334 TROPICAL NATURE, AND OTHER ESSAYS. 



dons, true horses, porcupines, and beavers, existed in 

 Europe long before tbey appeared in America ; and as 

 the theory of evolution does not admit the independent 

 development of the same group in two disconnected 

 regions to be possible, we are forced to conclude that 

 these animals have migrated from one continent to the 

 other. Camels, and perhaps ancestral horses, on the 

 other hand, were more abundant and more ancient in 

 America, and may have migrated thence into Northern 

 Asia. 



There are two probable routes for such migrations. 

 From Norway to Greenland by way of Iceland and 

 across Baffin Bay to Arctic America, there is everywhere 

 a comparatively shallow sea, and it is not improbable 

 that during the Miocene period, or subsequently, a land 

 communication may have existed here. On the other 

 side of the continent, at Behring Straits, the probability 

 is greater. For here we have a considerable extent of 

 far shallower sea, which a very slight elevation would 

 convert into a broad isthmus connecting North America 

 and North-East Asia. It is true that elephants, horses, 

 deer, and camels would, under existing climatal condi- 

 tions, hardly range as far north as Greenland and 

 Alaska ; but we must remember that most mysterious 

 yet indisputable fact of the luxuriant vegetation, 

 including even magnolias and other large-leaved ever- 

 greens, which flourished in these latitudes during the 

 Miocene period ; so that we have all the conditions of 

 favourable climate and abundant food, which would 

 render such interchange of the animals of the two con- 

 tinents not only possible, but inevitable, whenever a 

 land communication was effected ; and there is reason 



