in its bearing on the Distribution ^c. of Birds. 241 



American route. Being a genus unfitted for great heat, it 

 penetrated no further than Algeria, California, and North 

 China, retaining an identity all but specific everywhere, 

 except in the southern limits of its migration to Algeria and 

 California, where the species are markedly distinct. 



Of the Jays, Perisoreus remains in the Arctic regions, with 

 very slight variations. Garrulus, which left earlier, has 

 varied more rapidly as it worked southwards and, like many 

 other families, has never even crossed the Sahara. 



I assume that the progenitors of the tropical forms naturally 

 left the north long before those better adapted to endure 

 change of temperature, and thus had an infinitely longer 

 period in which to become more and more differentiated. 



I do not jjretend to explain how it happens that great families 

 are now restricted to particular regions, while other regions 

 not less fitted, so far as we can judge, for their maintenance 

 are without them : how, for instance, the Humming-birds, 

 the Toucans, the Tanagers, are solely Neotropical ; the 

 Plantain-eaters and the Guinea Fowls solely Ethiopian ; the 

 Pheasants solely Indian ; the Honey- suckers only Indian and 

 Australian; the Sun-birds only Indian and Ethiopian. But 

 the Polar theory seems to give the only possible solution, viz. 

 that their respective ancestors left by one or two routes only, 

 or that the other parties perished early in the struggle for 

 existence. Now the only marked exception to the close simi- 

 larity — I had almost said identity — of the North-American 

 land-bird fauna with that of the Old World is in one peculiar 

 family, the American Warblers, or Mniotiltidce. Structurally 

 these differ very widely from our Warblers ; but they are all 

 migrants, visiting North America only in summer, and we 

 cannot look on them as northern forms at all. They are 

 simply Neotropicals, which have no representatives in other 

 regions, and which have preserved their old traditions of 

 hugging the north as closely as they could, though greatly 

 modified in structure. The hereditary instinct in them has 

 been more permanent than the form. There is also another 

 possible cause, viz. that the North-American continent has 

 been in various parts repeopled from the south after its 



SER. v. — VOL. v. R 



