318 G. CARL HUBER 
half of the ninth day, contained stages which are younger than 
nearly all of those obtained the latter half of the eighth day. 
I am unable to state whether this is owing to a retardation in 
the rate of development of the ova in rats Nos. 17 and 35, or 
due to an error of record. The record gives date and hour of 
insemination and of killing, and I have no reason to doubt its 
accuracy. However, the two rats in question give the only 
instances of marked deviation from what appears as a normal 
rate of development as presented by the bulk of my material. 
Sobotta (11) has called attention to the difficulty of obtaining 
successively staged material in the mouse, and cites Kolster as 
contending: ““Man kénne auf die Altersbestimmung gar nichts 
geben.’’ During this stage of development the decidual erypts 
lodging the ova are deeper than in the preceding stage, their 
mesometrial portion being narrower, though they are not as 
yet separated from the uterine lumen. The orientation of 
the decidual crypts and the contained egg-cylinders is perhaps 
more readily made than in slightly younger stages, though not 
definitely enough to insure the cutting of sections in a given 
plane. Sections of the egg-cylinder cut in the longitudinal 
plane may be obtained by cutting parallel to the plane of the 
mesometrium or at right angles to the same. However, it is 
still largely a matter of chance as to whether the sections ob- 
tained pass through the midplane or at an angle thereto. 
In figure 25, there are reproduced representative sections of 
three germinal vesicles taken from the same uterus (rat No. 35, 
8 days, 18 hours) which show three closely approximated early 
stages in the development of the egg-cylinder. Noneof these three 
vesicles is cut in exactly the mid-longitudinal plane; especially is 
this true of the ends of the vesicles. Furthermore, the antimes- 
ometrial portion of each, lower part of the figure, composed of 
the thin-walled parietal ectoderm, shows a certain amount of 
folding, so that a portion of each wall is cut en face instead of 
en profile. The appearances here presented by the antimesome- 
trial portion of these vesicles is not to be confused with a ‘giant 
cell’ formation of this portion of the roof of the vesicle, described 
by Sobotta in his earlier publications, but corrected and retracted 
