160 bulletin: museum of comparative ZOOLOCxY. 



the mesodermic somites in that region." In the bead region, where 

 somites do not similarly press upon the neural plate, it still remains for 

 Locy to show how structures morphologically dorsal, as his "neural 

 segments" are, become converted into structures morphologically 

 ventral as well as dorsal, as are the "hindbrain neuromeres," for 

 example. 



My own evidence of their continuity in time is, as I have said, nega- 

 tive. Figure 4, Plate 2, represents an early stage with three or four 

 somites. One sees the " marginal bands " of which Locy has spoken, 

 but only the faintest traces of segments are visible. On one side — 

 the right — they are exceedingly irregular. At this stage the lateral 

 edges of the neural plate are not flexed ventrally, and puch segments as 

 are to be seen at all show best from the dorsal side. A quite regular 

 segmentation is seen on the left side of the cephalic plate, yet the seg- 

 ments are by no means all of the same size or distinctness, nor do they 

 equal in size the mesodermal segments. In the trunk region the lobes 

 of the edge of the neural plate show no definite relation to the meso- 

 dermal somites, the boundary between two somites coinciding in some 

 cases with the depression between lobes, in others with the apices or 

 with other parts of the lobes. I wish to call especial attention to the 

 fact that here, as in the embryos shown on Plate 1, the segments are con- 

 fined to the marginal hands, and therefore do not extend into the median 

 plate. Here, again, there is a considerable discrepancy between Locy's 

 observations and my own. 



I have found it impossible to trace definite segments into the later 

 stages, for in these stages, before the closure of the neural tnbe, in the 

 majority of specimens little or no evidence of segments along the cephalic 

 plate can be seen. 



Two embryos in later stages of development are seen in Figures 5 and 

 6, Plate 2. There is practically no evidence of segmentation or lobing 

 of the edge of the medullary folds. The segments which Locy has 

 numbered 1, 2, and 3 are visible in many specimens, in some very dis- 

 tinctlv, as shown in his photographs; but behind them there is an 

 irre(^ularly sinuous or entirely smooth edge, as shown in my Figures 5 

 and 6, and in Locy's photographic reproductions. These three anterior 

 segments, according to Locy, shift their position. Since, however, 1 

 do not find them constant in appearance and position, I have not been 

 able to regard them as of morphological importance. It is worthy of 

 note that they appear in the region of the neuropore, and that possibly 

 they may be partly accounted for as the result of the difficulty of fusion 



