492 Causes and Course of Organic Evolution 



But *'in Perameles obesula a further approach towards the 

 formation of a true allantoic placenta is seen, the alhmtois 

 giving rise to small vascular villi. In most marsupials the 

 allantois serves merely as a urinary reservoir, and in none 

 of them does it possess any important function as an organ of 

 nutrition, the young being born at a relatively early stage" 



It will thus be seen that graded stages in viviparous embry- 

 ology can be traced from a simple collateral uterine and egg 

 wall-absorptive relation (uterine placenta), through an at first 

 similar but later collateral uterine and branchial wall connec- 

 tion (branchial placenta), on to a collateral utricular-gland 

 iterine epithelium and yolk-sac union (yolk-sac placenta) and 

 finally reaching a more or less intimate villary allantoic union 

 between parent and embryo (allantoic placenta), from which 

 the transition to the most advanced condition of this, as seen 

 in higher mammals, is easy. So vivipary and easy transition, 

 from simple uterine development of many or two or one em- 

 bryo, to vivipary and uterine allantoic development of an 

 embryo in mammals, can be clearly traced in our suggested 

 line of ascent. Finally that the allantoic blood vessels are 

 branches of the pelvic arteries that ramify over the batrachian 

 bladder, and from it over the developing allantois, seems an 

 added anatomical feature of some value. 



But the amnion that is so precocious and striking a feature 

 of reptiles, birds, and mammals is wholly absent from urodeles 

 and other batrachians. This however is an embryological 

 character that may have arisen in early permian times, as vari- 

 ous means of attaching and feeding the embryo were evolved. 

 Stages in its evolving history are as yet entirely unknoTMi. 



Still another mammalian structure that we can only vaguely 

 trace or suggest the origin of- is the mammary gland. But it 

 should be i*emembered that the urodele skin is very richly 

 sui)i)lied T\ath sunken epidermal glands. Now, if forms derived 

 from these produced more and more feeble and dependent 

 viviparous embryos, these might have nestled against the 

 venter of the parent wliich may have curled round them. The 

 coiling of the parent round the eggs, up to hatching period, 

 in Ichthyophis and Hypogeophis amongst the Apoda, is sugges- 

 tive. By constant sucking on the part of the embryo and 

 stimulation then of the adjoining ventral surface of the mother 

 by the imperfect but viviparous young, some of the abundant 



