96 A. A. W. HUBRECHT:. 
the front and co-operates to bring about favourable conditions 
for respiration.” 
If we try to find out whether among living mammals there 
are such that would yet exemplify transitional stages, as 
they must have existed between such ancestral forms that 
possessed (as do yet the Primates) a more primitive “con- 
nective stalk ” and such that have come to evolve a free 
allantois, we must conclude that such forms are few. 
However, where they are found—among the Rodents and in 
Galeopithecus—we must recognise that we have one of the 
lower orders before us. In the Insectivores, where we might 
have expected to find traces, those that have hitherto been 
examined all possess a free allantois, but then they vary so 
considerably among themselves that we have reason to look 
forward hopefully to those Insectivores whose ontogeny has 
not yet been traced. 
Among Rodents are Cavia and the different mouse genera 
where the allantois offers peculiarities that might here be 
adduced. ‘These cases are, however, the same in which the 
phenomenon of the so-called inversion of the germinal layers 
occurs, and one might be doubtful as to whether this latter 
phenomenon were not rather responsible for the peculiarities 
of the allantois. However, I have attempted (on pp. 74, 
75) to connect this phenomenon with primitive features also 
noticed in the development of invertebrates. "We must for 
the present suspend our judgment and recognise that the 
whole process of inversion has still much that is obscure and 
can on no account be explained by apparently simple 
mechanical explanations as Selenka has attempted (84, 
p- 70). An early circumscribed attachment of the blasto- 
cyst to the maternal uterine wall was in his estimation 
the cause which rationally explained that the embryonic 
shield becomes bent upon itself (entypie). He neglected to 
consider that in a case of very early ‘‘ entypie” as it is found 
in Tupaja and was fully discussed above (p.10, Figs. 29—32), 
the blastocyst is during all these phases perfectly free and 
unattached in the much wider lumen. 
