304 



shaped, though we are not prepared to say that such a chauge of 

 shape always takes place. Elongated medullary cells may occur in the 

 wall of the neural tube. In view of the width of the apertures in the 

 external limiting membrane, we believe that elongation of escaping 

 medullary cells is not to be accounted for by compression due to 

 squeezing through a narrow orifice. Along the ventral root to its 

 union with the dorsal root, and beyond, fusiform cells can at this time 

 be seen lying among the fibers that compose the embryonic nerve. 

 Occasionally one of there accompanying cells can be observed under- 

 going mitotic division. 



Examination of sections of a cat embryo has shown that in this 

 animal medullary cells appear to migrate as in the pig. 



Several years ago Schaper ('97) renewed the investigation of the 

 embryonic central nervous system of vertebrates. His painstaking 

 study of the earhest differentiations in the medullary tube of the 

 higher vertebrates showed that the Keimzellen of His — the dividing 

 epithelial cells near the lumen of the tube — produce by mitosis 

 cells to which he gave the name "indifferent". These indifferent cells 

 possess a capacity for locomotion, and a capacity for further pro- 

 pagation by division. Leaving the region in which they were produced, 

 they migrate outward to the mantle layer, and here are transformed 

 either into neuroblasts, which later become nerve-cells (neurones), or 

 into embryonic supporting cells, which later form the neuroglia. 



In these medullary cells escaping from the neural tube we re- 

 cognize the indifferent cells of Schaper. The migrant cells can easily 

 be traced to their source in the mantle layer, where the majority 

 of indifferent cells are to be found. They show a marked capacity 

 for locomotion, and a capacity for further propagation by division. 

 Each cell consists of a nucleus with which is associated only a minute 

 amount of cytoplasm ; there is, as yet, nothing to indicate either its 

 nervous or its supporting nature. The migrant cells appear to have 

 no direct connection with the nerve-fibers among which they lie. 



It, therefore, appears that some of the indifferent cells continue 

 their outward migration beyond the mantle layer. Passing into the 

 niduli of the ventral nerve-roots, they escape from the neural tube 

 along with the peripherally directed processes of the neuroblasts of the 

 niduli, as indicated in the accompanying diagram. 



There are two points to which we wish especially to call attention 

 in this paper. First, it seems to us probable that the migration of 

 medullary cells into the ventral nerve-roots of mammalian embryos is 



