162 bulletin: museum of compaeatiye zoology. 



regard such irregularities of the edge of a rapidly expanding plate of 

 tissue as of morphological importance. It is very significant that the 

 segments show most prominently in the cephalic-plate region just before 

 the edges of the plate begin to rise dorsally, for it is likewise at this 

 stage that I find the first evidence of the disassociation of cells along 

 the edges of the neural crest. Such a disassociation of cells, or even a 

 rapid proliferation of cells, — which certainly does occur in this region, 

 — would lead to such phenomena as those reproduced in Figures 1 and 

 2, Plate 1. An examiuation of cross sections of the cephalic plate 

 (Plate 7, Figs. 55 and 56) before the edges have fused dorsally to form 

 a closed tube shows that the neural crest is already differentiated from 

 the tissue which will form the walls of the neural tube ; it is differen- 

 tiated as a region of rapid cell proliferation and of less compactly 

 arranged nuclei. If the centres of cell proliferation were fixed, then 

 we should have a segmented neural ridge, as affirmed by Beard ('88). 



My interpretation differs from Locy's, since he finds the "neural 

 ridges " segmented regularly, and considers the segments as sur- 

 vivals of an ancestral segmentation ; whereas I find the edges of the 

 neural plate irregularly and somewhat transitorily segmented, the 

 irregularity and inconstancy of the segments precluding, in my opinion, 

 a phylogeuetic interpretation. Locy's results from surface studies seem 

 to me to be a confirmation of those reached by Beard ("88), who, in 

 studying the development of the peripheral nervous system in Selachii, 

 found from the examination of sections that the neural crest is differen- 

 tiated before somites appear, and that it is from the beginning segmented. 

 Beard's conclusions have, however, never been confirmed, and have in- 

 deed been regarded by Dohrn ('90, p. 55) as quite untenable. 



To demonstrate that Locy has not accurately traced the '•' neural seg- 

 ments " onward in unbroken continuity until they become the " neuro- 

 meres of other observers," I propose to discuss the relation of the 

 neuromeres to the posterior limit of the cephalic plate. 



f. Limit of "Cephalic Plate." 

 Locy ('95, p. 543) has stated that in early stages of the embryo, before 

 the neural plate has formed a closed tube, head and trunk may be distin- 

 guished. " It is possible," he says, " in very young stages to draw a line 

 indicating where the expanded part of the cephalic plate joins the non- 

 expanded part of the embryo. . . . This is, in Squalus acanthias, just in 

 front of the point where, subsequently, the vagus nerve begins. ... In 

 this animal, we may identify that part of the head which lies in front of 



