92 BULLETIN OF THE 



But during the stages accompanying this process of folding and 

 immediately following it, the outlines of the inner margins of the cell 

 plates are preserved with such sharpness as to render highly improbable 

 the idea that individual cells are detached from these plates, while at 

 the same time, as P'igures 3 to 5 show, the lower margins of the proto- 

 vertebree are brought into a position at least equally favorable for the 

 production of cells destined to suri-ound the fore-gut. 



Thus the conditions indicate that these cells arise from the protover- 

 tebral plate, and the fact that in a number of cases cells can be seen in a 

 position of actual emergence from the protovertebrse to the intermediate 

 cell-mass proves conclusively that such is their origin. Figures 2 and 

 4, and Figures 12 to 15, give an idea of the relation of these cells to the 

 protovertebrpe, Figures 14 and 15 showing particularly well the connec- 

 tion of the cells with the lower inner margin of the protovertebrse. The 

 older stages of development are perfectly reconcilable with this view, some 

 stages, as that seen in Figure 9, showing very plainly that the connec- 

 tion persists for some time. The increase of the protovertebree in size, 

 and tlie pressure on them from the lateral plates as these approach the 

 median plane, satisfactorily explain the condition from a mechanical point 

 of view. 



From these facts we can, I think, unquestionably conclude that in 

 the heart region the intermediate cell-mass arises from the lower inner 

 margin of the protovertebnx'. 



The mesoderm is undifferentiated in the cod anterior to the ear region, 

 but in support of a mesodermic origin of the lining of the heart I give a 

 figure (Fig. 16) of a section cut obliquely in the eye region of a cod em- 

 bryo, nine and a half days old, which shows a direct connection of the 

 endothelial lining of the heart (here the aorta) with the undifferentiated 

 mesoderm of the head. 



