376 PALEONTOLOGY AND EVOLUTION XI 



in point of extent or variety with those which are 

 yielded by the Miocene strata. However, what we 

 do know of this Upper Eocene Fauna of Europe 

 gives sufficient positive information to enable us 

 to draw some tolerably safe inferences. It has 

 yielded representatives of Insectivora, of Cheir- 

 optera, of Rodentia, of Carnivora, of artiodactyle 

 and perissodactyle Ungulata, and of opossum-like 

 Marsupials. No Australian type of Marsupial has 

 been discovered in the Upper Eocene strata, nor 

 any Edentate mammal. The genera (except per- 

 haps in the case of some of the Insectivora, Cheir- 

 optera, and Ecdentia) are different from those of 

 the Miocene epoch, but present a remarkable 

 general similarity to the Miocene and recent 

 genera. In several cases, as I have already shown, 

 it has now been clearly made out that the relation 

 between the Eocene and Miocene forms is such 

 that the Eocene form is the less specialised ; while 

 its Miocene ally is more so, and the specialisation 

 reaches its maximum in the recent forms of the 

 same type. 



So far as the Upper Eocene and the Miocene 

 Mammalian Fauna3 are comparable, their relations 

 are such as in no way to oppose the hypothesis 

 that the older are the progenitors of the more 

 recent forms, while, in some cases, they distinctly 

 favour that hypothesis. The period in time and 

 the changes in physical geography represented by 

 the nummulitic deposits are undoubtedly very 



