"lyo Grinnell, Chestnut-backed Chickadee. Lfjiv 



words, the divarication is now wholly complete and there are two 

 separate twigs. The area of intermediate faunal conditions be- 

 tween the humid coast belt and the arid interior region of British 

 Columbia and Alaska is very narrow, consisting, in places per- 

 sonally traversed by me, of but a few miles over a mountain ridge. 

 This very narrowness of the area of faunal mergence probably 

 accounts for the lack of intermediates at the present day between 

 hudsofikus and rufescens. 



The center of distribution of any animal is where the greatest 

 rate of increase is. The greatest rate of reproduction is presum- 

 ably where the species finds itself best adapted to its environment ; 

 and this is also where the death rate is least, unless an enemy 

 rapidly multiplies so as to become a serious check. In a wide- 

 ranging species, or one that is rapidly spreading over a region of 

 varying climatic and associated conditions, sub-centers of distri- 

 bution will arise at points which prove to be more favorable, in 

 point of food supply and minimum of enemies, than intervening 

 areas. From each of these new centers of distribution there will 

 be a yearly radiating flow of individuals into the adjacent country, 

 so as to escape intra-competition at any one point. 



Such centers of distribution will obviously, as time goes on, har- 

 bor only locally pure-bred individuals, for foreign individuals will 

 not stem the tide of population from season to season slowly 

 emigrating. This will amount to operative isolation and allow of 

 the time necessary for the impress, by local factors of environment, 

 of incipient characters, which, through cumulative inheritance as 

 the element of time further increases, become to us perceptible 

 and characterize this set of individuals as a geographical race or 

 * subspecies.' 



Let us suppose that descendants from the interior Parus pre- 

 hudsonicus from season to season pushed their way further and 

 further into the primaeval coast belt until the latter supported a 

 vigorous colony. The coastal humidity was very likely at that 

 time but slightly greater than that of the interior, having gradually 

 increased through slow shifting of ocean currents or other causes, 

 so that the faunal boundary was not so abrupt and did not then as 

 now constitute a formidable barrier to invasion. 



Faunal conditions are without doubt undergoing- constant alter- 



