409 



the parallel folds begin to diverge. This makes an interesting cor- 

 respondence in number between those observed in Squalus acanthias, 

 in the same region, and those occurring in Amblystoma. It is only in 

 Squalus that I have made a connected study of these structure. In 

 the newt, the conditions are very similar and I have photographs of 

 newt's eggs, in very early stages, in which these metameres show 

 clearly. In Torpedo ocellata, I have also noted the occurrence of 

 this metameric segmentation in several stages, beginning with the 

 stage designated " C .", by the Zieglers, and tracing it along in 

 stages with a widely open neural groove. I have not, in this animal, 

 made any study of the number or arrangement of the segments. 



III. Considerations based on the Facts recorded. 



There are some points suggested by the account in the preceding 

 pages that require further consideration. It is evident, from the very 

 early appearance of this segmentation, that we are dealing with a 

 characteristic of fundamental importance. It will be remembered, 

 that the division of the nerve cords into segments, in this way, is 

 one of the earliest phenomona to appear after the embryo is outlined. 

 It is clearly defined before the vestiges of any organs appear, and is, 

 therefore, very primitive, and may be looked upon as a survival of 

 that segmentation, which all agree in assuming for the (not too 

 remote) ancestral form. The cause for its appearance reaches back 

 to the cause for the existence of metamerism, and, whatever hypothesis 

 will explain the occurrence of the latter phenomenon in animals, will 

 also explain this neural segmentation, for, it is merely a survival of 

 a primitively segmented condition of the body. 



It is a point of some interest that this segmentation is primitively 

 epiblastic, and is clearly defined throughout the length of the embryo 

 before the mesoblast has, to any extent, become divided into somites. 

 This fact, will tend to modify a current conception of metamerism 

 which assumes as fundamental that division into metameres begins in 

 the mesoblast, and that the epiblastic segmentation is moulded over 

 it. The point in question is brought out in the following brief 

 quotation from Sedgwick's recent paper *)■ He refers incidentally to 

 metamerism saying: "The segmentation which obviously persists in the 

 trunk region, and which begins with segmentation of the mesoderm, 

 and is moulded upon it in the manner characteristic of all meta- 

 merically segmented animals." But from examination my embryos, 



1) Quart. Journ. Jfic Sei,, June 1892. 



28 



