RANA. 359 
hence the segmentation is total and unequal, the hypoblast half being 
retarded. A modified blastula is, however, still produced. In the 
skate the bulk of yolk is so enormous that only the epiblast cells and a 
portion of the hypoblast can segment. Hence the segmentation is only 
partial or meroblastic, producing, not a true blastula, but a cap or 
blastoderm resting on a mass of unsegmented hypoblast and yolk. 
Gastrula.—The yolk in the frog is already sufficient to prevent the 
normal archiblastic invagination, and as the hypoblast is too bulky to 
be tucked into the epiblast, the epiblast perforce extends round the 
hypoblast, producing finally a small blastopore in the same position as 
that of 4mphioxus, viz., at the postero-dorsal part of the embryo. In 
the skate this is carried still further, and the rim of the epiblast has to 
extend so far round and over the enormous mass of yolk that the 
embryo differentiates during the process. The final result is the same 
as before, the epiblast eventually meeting round a small blastopore at 
the posterior end of the embryo. 
But in both the frog and the skate it is doubtful how far the archen- 
teron is produced by true invagination. Probably in both cases it 
arises mostly by a split in the hypoblast. In the frog the archenteron 
is largely filled by a ventral mass of yolk-cells, but in the skate this is 
*so enormous that it protrudes as a large separate mass of the body. 
Chordula.—The frog embryo, a day or two before hatching, as has 
been seen, can be directly compared with the chordula larva, but here 
again there are modifications. The notochord is not folded off, but 
arises as a solid mass, and the mesoderm no longer arises as a paired 
series of pouches with cavities in continuity with that of the gut, but as 
solid masses, with cavities produced later by splitting. 
In the skate much the same as in the frog occurs, if we consider the 
embryonic portion only. We may note that both the frog and skate 
appear to have at least the main part of the mesoblast formed from two 
posterior sheets or plates comparable to the posterior sacs of Amphioxus 
and of Ascidia. Again, in Amphioxus the whole of these sacs divide 
into mesoblastic somites, the ventral parts of which fuse later to form 
the perivisceral ccelom, the dorsal parts remaining segmented. On the 
other hand, in the skate and frog the ventral part or /ateral plate is never 
segmented, but splits at once to form the perivisceral ccelom, only the 
dorsal part or vertebral plate being segmented, as in Amphioxus. The 
result in all three types is the same though brought about in a different 
manner. 
Foetal Membranes.—In the frog the yolk distends the abdomen, 
but is not sufficient to cause the formation of a complete yolk-sac. In 
the skate, however, the yolk is so abundant that the embryo cannot 
possibly be built up to include the yolk, and the latter has to be held in 
a special sac. The outer wall of this sac is the serous membrane, a 
continuation of the body-wall, and the inner is the yolk-sac proper, 
4 continuation of the gut-wall. Hence we see that in the skate the 
abundance of yolk (and lecithal nutrition) has caused the formation of 
two extra-embryonic fetal membranes, the serous membrane and the 
yolk-sac membrane. In all the Ammzofa not only are these two present, 
but two more, the amnion and the allantois, are superadded. 
