484 Causes and Course of Organic Evolution 



The absence of an amnion and well-developed allantois 

 in Batrachia and their presence in reptiles and mammals seem 

 highly favorable for an alliance of the two last. But the un- 

 doubted developing rudiment of the allantois as a cloacal 

 outgrowth exists in Batrachia, while in what we would regard 

 as the most primitive group of mammals, namely the Mar- 

 supialia, an allantois is not seen in most, and only attains 

 to the condition of an allantoic placenta in Perameles. The 

 amnion as such is entirely unrepresented in the last, but when 

 we consider how striking and fundamental a modification 

 occurs in Salamandra maculosa and S. atra (155: 117) so as 

 to make these viviparous, while the group is oviparous, such 

 is proof of a degree of adaptability and modification that might 

 readily have stimulated the further marked enlargement of 

 the allantoic rudiment, and equally have started new forma- 

 tion of an amnion during the mid- or late carboniferous, or at 

 latest in the early permian, epoch of the earth's history. 



Mention of the above formation of a viviparous habit in 

 some of the higher urodeles forms a point of contact between 

 Batrachia and Mammalia that we cannot afford to overlook. 

 That the habit should appear amongst such Apoda as Der- 

 mophis and Typhlonedes, and that it should reassert itself 

 in Spelerpes and Salamandra amongst the Urodela, is proof 

 that here at least a typically mammalian habit has been fairly 

 well established. But attention might also be drawn to the 

 post-embryonic and yet premature nourishment effected by 

 members of the Marsupialia, a group that we would look on 

 as the basement series of the mammals. For, when the freed 

 foetus or embryo is transferred by the mother marsupial to 

 its pouch, it continues — unwittingly doubtless by chemotactic 

 action as with milk for a sucking child — to absorb food from 

 a free source, much as do the foetal salamanders, when they 

 at one stage absorb the products of surrounding and arrested 

 eggs alongside them. 



Mention of the marsupials can now serve for introduction 

 of the question as to which of the mammalian groups the Ba- 

 trachia most nearly approach. It must freely be confessed 

 that palseontological knowledge is here so scant, and the present 

 gap that separates Batrachia and Mammalia is so wide, that 

 only an approximate and suggested answer can be given. But 

 many reasons tend strongly in favor of the marsupials as at 

 least the commencing members of tlie main line. For it in- 

 cludes such diverse types as the edentate-like MyrmecobiuSy 



