1891.) _ Embryology. 163 
surface, which Van Beneden mistook for the blastopore. . The blasto- 
derm proper consists of ectoderm, several cells in thickness, and below 
scattered endoderm cells. The mesoderm is formed later between the 
FIG. 1.—(Diagram A.) - Fic. 2.—(Diagram B.) 
two. The strong point of this explanation is that it seems to refer the 
germ-layers back to a condition found in the reptilian embryo, and 
the weak point, it seems to me, is that it does not clearly illustrate the 
method by which the yolk has been lost, and what cells originally 
contained it. 
Minot gives the following hypothetical stage to explain the homolo- 
gies of the mammalian germ-layers (see Diagram B). The uei 
central cavity, which he calls the segmentation cavity (the yolk-cavity 
of Haddon), is surrounded by endodermal cells, which formerly con- 
tained yolk. (Hence they do not represent epibolic ectoderm, T 
believed by Haddon.) The walled blastoderm (embryonic knob) at 
the upper pole of the figure is composed probably entirely of ectoderm 
cells. (Again a contradiction to Haddon’s view.) The endoderm 
which later appears under the blastoderm comes from the sides where 
the endoderm cells around the segmentation cavity pass into the ecto- 
dermal blastoderm, Here, it seems to me, is the weakest part of the 
hypothesis, and Dr. Minot seems to have expected to find the oy 
formation of endoderm as in the teleost. This is flatly contradicted 
by well-supported statements. (See Hubrecht below.) The author has 
jumped from the frog’s gastrula to that of the mammal, not giving, I 
believe, due weight to the intermediate reptilian stage, assuming that 
With the loss. of yolk in the ancestral mammal there was a return to 
the more primitive condition of the amphibian stage ; but it seems this 
is hardly a fair assumption as a basis for further hypotheses. ; 
Prof. Hubrecht gives a second paper‘ in his studies Z wee 
embryology, entitled “ The Development of the ss cp 
t Ouar., Jour. Micro. Science, 
