158 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



region, are not equally distinct, it being very difficult, if not impossible, 

 to determine tbe boundaries of some of them. They also differ cousidei'- 

 ably in length and apparently without any regularity, a condition not 

 easily reconciled with the interpretation of them as true segments. It 

 would certainly be impossible even in this specimen to point out with 

 certainty corresponding segments on opposite sides of the cephalic plate. 

 In the trunk region of the same specimen no correspondence between 

 somites and "neural segments" is seen. However, a faint lobing of 

 the inner margin of the tail fold is seen on the right side of the embryo. 

 Locy's ('95, p. 519, Fig. 29) description of a stage as close to this as 

 any figured by him is as follows: "They [the segments] appear like. a 

 row of beads running along the ventrally recurved maj'gin, and extend 

 with great distinctness the entire length of the embryo. Those in the 

 trunk region are continuous with those of the head, and pass into the 

 latter without any transition forms. There is, however, some individual 

 variation in size of the neuromeres, and they are not absolutely sym- 

 metrical on the right and left sides, but the significant thing is, [that] there 

 is uniformlv the same number on each side in a given region, such as the 

 hindhrain, or the bi"ain region as a whole. . . . There seems now to be 

 a natural landmark separating the ' cephalic plate ' from the rest of the 

 embryo ; this is an abrupt downward bending in the medullary folds, 

 wliich, as I have determined, lies just in front of the future origin of 

 the vairus nerve. There are eleven metameres'^ in the lateral margins 

 of the cephalic plate, including the ones embraced in this fold." The 

 accuracy of this conclusion I shall discuss in treating of the question of 

 the limit of the cephalic plate (p. 162). I wish here only to call atten- 

 tion to the fact that none of the reproductions of Locy's photographs, 

 with two possible exceptions (his Figs. 2 and 23), show a segmentation 

 of the neural folds in either the trunk or the embryonic rim. 



If now we turn to Figure 3 (Plate 1), we find an embryo of about 

 the same stage as that shown in Figures 1 and 2 ; at least, it has the 

 same number of somites (6 to 6^). The conditions are these. The " seg- 

 ments " at the margin of the neural plate differ markedly in distinct- 

 ness, and are irregular in size. In the region of the cephalic plate — 

 the posterior boundary of which is marked by the arrow — the number 

 of segments on the right and left sides is not the same. I was not 

 able to assert this with so much confidence in regard to the embryo of 

 Figures 1 and 2, since in that embryo the limits of the cephalic plate 

 were less clearly defined. If the segments of the two sides of the neu- 



1 Italics mine. 



