xiir THE GEOLOGICAL EVIDENCES OF EVOLUTION 393 



the tree-kangaroos of New Guinea ; a large wombat as large 

 as a tapir ; the Diprotodon, a thick-limbed kangaroo the size of 

 a rhinoceros or small elei:)hant ; and a quite different animal, 

 the ISTototherium, nearly as large. The carnivorous Thyla- 

 cinus of Tasmania is also found fossil ; and a huge phalanger, 

 Thylacoleo, the size of a lion, believed by Professor Owen 

 and by Professor Oscar Schmidt to have been equally carni- 

 vorous and destructive.^ Besides these, there are many other 

 species more resembling the living forms both in size and 

 structure, of which they may be, in some cases, the direct 

 ancestors. Two species of extinct Echidna, belonging to the 

 very low Monotremata, have also been found in New South 

 Wales. 



Next to Australia, South America possesses the most re- 

 markable assemblage of peculiar mammals, in its numerous 

 Edentata — the sloths, ant-eaters, and armadillos ; its rodents, 

 such as the cavies and chinchillas ; its marsupial opossums, and 

 its quadrumana of the family Cel)idae. Eemains of extinct 

 species of all these have been found in the caves of Brazil, of 

 Post-Pliocene age ; Avhile in the earlier Pliocene deposits of the 

 pampas many distinct genera of these groups have been found, 

 some of gigantic size and extraordinary form. There are 

 armadillos of many types, some being as large as elephants ; 

 gigantic sloths of the genera Megatherium, Megalonyx, Mylodon, 

 Lestodon, and many others ; rodents belonging to the American 

 families Cavidce and Chinchillida3 ; and ungulates allied to the 

 llama ; besides many other extinct forms of intermediate types 

 or of uncertain affinities. ^ The extinct Moas of New Zealand 

 — huge wingless bii'ds allied to the living Apteryx — illustrate 

 the same general law. 



The examples now quoted, besides illustrating and enforcing 

 the general fact of evolution, throw some light on the usual 

 character of the modification and progression of animal forms. 

 In the cases where the geological record is tolerably complete, 

 we find a continuous development of some kind — either in 

 complexity of ornamentation, as in the fossil Paludinas of the 

 Hungarian lake-basins ; in size and in the specialisation of the 



^ See The Mammalia in their Relation to Priyneval Times, p. 102. 

 2 For a Lrief enumeration and description of tliese fossils, see the author's 

 Geographical Distribution of Ayiimals, vol. i. ji. 14G. 



