RELATIONS OF YOLK TO GASTRULA IN TELEOSTEANS. 17 
invaginated layer beneath the axis of the embryonic rudiment 
increases at the expense of the invaginated layer of the 
embryonic ring, the rest of which goes to form the external 
part of the embryonic rudiment. The whole of the embryonic 
ring thus belongs to, and is formed into, the dorsal region of 
the embryo. These phenomena are discussed at greater length 
in a subsequent section. The segmentation cavity is seen at 
the stages represented in fig. 6 and 7, both superficially, and 
in optical section at the side of the ovum opposite to the 
embryo. At the stage shown in fig. 8 it is no louger visible in 
section, though the line s s marking the extent of the in- 
vaginated layer is still distinct. In fig. 9 the line ss is not 
seen, and at this stage the segmentation cavity is not visible. 
As far as can be judged from my observations the segmenta- 
tion cavity is obliterated at the stage represented in fig. 8 by 
its epiblastic roof coming into contact with the periblast. 
Before the yolk is completely enveloped by the blastoderm the 
differentiation of organs has begun to take place in the embryo. 
The cerebral part of the eye is visible at the stage shown in 
fig. 7, and at the time the envelopment is complete the separa- 
tion of the mesoblastic somites has commenced. It will be 
seen that, on the supposition that the anterior end of the 
embryonic rudiment is a fixed point in relation to the yolk, 
the embryonic side of the blastoderm at first grows faster 
than the opposite side; the distance from 2 to z in figs. 6, 7 is 
less than the distance from x to y. In fig. 8 the contrary is 
the case. After a certain time the growth in length of the 
embryo takes place more slowly, and at the time when the 
embryo is completely enveloped the embryo in the haddock 
and whiting occupies half a meridian of the yolk. In the 
herring ovum (fig. 25) the embryo at the corresponding stage 
is longer in proportion to the yolk. The herring ovum is far 
from being as transparent as are pelagic ova, and in conse- 
quence of this in the former the relations of the internal parts 
cannot be distinguished in the living condition; thus in fig. 23 
the limits of the segmentation cavity and invaginated layer are 
not distinct. From my sections of ova at the stage shown in 
VOL. XXVI.—NEW SER. B 
