1891.] Embryology. 671 
tation, those over one pole being very large, and are, as shown by their 
fate, the endoderm cells, while those at the other pole go to form the 
ectoderm. Both germ-layers are filled with large yolk masses. The 
ectoderm cells partially surround the endoderm cells. Where ecto- 
derm and endoderm come in contact around the periphery of the large, 
open blastopore there is found a ring of small ectoderm cells, which go 
to form the nerve-chord of the older larva, It is thus seen that the 
nerve-chord is formed from two bilateral parts lying along each side of 
the blastopore, then subsequently coming into contact form the nerve 
plate. Before the blastopore is closed in, however, the large endo- 
derm cells, which are still at the surface (within the rim of the ecto- 
derm forming the blastopore lips) delaminate into a row of outer, 
smaller cells—the endodermal plate—and larger cells in the interior of 
the embryo. From the former there develops, at the sides of the 
blastopore, the peristomial mesoderm. Later, as said above, the lips 
of the blastopore close over the endodermal plate, and the nerve-chord 
- is formed out of the cells from the two sides of the blastopore meeting 
over the endodermal plate. J is thus seen that the blastopore corre- 
sponds to the dorsal side of the embryo. In other words, the animal pole 
of the eggs, where the ectoderm first forms, corresponds to the ventral 
side of the Ascidian, and by inference to the ventral side of all Chor- 
data. (Van Beneden had previously pointed out this fact, which is of 
the greatest importance, since it bears directly upon two of the most 
interesting problems of embryology,—viz., the relation of vertebrates 
to other groups, and the polar relations of the egg to the adult ani- 
mal.) There is no invaginate gastrula in Distaplia, andt he cavity of 
the digestive tract appears later as a split in the endoderm. After a 
review of gastrula within the group, the author concludes as follows : 
“ Segmentation in the so/fary Ascidians is nearly equal, and leads to 
a one-layered blastula, of which one-half flattens, then invaginates. 
By this means there is formed a gastrula which comes nearest of all 
Ascidians to the primitive type,—7.e., to a Archigastrula. In the social 
Ascidians a modification is brought about in that the blastula is not 
formed. With the disappearance of the latter the segmentation- 
Cavity is reduced to a split between the embryonic cells, or fails pont 
pletely. The result of segmentation is a two-layered plakula, into 
which the elements of the two germ-layers differentiate quite early,— 
at the eight-celled stage. The gastrula is here formed not by invagi- 
nation of blastoderm cells, but by a splitting in the endoderm, while the 
borders (periphery) of the plakula rise up and grow towards one 
another,—a process that is brought about by unequal growth (increase} 
