THE DESCENT OF THE PRIMATES 31 



backs, now on their bellies, sometimes in special 

 pouches, sometimes in carefully constructed nests. 



Others again may have developed much larger 

 eggs, with an albumen layer and a shell. There 

 can be no doubt on the other hand that a number 

 of them have retained their eggs in the maternal 

 oviducts and have hatched them there, forming a 

 viviparous section. 



In fact, there is really not one cogent reason 

 which woiild prevent us from deriving arrange- 

 ments as we find them in jplacental mammals 

 directly from viviparous amphibian ancestors. 

 The spherical embryonic vesicle with the en- 

 closed umbilical sac and with the embryonic area 

 spread out flat on the top, is not necessarily a deri- 

 vate of a preceding one in which this umbilical 

 vesicle enclosed an enormous quantity of fluid 

 yolk substance. This spherical extension of the 

 vesicle may also have been reached more directly. 

 At all events, this possibility has certainly no less 

 a claim to our careful consideration. Be it well 

 understood, however, that I do not commit myself 

 to professing that I feel sure that it really has in 

 placental mammals developed thus differently. 

 What I wanted to point out is, that the genera- 

 tive adaptations being so varied even amongst 

 living amphibians, they must have been so on 



