^°I'o^^'] Grinnell, Chestnut-backed Chickadee. 3^3 



to me to be the absolutely essential condition for the differentia- 

 tion of two species, at least in birds. 



A strong argument in support of this conviction is that we never 

 find two ' subspecies' breeding in the samefaunal area, and no two 

 closely similar species, except as can be plainly accounted for by 

 the invasion of one of them from a separate center of differentia- 

 tion in an adjacent faunal area. An appropriate instance in illus- 

 tration of the latter is the occurrence together in the Siskiyou 

 Mountains of northern California of the brown Pants riifescens of 

 the wet coastal fauna and the gray Parus gavibeli of the arid 

 Sierran fauna. (See Anderson & Grinnell, Proc. Ac. Nat. Sc. 

 Phila., 1903, p. 13.) The Siskiyou Mountains occupy a line of 

 mergence between the two fauna;, and the two respectively repre- 

 sentative chickadees have evidently extended their ranges toward 

 each other until now over this one small area they occupy com- 

 mon ground. Several parallel cases could be cited ; their signifi- 

 cance seems obvious. 



We come now to consider the origin of the races of Parus 

 rufescens. In a species of recent arrival into a new region (by 

 invasion from a neighboring faunal area), as it adapts itself better 

 and better to its new surroundings, granted the absence of closely 

 related or sharply competing forms, its numbers will rapidly 

 increase. This means that there will be increased competition 

 within the species itself, on account of limited food supply. The 

 alternative results are either starvation for less vigorous indi- 

 viduals during recurring seasons of unusual food scarcity, or dis- 

 semination over a larger area. In a way the first might be 

 considered as beneficial in the long run, as doubtless leading to 

 the elimination of the weaker ; such a process evidently does take 

 place to a greater or less degree all the time, and is important for 

 the betterment of the race. But as a matter of observation Nature 

 first resorts to all sorts of devices to ensure the spreading of indi- 

 viduals over all inhabitable regions ; in other words, the extremest 

 intra-competition does not ensue until after further dissemination 

 is impossible. In birds we find a trait evidently developed on 

 purpose to bring about scattering of individuals. This is the 

 autumnal ' mad impulse ' which occurs just after the complete 

 annual moult, when both birds-of-the-year and adults are in the 



