514 LOCY. [Vol. XI. 



segments is more plainly seen. Near the middle part of the 

 embryo the lines of segmentation are faintly traceable from 

 the margins towards the median furrow. The two lines of 

 segments are joined in front by a single median piece or seg- 

 ment. This unsegmented anterior tip becomes more promi- 

 nent in the immediately following stages. There is no evidence 

 to show whether this represents the primitive anterior segment 

 or several aggregated anterior segments. These segments, 

 once established in this very early stage, may be traced onward 

 in an unbroken continuity until they become the neuromeres 

 of other observers, and sustain definite relations to the spinal 

 and cranial nerves. Ryder, in 1881, observed segmental divi- 

 sions extending into the embryonic rim of Elacate, one of the 

 Teleosts. In 1885, he figures such structures in a stage in 

 which the neural groove is closed and the eye vesicles are 

 well established. Although the figure shows a considerably 

 later stage than we are now dealing with, and he does not 

 speak of their earliest origin, nevertheless, the feature of 

 their extending beyond the embryonic axis into the blasto- 

 dermic rim agrees with my observations on Acanthias, and 

 I think it not improbable, that Ryder's segments correspond 

 with those I have described. These segments, observed 

 by Ryder under such unusual conditions, have generally been 

 interpreted by morphologists as due to precocious segmenta- 

 tion in the non-axial mesoderm. The segmentation I have 

 just described is not capable of such interpretation, for sec- 

 tions show that the mesoderm is not yet divided into proto- 

 vertebrae at this stage, and that the epiblast is the seat of the 

 segmental divisions. The mesodermic somites of Squalus are 

 formed later in the usual way, and the first ones appear in the 

 trunk or neck region at a later period. 



In Fig. 26 the embryo is relatively more slender in the trunk 

 region, and there is coming to be an observable distinction 

 between the broadly expanded cephalic plate and the narrower 

 body. Upon the anterior end there is being formed a pro- 

 truding unsegmented median tip, which is much better seen in 

 Figs. 27 and 5. The median furrow ends in front in a broadly 

 expanded depression. The gastrular cavity has become nar- 



