430 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
inflected at the margin of the blastopore and must meet somewhere below the ecto- 
derm ; and, by a further comparison of figs. 33 and 41 , it becomes evident that the 
bulk of the cells of the median portion of the blastoderm of fig. 33 must migrate 
toward the margin before the infolding can produce the conditions seen in fig. 38. If 
the formation of the gastrula of Cymatogaster follows the usual method, fig. 38 is 
derived from fig. 33, after the manner just described, granting, of course, that both 
figs. 33 and 38 represent normal conditions. Considering, however, the great rapidity 
with which this process takes place, and examining fig. 41 in connection with fig. 38, 
it seems probable that the infolding process may be slurred over, and the lower layer, 
or primitive entoderm, be formed by a process of delamination.* 
In this connection the stage represented in fig. 35 becomes interesting. If this 
stage be compared with the one represented in fig. 33 the great probability that the 
condition represented in fig. 35 is derived by simple delamination from fig. 33 becomes 
quite evident. 
The shape of fig. 33 is retained by fig. 35, which can readily be derived from the 
former by further division and a slight progress of the epidermal layer over the yolk 
nucleus. But in fig. 35 the entoderm is already well separated from the ectoderm 
near the margin, and the division, while not so distinct at the median portion, is evi- 
dent both as a line and in the size of the cells. While, then, figs. 41 and 38 make it 
seem possible that the lower layer is formed by delamination, fig. 35 makes it certain 
that it is so formed. 
Now, as to the changes in fig. 35 necessary to produce fig. 41. We see in the 
fig. 35 the beginning of the constriction of the yolk at the margin of the entoderm 
which gives the yolk in fig. 41 its peculiar shape. The margin of entoderm has 
advanced toward the entodermic pole at a uniform rate on all sides while the epidermal 
layer of cells has been apparently passive. In fact, the blastopore in fig. 41 is larger 
than in figs. 35 and 36, but the margins in fi g. 41 do not correspond to the margin of the 
epidermal blastopore of figs. 35 and 36, but to the shaded portion of fig. 36. The epider- 
mal layer in fig. 41 is not everywhere evident, and where it is most conspicuous the cells 
are distended and feebly stained. This is probably due to intracellular digestion, as 
was mentioned above, but it may be due to the degeneration of this layer of cells. 
Certain it is that in this stage the epidermal layer is not distinguishable near the 
blastopore, where it is so evident in stage 35. 
In the stages immediately succeeding those represented in fig. 41 the yolk nucleus 
disappears and the yolk assumes a more or less spherical outline. Its fate will be 
discussed later. With its disappearance, the “ fixed point” souseful in orientation 
is lost. 
From the fact that the entoderm is formed by delamination and the anterior end 
of the embryo lies at one margin of the blastopore and the posterior end at theother ( ?), 
the fundamental difference between the gastrula of Cymatogaster and that of teleosts 
in general may be gathered. The embryo of teleosts in general is formed by the con- 
crescence of the embryonic ring from before backward, and the embryo rarely encircles 
more than half of the yolk. ( Clupea , a genus with a small yolk, forms a notable exception.) 
* Wilson, p. 220, says : “ It is, to be sure, very doubtful whether there is any vertebrate in which 
the primitive hypoblast is really delaminated from the upper layer.” 
