414 NATURAL SELECTION. CHAP. XXT. 



IQ such cases the affinities of species are often more readily 

 discerned by reference to these imperfect structures than by 

 others of much more physiological importance to the in 

 dividuals themselves. 



The same hypothesis would explain why there are no mam 

 malia in islands far from continents, except bats, which can 

 reach them by flying ; and also why the birds, insects, plants, 

 and other inhabitants of islands, even when specifically 

 unlike, usually agree generically with those of the nearest 

 continent, it being assumed that the original stock of such 

 species came by migration from the nearest land. 



Variation and natural selection would also afford a key to a 

 multitude of geological facts otherwise wholly unaccounted 

 for, as, for example, why there is generally an intimate con 

 nection between the living animals and plants of each great 

 division of the globe and the extinct fauna and flora of the 

 post- tertiary or tertiary formations of the same region ; as, for 

 example, in North America, where we not only find among the 

 living mollusca peculiar forms foreign to Europe, such as Grna- 

 thodon and Fulgur (a subgenus of Pyrula), but meet also with 

 extinct species of those same genera in the tertiary fauna of the 

 same part of the world. In like manner, among the mammalia 

 we find in Australia not only living kangaroos and wombats, 

 but fossil individuals of extinct species of the same genera. 

 So also there are recent and fossil sloths, armadilloes, and other 

 edentata in South America, and living and extinct species 

 of elephant, rhinoceros, tiger, and bear in the great Europeo- 

 Asiatic continent. The theory of the origin of new species 

 by variation will also explain why a species which has once 

 died out never reappears, and why the fossil fauna arid flora 

 recede farther and farther from the living type in propor 

 tion as we trace it back to remoter ages. It would also 

 account for the fact, that when we have to intercalate a new 

 get of fossiliferous strata between two groups previously 



