564 Stewart Paton 



Occasionally one uotices a cell witliin the cord throwiug- out a pro- 

 cess which tapers rapidly as it approaches the base of the bridge, 

 and either before or soon after entering this structure loses its iden- 

 tity in the surrounding mass of protoplasm. In a later stage, re- 

 jiresented by Torpedo embryos of 7 nini, in length, the relation of 

 the librils to the cell processes and nuclei may be studied to ad- 

 vantage. A representation of these relations is given in Fig. 5. 

 Two nuclei are visible with very faintly stained processes [b) and 

 [d] which may be followed in the direction of the ventral roots (c). 

 A sraall conuecting brauch (e) seems to unite the fibrils in one pro- 

 cess with those in another. An important point to be noted is that 

 no process distinct from the surrounding matrix can be distinguished 

 at the inner pole of the nuclei. The thin tapering spirai processes 

 capping the nucleus and running in the direction of the ventral 

 roots are the only definite outlines iudicating cell boundaries that 

 exist in these sections. At [a] is a single tìbril which has crept 

 past the nucleus and entered the field where cell boundaries, if they 

 exist, are indistinguishable. In Fig. 8 of the same Piate these con- 

 ditions are even clearer, although the embryo from which the sections 

 were preparedwas only 6 mm. long. This difference in the length serves 

 to emphasize the fact that the period marking the beginning of 

 neurofibrillation as well as the rate at which it proceeds varies 

 considerably even in embryos of the same species. 



At (c) in Fig. 8 delicate fibrils may be seen lying either upon 

 or at the side of the nucleus, while at [D] two coarse and several 

 finer fibrils contained in the common envelope [N] diverge from the 

 others, passing in the direction of the two nuclei at the extreme 

 right in the drawing, to skirt the edge of the cord. Again [b') two 

 fibrils running quite parallel to each other are represented passing 

 a nucleus in the outer layer aud continuing their course as far as 

 one in the seeond row. 



In Fig. 9, Piate 23 there is a drawing representing a section of 

 the cord with the ventral roots in one of the early stages of Lacerta 

 muralis. The intense staining of the fibrils and the fact that they are 

 of a more uniform thickness and generally coarser than is the case in 

 Selachians is a prominent characteristic. These structural variations 

 in different species deserve closer study than has yet been given to them. 



The neurofibrils within the bridges and the triangulär area in 

 the cord begin to multiply quite rapidly (Figs. 8 and 9, Piate 23) 

 and at the same time single fibrils or bundles begin to appear in 



