116 A. A. W. HUBRECHT. 
owe to Owen and others, our knowledge of their ontogeny has 
recently been furthered, in the first place, by Selenka (’87) 
and Hill ’97). And the researches of the latter, that have 
already been alluded to above more than once (p. 3), have 
shattered the old notion that this specialised group of mam- 
mals was intermediate between the Ornithodelphia and the 
Monodelphia. They have furnished most weighty data from 
which we must conclude that—previously to the very quaint 
modifications which have taken place when the growth of the 
foetus was in part transferred from the uterus to the 
marsupium—these animals were more closely related to mono- 
delphian contemporaries than they are now. Most of them 
have now, during the short period of pregnancy, a well- 
developed area vasculosa on the umbilical vesicle, which, 
thanks to a quite extraordinary development of the proamnion 
(Fig. 180), can most efficiently serve as a means of osmotic 
exchange between the foetal blood and the maternal, which 
circulates in deep folds of the uterine mucosa. At the same 
time most of them show an allantois which hes hidden in a 
recess of the umbilical vesicle, and does not in any way come 
to the surface or partake in nutritory exchanges. 
Hill’s researches (’97, 1900) on Perameles and Dasyurus, 
and what Caldwell (’87) found many years ago in Phasco- 
larctos, show that this passiveness of the allantois and 
its ineffective and hidden situation are not the general rule. 
In the genus Perameles the allantois partakes in a very 
effective placentation, histologically corresponding to what 
we observe in the Monodelphia; in Phascolarctos we notice 
a first step in a degenerative direction, the allantois yet 
touching in one circular spot the outer wall of the foetal 
vesicle, but not entering any more into vascularised connec- 
tion with the mother. 
We must now more fully discuss the earliest phenomena 
that are described by Hill for his Perameles blastocysts, 
more particularly as far as the proliferation of the tropho- 
blast is concerned. 
Hill comes to the conclusion (’97)—and Strahl (’06, p. 277) 
