VIVIPAROUS FISHES OF THE PACIFIC COAST. 
429 
slightly deeper than the other cells. The cells of this layer are the only ones which 
cover the yolk nucleus, leaving but a minute naked portion, the blastopore (fig. 36). 
They are more distinct in some sections than in others, but are everywhere quite 
evident. The axes of the remainder of the cells are more alike. These cells do 
not extend much, if any, beyond the margins of the yolk. The plane of the sections 
of this egg is meridian, probably but slightly inclined to the sagittal. The lower 
layer of cells is well separated from the upper layer. At the median portion the 
blastoderm is still many (6 to 7) cells deep and the separation between ectoderm and 
entoderm is not yet so distinct. The cells, however, which will form the entoderm 
are smaller than those overlying them. 
Figure 38 represents a section of a gastrula in all probability cut in a sagittal 
plane. The yolk in this section is pear-shaped, the narrow' end being composed of 
the yolk nucleus. The latter is composed of a more granular refringent portion 
and a more uniform part resembling the protoplasm of the cells. The cells are 
grouped into a one- to three-cell-deep ectoderm and a one-cell-deep entoderm which 
becomes two and then several cells thick towards the dorsal edge of the blastopore. 
The outer layer of the ectoderm is flattened in places; the inner cells are rather 
loosely connected. On the upper part the entoderm cells lie close against the yolk 
and form a continuous series. The division in this case between the entoderm and 
ectoderm is well marked by a space which was probably produced by the reagents 
(osmic, chromic, and acetic strong mixture). Near the blastopore, in what would 
correspond to the embryonic ring in other teleosts, the two layers merge more or less. 
Figs. 41 and 42 represent two parallel sections from another egg of the same 
ovary, cut transversely or in a vertical plane at right angles to the sagittal. All these 
eggs are in the same stage of development. In the eggs represented in fig. 40 the 
yolk nucleus consists of an irregularly biconvex portion, whose peripheral part 
stains deeper, and which contains some bodies which stain less and a more uniform 
part forming the matrix of the former. The cells, as in fig. 38, are divided into 
an ectodermic layer and an entodermic layer. Some of the outer layer of cells of 
the ectoderm have become greatly distended, probably by absorption of the intra- 
ovarian fluid, which fills the breathing chamber and stains as these distended cells do. 
The line separating the entoderm from the ectoderm is quite distinct, but along the 
median portion, or at the axis of the embryo, the entoderm is several cells thick. The 
same is indicated in the horizontal sections. The shape of the yolk is different in the 
three eggs, but they agree in all essential respects, and figs. 38 and 41 must represent 
a normal stage in the development of Cymatogaster. The question now arises as to 
how this gastrula has been formed from the condition represented in fig. 33. 
The fact that there is a marginal ingrowth of cells to form the lower layers in 
teleosts generally has been observed by so many authors, and is so evident from the 
fact that the space between the inner margins of the embryonic ring is reduced in 
some cases, notably Stolejjhorus, that it is beyond dispute. Does the method of the 
formation of the gastrula in Cymatogaster form an exception to the rule? It must be 
borne in mind that the changes undergone between the stages represented by figs. 
33, 35, and 38, all take place during the time occupied by the latter half of the 
eleventh segmentation and the first half of the twelfth. During this time, if the 
usual course is followed, all but the outermost layer of the blastoderm must be 
