT 203 



oblongata and cerebellum. It will be seen (Fig. 65) that 

 the medullary groove is lined by ectoderm and that 

 therefore the whole central nervous system is ectodermic 

 in origin ; and this is also true of the nerves. 



A longitudinal thickening of the endoderm on the donsal 

 side of the archenteron becomes constricted off as a solid 

 rod of cells, lying between the medullary cord and the 

 archenteron. This is the noiochord (Figs. 64 K and 65, 

 ncK) ; it forms the primary axial skeleton around which the 

 vertebral column is subsequently developed. The meso- 

 derm is for some time a solid mass occupying all the space 

 between the ectoderm and endoderm, so that there is no 

 body-cavity. But at a later stage a cavity appears divid- 

 ing the mesoderm into two layers, one in contact with the 

 ectoderm {parietal layer, Fig. 65, tiles'^), the other with the 

 endoderm {visceral layer, mes ^). The space between them 

 is the ccelome {c). The dorsal part of the mesoderm on 

 either side of the medullary cord and notochord becomes 

 divided transversely, and gives rise to muscle-segments or 

 tnyoineres (Fig. 64, L). 



The embryo now begins to elongate in a definite direc- 

 tion (Fig. 64, J) : its dorsal surface, still marked by the 

 medullary grove, which now soon becomes closed, is slightly 

 concave, its ventral surface very convex. The blastopore 

 closes up, the yolk-plug becoming entirely covered by ecto- 

 derm ; but for a short time the medullary canal and 

 archenteron still communicate by the neurentcric canal 

 K, H.e.c). This soon disappears, and at the hinder end 

 of the embryo a little conical outgrowth forms the 

 rudiment of the tail (/). The opposite end is rounded, and 

 on its ventral surface appears a little half-moon shaped 

 groove (which afterwards becomes subdivided into two. 

 Fig. 66, C), the rudiment of the sucker by which the tad- 



