EARLY ONTOGENETIC PHENOMENA IN MAMMALS. 87 
A strongly developed vascular network on the umbilical 
vesicle is also found in many Didelphia (Fig. 130), where the 
vascular area is undoubtedly in a high degree of nutritive 
significance during the short stay of the blastocyst in the 
uterus. 
In the Ornithodelphia viviparity has been replaced by 
oviparity. There is no voluminous albumen layer, and the 
yolk-laden blastocyst is narrowly enclosed in the egg shell, a 
detail which renders the safe handling of the early em- 
bryonic shield exceedingly difficult. I believe this mero- 
blastic arrangement in the Ornithodelphia to have been 
preceded by viviparity and by a relation of the spacious 
trophoblast to the formative ectoderm such as it was 
described and discussed above, the evolution of the allantois 
having come about during this viviparous phase. Perhaps 
we may yet regard the aberrant way in which the area 
vasculosa spreads over the yolk sac as a primitive character. 
Instead of being restricted to a cireular region, the blood- 
vessels of Echidna invade the total surface of the umbilical 
vesicle (here: yolk sac) although they do not form so dense 
(Semon, *94, Figs. 610 and 61s) a network as they do in 
the Primates. 
In the Sauropsida the phenomena of yolk-increase, and 
specialisation in the area vasculosa have varied in different 
directions. 
3. The Allantois. 
We now come to discuss the last of the embryonic append- 
ages or fcetal membranes, which are looked upon as charac- 
teristic for Sauropsida and Mammalia, viz. the allantois. 
This again was naturally first known in the chick, and what 
was revealed about it by this venerable archetype of verte- 
brate embryology was applied and adapted as best could be 
to the other higher vertebrates. 
In the didelphic and monodelphic mammals its function 
could not be, as in birds, an extension against the egg-shell 
for respiratory purposes. But even in these it is seen soon 
