117 
the formation of a comparatively large cavity which, both by reason 
of the time of its appearance and its position, is comparable, in the 
first instance, with the segmentation cavity of the lower vertebrates, 
but which becomes converted into a modified archenteric space by the 
extension of the hypoblast around it. 
With this probability in view the published figures of young 
mammalian ova were carefully examined and it was found that they 
not only failed to furnish a positive proof that the young mammalian 
blastodermic vesicles they represented consisted of an epiblastic wall 
which was lined, over only a small area of its boundary, by a layer 
of hypoblast but that, on the contrary, they were just as well adapted 
to support the conclusion that each of the blastocysts consisted, mainly, 
of a comparatively large hypoplastic vesicle bearing upon one of its 
poles a small mass of epiblast, and that the didermic stage of the 
cysts was completed not by the extension of the hypoblast round the 
inner surface of an epiblastic wall but by the extension of the epiblast 
round the exterior of the hypoblast. 
If this explanation was substantiated the conclusion which naturally 
followed was that in the early stages of development the ova of the 
mammalia did not swerve from the usual path of vertebrate onto- 
genetic evolution, but that they presented all the phenomena which 
might be expected to appear in comparatively large-yolked ova which 
were placed under special nutritive conditions. | 
This conclusion, however, was only supported by direct observation 
upon the ova of rats and mice, for, although the figures of other 
mammilian ova were well adapted to sustain it, the phenomena 
presented by the ova from which they were drawn were interpreted 
in a different manner by the investigators who examined them. Further 
proof was essential before the question could be definitely settled and 
with the object of obtaining the necessary evidence I commenced, 
during the past summer, a series of observations upon the ova of 
the ferret. 
Unfortunately the commencement of this work was unavoidably 
delayed until the breeding season was far advanced and only a few 
ova were secured. These however were in an extremely interesting 
stage of development. 
Before describing them attention must be drawn to the fact that 
the exact age, in hours, cannot be given because the act of coition 
occupies a considerable space of time the animals lying together in 
some cases for a period of 3 hours or more. Further, impregnation 
does not always result from one sexual congress, for out of a total 
