RELATIONS OF YOLK TO GASTRULA IN TELEOSTEANS. 238 
who considered that the embryo increased in length from the 
beginning by the addition of fresh somites at the posterior 
end, as in Chetopods, that is to say, by the multiplication and 
differentiation of the cells in siti at the posterior end of the 
embryo, no addition being made to them from other parts of 
the ovum. As I have pointed out, if Balfour’s view were 
correct it would be difficult to account for the disappearance of 
the germinal ring in Teleostean embryos. The ring is seen 
to become part of the embryo, and cannot be accounted for in 
any other way. Agassiz and Whitman (1) express their adhe- 
rence to the views of His and Rauber as to the occurrence of 
concrescence without entering into the question of the signi- 
ficance of the process. Ryder (3) is also fully convinced that 
the embryo is lengthened by the concrescence along the neural 
axis of the edge of the blastoderm. 
The quotation from His given by Balfour (loc. cit.) is, so far 
as it goes, completely in accordance with the views in which 
- my own studies have resulted, namely, that in osseous fishes 
the embryo grows together lengthwise from the two symmetri- 
cally-placed halves of the edge of the blastoderm, and that the 
whole edge of the blastoderm is the blastopore, which thus 
extends along the whole dorsal median line; but as no re- 
examination of the subject has been published in this country 
since the rejection of the whole concrescence theory by Balfour, 
I have thought it well to show how inevitably the acceptance of 
the theory follows from a study of Teleostean development in 
pelagic ova; and I also wish to show how certain of the argu- 
ments of His and Rauber concerning the medullary folds in 
Elasmobranchs are irrelevant, at least to the question of the 
significance of the concrescence from the point of view of evo- 
lution, and to examine into certain consequences which follow 
from the facts of Teleostean development, consequences which 
concern the features of development in other vertebrate types, 
and the relation of these features to ancestral history. 
To resume, then: the edge of the blastoderm in Teleostean 
ova is the primitive blastopore, and the embryo is formed, at 
all events to a great extent, by the coalescence of the edges of 
