RELATIONS OF YOLK TO GASTRULA IN TELEOSTEANS. 21 
wall of the latter. I have no additional evidence to bring 
forward; my recent observations have confirmed the view I 
adopted concerning Kupffer’s vesicle, but the formation of the 
intestine in the region anterior to that vesicle still requires 
investigation. 
It is worthy of remark that the figures and description of 
the origin of the intestine in the salmon given by Oellacher (11) 
are not in any way inconsistent with the view for which I am 
contending, provided it be remembered that the multiplica- 
tion of nuclei in the periblast probably takes place faster than 
the consumption of such nuclei in the formation of the floor 
of the intestine. 
The above reasoning leads, then, to the conclusion that the 
invaginated ring is the dorsal hypoblast, and is therefore 
homologous with the dorsal hypoblast in other vertebrate 
types, with the invaginated layer beneath the embryonic 
rudiment in Elasmobranchs and Amphibians. In the Tele- 
ostean there is no wide cavity separating dorsal hypoblast from 
yolk as in those two types, the only representative of such a 
cavity is Kupffer’s vesicle. 
The next question to be considered concerns the method by 
which the invaginated layer passes, during the growth of the 
blastoderm, over the yolk from its original position to its 
final place beneath the embryonic rudiment, and the embryonic 
ring is absorbed into the dorsal region of the embryo. The 
first change is part of the second. The main features of the 
process have already been pointed out, but they must now be 
examined more closely. In following the process in the living 
ovum it seems perfectly certain that the centre of the blasto- 
derm (#) at the stage shown in fig. 5 is a fixed point in relation 
to the yolk, while the non-embryonic portion of the germinal 
ring slides bodily over the surface of the yolk. There is no 
doubt a multiplication of cells going on in all parts of the 
blastoderm during this period of the development as in all 
others, but this does not account completely for the growth of 
the embryonic rudiment if its relation to the germinal ring be 
carefully considered. After the blastoderm has passed the 
