RELATIONS OF YOLK TO GASTRULA IN TELEOSTEANS. 33 
coelomic cavity instead of being obliterated, as Ryder declares 
to be the case in the cod, I am unable to decide. 
SuMMARY. 
The ova of Gadus eglefinus, L., G. morrhua, L., and 
G. merlangus, L., have a perfectly transparent and homo- 
geneous yolk, and a blastoderm of light terra-cotta colour, which 
disappears at the end of the period of simple segmentation. 
The ovum of Trigla gurnardus, L., is larger than those 
of the three species of Gadus described, and is distinguished 
by the presence of a large brown-yellow oil globule, which is 
capable of free motion in the perivitelline space. 
The continuity between the cells of the blastoderm and the 
periblast in its earlier condition can be seen in surface views of 
the living ova of the cod and haddock. 
The invaginated layer of the germinal ring is never con- 
tinuous beneath the segmentation cavity ; it is not continuous 
with the periblast ; it passes to a position beneath the axis of 
the embryo by the concrescence of the edge of the germinal 
ring, and the layer constitutes from its first appearance the 
dorsal hypoblast. 
The floor of the intestine is in all probability derived from 
the periblast. 
The whole edge of the blastoderm represents the ancestral 
blastopore, and the formation of the embryo by concrescence 
is simply the closing of the blastopore from before backwards. 
The edge of the blastoderm in Amphibia, Petromyzon, and 
Ganoids is homologous with the edge of the blastoderm in 
Teleostei. 
There is some doubt whether the primitive blastopore ends 
at the posterior end of the embryo, although all that is known 
concerning the behaviour in Teleosteans favours the conclusion 
that it does so. 
The edge of the blastoderm in Elasmobranchs is not homo- 
logous with the edge in Teleostei; the inflected part in the 
former represents the whole of the latter; the rest of the 
VOL. XXVi.—NEW SER, Cc 
