THE PRIMARY SEGMENTATION. I 35 



On this point I wish to observe that in all the forms studied, 

 embracing representatives of birds, amphibia, and selachians, 

 the neural segments are among the first anatomical structures 

 to be established ; before the vestiges of any organs have ap- 

 peared, the embryo is divided throughout its length into similar 

 segments. These metameric divisions, therefore, antedate 

 myotomes, branchiae, cranial nerves, or any other structures 

 that exhibit metamerism. They persist through the early 

 stages of development, and become definitely related to seg- 

 mental nerves and segmental sense-organs. In the light of 

 their early appearance and their history, I think we are justi- 

 fied in saying they are the most satisfactory traces of primitive 

 metamerism that are preserved in the group of vertebrates. 



It should also be observed that the entire embryo is seg- 

 mented, and the term " metamerism of the head " should be 

 understood to -signify merely regional metamerism, and not a 

 different kind of segmental division from that occurring in the 

 rest of the embryo. 



The next point to be noted with regard to these segments 

 is that they are formed independently of mesodermic influence. 

 I have shown that the neural segments appear much earlier 

 than those of the mesoderm, and that they extend throughout 

 the embryo ; when, however, the protovertebras appear they 

 are localized, and are formed backwards and forwards from the 

 point of their first appearance. But the final appeal must be 

 made to sections. A careful study of sections of shark and 

 chick embryos shows that the mesoblast is not divided into 

 protovertebrse in the head, even after that region is completely 

 segmented. In the sharks also, the neural folds that are so 

 evidently segmented, are at first wing-like expansions from 

 the body, and during this stage no mesoblast enters into them. 

 Therefore, the neural segments cannot depend upon the seg- 

 mental folds in the mesoblast. The combined facts place the 

 neural segments on a good basis for independent consideration 

 as survivals of primitive segmentation. 



If the brain walls are completely exposed by removing the 

 overlying tissues, we may count the number of neural seg- 

 ments with complete satisfaction. In the brain of shark em- 



