164 The American Naturalist. (February, i 
Sorex vulgaris.’ The paper is a detailed description of the origin of 
the germ-layers. The earliest stage obtained had a single layer of 
flattened cells (ectoderm) lying beneath the zona and bounding a cen- 
tral cavity filled with fluid. These ectodermal cells he calls the tropho- 
blast. At one point in the periphery there is an accumulation of cells 
—the embryonic knob—which contains the material for the embryonic | 
Se ae ee y ~ 
ectoderm and endoderm. The cavity in the center surrounded by the 
trophoblast and filled with fluid is the segmentation cavity. The em- 
bryonic knob gives rise to the early endoderm cells from its more cen- 
tral part, and some of these then migrate around the periphery of the 
central cavity and apply themselves to the inner side of the trophoblast 
(ectoderm). See Diagram C. This contradicts part of Minot’s 
hypothesis given above.) The trophoblast cells seem to grow over the | 
embryonic knob, causing an “‘ inversion ” of the embryo. After the 
differentiation of the endoderm from the embryonic knob the remain- 
ing ectoderm is spoken of as the embryonic shield (emd.sh.) The 
endoderm first forms part of the notochord and mesoblastic plates. 
Thus under the anterior end of the embryonic shield the endoderm — 
is spoken of as the protochordal plate (mo,ch’.) The rest of the noto- 
chord differentiates later and in a different way. The mesoderm has 
not yet appeared, but is now inaugurated by the appearance of the 
primitive streak. - The mesoderm originates from three different 
points: 1st, from the sides of the protochordal plate (see above) ; 2nd, 
from the primitive streak, from which it advances forward between 
ecto- and endoderm ; and 3d, from an annular zone of endoderm lyiag 
around and under the periphery of the embryonic shield. The details aS a 
of this process are shown in a large number of figures. pee A 
We may now pass to the theoretical considerations of the gastrula- soa 
tion of mammals, (The process of inversion, or the sinking of the oe 
embryo into the cavity of the vesicle, may be left out of account, ai 
produces no important changes in the germ-layers of the embryo, pe ie 
may in a general way be compared with the later formation of " 
amnion.) We have seen in the early differentiation of the endoderm = 
from the embryonic knob that part of the endoderm is formed TA 
the actual process of gastrulation has set in,—that is, before the ap i 
ance of the primitive streak. This the author calls precocious #6 
gation, and is an ontogenetic phênomenon. Later, when the p pat 
streak is formed (the coalescing of the lips of the blastopore), S 
endoderm arises in this region and is added to that already pe gees 
and this latter is the phylogenetic endoderm, and alone 1$ om 
pared to the Sauropsidan type. The remaining part of the 2o 
