400 WILSON. [Vot. III. 
form, and arrange themselves in a perfectly flat plate, their 
nuclei remaining near the peripheral ends. The smaller cells 
of the upper pole become greatly flattened and spread out, so as 
to form a pavement epithelium covering the plate of larger cells 
above and on the sides. The primary mesoblasts lie side by 
side near one end, closely surrounded by the outer cells, and 
from each a somewhat irregular row of cells extends downwards 
and forwards. The cleavage cavity is much reduced, but does 
not disappear until a little later. 
The embryo is now bilaterally symmetrical, with well-defined 
dorso-ventral and posterior axes, and, except for the precocious 
development of the mesoblast, is closely similar to the ‘“ Pla- 
kula”’ of Biitschli. The germ-layers are perfectly differentiated, 
the entoblast occupying the entire ventral surface, the ecto- 
blast covering the sides and dorsal surface, while the mesoblast 
is arranged in two diverging rows that lie near the sides of the 
embryo and meet each other in the median line behind. Series 
of sections through such embryos (see Figs. 45, 46), as well 
as optical sections of entire embryos (Figs. 35, 42), show in 
the clearest manner that the mesoblastic rows end abruptly 
in front at the sides, without extending across the median line. 
At a much later period they grow forward and join above the 
mouth, this forward growth being effected in the main by an 
actual migration of the cells, and in no case by proliferation 
from either ectoblast or entoblast. 
The embryo now becomes concave on its lower side, so as 
to form a large blastopore occupying the entire ventral aspect 
(Fig. 35). By the folding in of its sides this blastopore then 
assumes the form of a longitudinal slit (Fig. 37), and rapidly 
closes from behind forward, the anterior part remaining open as 
the mouth (Figs. 36, 42). During these changes the anterior 
lip of the blastopore undergoes no change; the general history 
of the blastopore is, therefore, closely similar to that Ewpomatus, 
as described by Hatschek (No. 23). The anus is not formed 
until the end of larval life.? 
1 There is a considerable variation in the closure of the blastopore, owing to differ- 
ences in the rate of folding between the sides and the posterior margin of the blasto- 
pore. Asa rule the sides fold in more rapidly than the hinder lip, thus giving rise to 
a slit-like blastopore, but in some cases the reverse is true, so that the blastopore 
never appears as a slit, but always as a rounded opening. 
