MUSEUM OF COMPARATIVE ZOOLOGY. 85 



hand plate already shows the begimiiug of a double row of cells. Figures 

 7 and 7* show the ends of the plates to be two or three rows thick. The 

 cells lying between the edges of the approaching plates are seen to be 

 arranged somewhat in rows parallel to these edges, this being especially 

 true close to the plates. This condition might lead one to think that 

 the thickening of the plates is due to additions received from these cells. 

 The fact that in the younger stage (Fig. 6) there are almost twice as 

 many cells between the ends of the lateral plates as in the later stage 

 (Fig. 8), might also tend to support the idea that these cells add them- 

 selves to the lateral plates ; but it is clear that these loose cells might be 

 pushed by the approaching plates into a position between the lateral 

 plates below and the gut above, and thus caused to extend toward the 

 protovertebrse ; in fact, cells unquestionably from this source can be seen 

 in this position between the gut and the plates in Figures 6 and 8. In 

 my opinion there is no ground for maintaining that the edges of the two 

 latei'al plates receive accessions from the cells that lie between them, for 

 the outlines of the plates can always be distinguished as sharp bounda- 

 ries separating the cells of the plate from those that lie between them. 

 The increase in the number of cells in the plates comes exclusively from 

 the division of the cells of the plates themselves. 



At six and a half days, as shown in Figure 8, the plates are much 

 closer together, and a tendency begins to be seen for the plate of the 

 right side of the embryo (also right in the figure) to push upwards 

 towai-d the gut and past the end of the left plate, which is at the same 

 time pushing downwards. 



This condition is continued to the seventh day (Plate III. Fig. 9), 

 when the edges of the plates are seen to overlap each other for a con- 

 siderable distance. 



Up to this stage of development the histological condition of the cod 

 is fairly serviceable for study, but after this the elements become very 

 closely packed, the different cell layers and masses being almost indis- 

 tinguishable. Their study, therefoi-e, becomes very difficult. However, 

 the histological conditions of the older stages of development are not of 

 so much importance in the question of the origin of the heart as are 

 those of the younger stages. I have therefore been able to make out all 

 that is necessary for the present investigation without concerning myself 

 with the histological details of later stages. I could trace the develop- 

 ment accurately enough to show the continued approach of the thickened 

 edges of the plates up to the time of their union and the complete estab- 

 lishment of the heart. Figure 10, Plate III., from a cod embryo of seven 



