XIII EVOLUTION AND DISTRIBUTION OF ANIMALS 283 



the undoubted fact that every group of animals whose 

 distribution is discontinuous is now more or less in a 

 fragmentary condition, and has, in all probability, once had 

 a much more extensive range, to which its present dis- 

 tribution may offer no clue whatever. Who would ever 

 have imagined, for example, that the horse tribe, now 

 confined to Africa and Asia, formerly ranged over the 

 entire American continent, north and south, in great 

 abundance and variety; or that the camel tribe, now 

 confined to Central Asia and the Andean region of South 

 America; formerly abounded in North America, whence in 

 fact our existing camels were almost certainly derived ? 

 How easy it is to imagine that analogous causes to those 

 which have so recently exterminated the horses of 

 America and Europe might have acted in a somewhat 

 different direction, and have led to the survival of horses in 

 South America and Africa, and their extermination else- 

 where. Had this been the case, how strong would have 

 been the argument for a former union of these two 

 continents ; yet we now know that these widely 

 separated species would be merely the relics of a once 

 dominant group which had occupied and become extinct 

 in all the northern continents. 



Discoveries of extinct forms remote from the countries 

 they now inhabit, are continually furnishing us with new 

 proofs that the great northern continents of the two 

 hemispheres were really the birthplace of almost if not 

 quite all the chief forms of animal life upon the globe ; 

 while change of climate, culminating in the glacial epoch, 

 seems to have been the motive power which has driven 

 many of these forms into the tropical lands where they 

 now alone exist. 



If we give full weight to these various considerations, 

 and at the same time bear constantly in mind the 

 extreme imperfection of our knowledge of extinct land 

 animals, we shall, I believe, have no difficulty in explaining 

 most of the apparent anomalies in zoological distribution, 

 and in imagining a possible and even probable solution 

 for those extreme cases of difficulty which the facts at our 

 command do not yet permit us to explain in detail. 



