THE ORIGIN OF SPECIES AND GENERA 297 



Colour also varies greatly in correspondence to latitude 

 and longitude. Dark-coloured birds are said to become 

 blacker towards the south ; in others the yellow or red 

 bands become deeper: while in those transversely banded, 

 the dark bands become broader, and the light ones nar- 

 rower. Those with white spots or bands have them 

 smaller in the south, and sometimes lose them altogether. 

 These differences are sometimes so great that the ex- 

 treme northern and southern forms might be considered 

 distinct species were it not for the perfect gradation of 

 intermediate types in the intervening localities. There is 

 also an increase of intensity of colour from east to west, 

 as exhibited by the same or b}^ closely allied representa- 

 tive species inhabiting the Atlantic and Pacific coasts 

 respectively. In the desert plains of the interior, how- 

 ever, the colours are paler than on either coast ; but this 

 is no doubt a protective modification, assimilating the 

 tints of animals to the rock or surface soil on which they 

 dwell. In some cases well-marked varieties of the same 

 species appear to be confined to the Eastern States and 

 to California respectively, as, for example, the eastern and 

 western forms of Bewick's Wren ( Thrijothorits Bewickii), 

 which differ greatly in the length of the bill, although 

 otherwise almost identical; and as these two forms do 

 not, so far as yet known, anywhere intermingle, they 

 afford a good example of the first step in the formation 

 of a new species. The beautiful purple finch {Carpodacus 

 purpureus) of the Eastern States, and its western form, 

 which has been named californiciis, perhaps form another 

 example ; but until the range of these birds is fully and 

 accurately determined, we cannot be sure that there is not 

 some limited area where the two forms intermingle and 

 their distinctive characters disappear. 



From the fact of variation, so extensive as regards the 

 number of variable characters and so large in absolute 

 amount as has now been proved to exist in many species, 

 we may fairly draw the conclusion that analogous varia- 

 tions, sometimes of less and sometimes of greater extent, 

 is a general characteristic of animals in a state of nature ; 

 and with such materials to work with, it becomes easy 



