1618 
grows out much earlier than the true entodermal one, the latter 
penetrating into the former, is an indication that the evolution of 
the allantois may have taken place in the way outlined above. 
2. The yolksack. Our starting point has been an organ containing 
a large amount of yolk but divided completely into cells, such as 
occurs in many Amphibian larvae and embryos. If viviparity has 
led to the attachment. of the germbladder to the uterine-wall, the 
deposition of reserve-material in the cells of the yolksack becomes 
superfluous and thus the amount of yolk will diminish in the course 
of evolution and finally disappear. The yolkfree umbilical bladder 
will retain however some importance as a haemopoietic organ, while 
in some cases respiratory and nutritive functions may be also retained 
to a certain degree. This condition leads in some cases to omphaloid 
placentation (Marsupials, young stages of horse). Though I have 
not been able to remove all difficulties encountered in this point, 
I am inelined to consider the micromphaloid germbladder of Primates 
as primitive and the macromphalon especially if connected with 
omphaloid placentation, as a secondary phenomenon. 
Now as the trophoblast bladder increases in size and thus the 
umbilical vesicle relatively diminishes in extent, two possibilities 
present themselves. The wall of the umbilical vesicle may separate 
on the whole surface from the diplotrophoblast, with exception of 
the proamnion-area (the entyped forewall of the germbladder). In 
this ease the typical fruitbladder of Primates appears, where the 
embryonic rudiment with the small and free yolksack is suspended 
in the large trophoblastbladder by an archamnion-stalk. If however 
the umbilical vesicle remains attached to the diplotrophoblast at 
various points, its wall, being outdistanced in growth by the latter, 
will be torn to pieces as the wall of the fruitbladder increases in 
extent. This happens in Rodents with inversion of germlayers, where 
the yolk-entoderm remains in connection with the trophoblast-layer. 
In this case it presents a continuous layer below the embryonic 
shield only, but it is not able to coat the wall of the whole gerin- 
bladder on the inner side‘). Consequently the cavity of the yolksack 
opens into the-trophocoel. 
I think we may start from similar conditions, if we wish to 
understand the phylogenetic evolution of the sauropsidan egg. We 
have only to imagine that in the common umbo-trophocoel yolk- 
material is deposited, as the primitive viviparity of Protamniotes 
changes into the secondary oviparity of Prosauropsida, that further 
1) See esp. SrLenKa, l.c. H. 1, 1883, p. 16, T. Land I and H. Ill, 1884, T. Xl and XII. 
