8 J. T. CUNNINGHAM. 
especially close to the edge of the blastoderm where the layer 
is thickest. Study of the living ova does not enable one to see 
the segmentation cavity or any separation between the centre 
of the blastoderm and the yolk at the early stages now under 
consideration, but there can be no doubt that Agassiz and 
Whitman have given the true history of the origin of the peri- 
blast. The history of the relations of the yolk to the blastoderm 
during the period of segmentation may from their results, and 
from what can be seen in living pelagic ova, be read as follows: 
—When the blastodisc is formed it is not completely separated 
from the yolk, although its limits appear to be so distinct, it 
remains continuous with protoplasm in the yolk-mass, proto- 
plasm of which the greater part forms a thin superficial layer 
extending from the edge of the blastodisc over the surface of 
the yolk. When the blastodisc divides into two parts, each of 
these is to be considered as being continuous with half the 
yolk through the yolk protoplasm. When the blastoderm con- 
sists of sixteen cells the twelve peripheral at least of these are 
still continuous with the yolk protoplasm. The cells at the edge 
of the blastoderm remain during segmentation continuous with 
the yolk protoplasm, and at all stages the yolk-mass may be 
considered as potentially divided into a number of portions, 
each belonging to one of these cells. 
As is seen in fig. 12, during the process of segmentation the 
division of a nucleus at the edge of the blastoderm is accom- 
panied by the formation of one plane of cell division parallel 
to the edge of the blastoderm between the outer daughter 
nucleus and the blastoderm, and of other planes perpendicular 
to the edge of the blastoderm, while there is no plane of division 
between the outer daughter nucleus and the yolk. When the 
stage at which invagination begins is approached the formation 
of planes of division perpendicular to the edge of the blasto- 
derm ceases ; thus the outer row of nuclei are separated by cell- 
division planes from the rest of the blastoderm, but not from 
one another. After this these nuclei continue to divide, but 
no planes of division are formed to divide the protoplasm in 
which they are situated into corresponding cells. These nuclei 
