MIGRATION AND ORIGINS 281 



The physical characters of some races may be adaptive, but it is 

 a difficult task to demonstrate the adaptive nature of most physiologi- 

 cal characters even in full species. In a few important studies, local 

 populations have been shown to have adaptive distmctions. In his 

 classical demonstration of physiological adaptation of local popu- 

 lations K. Goldschmidt (1948) found populations of Lymantria dis- 

 pars at high elevations which were more hardy in cold than popu- 

 lations living at lower and warmer elevations. Some of these 

 physiologically distinguishable neighboring populations were shown 

 to be genetically differentiated by incompatibility for reproduction. 

 J. A. Moore (1949) has shown that examples of Rana sylvatiea from 

 northern ranges produce eggs which are viable in colder water than 

 can be endured by the eggs of southern populations. Some incom- 

 patibility in fertilization by parents from remote ranges also indi- 

 cates their genetic distinction. The differentiation of Rana sylvatiea 

 has occurred in northern areas and is probably of recent origin. These 

 obviously adaptive and genetic variations distinguishing populations 

 show that adaptively significant variations sometimes become 

 established within species. 



I consider it inescapable that natural selection will favor any adap- 

 tive variation according to its value to the possessor. If the variation 

 is an inheritable character it will proceed through the genetic mecha- 

 nism with a bias for its preservation in proportion to its adaptive 

 value. However, if we consider how rarely the small variations 

 within a species have been proved to be inheritable adaptations, we 

 must admit that there are few demonstrations of opportunities for 

 natural selection, per se, to operate in the differentiation of discrete 

 populations within a species. Some other factor or agency must then 

 be sought for the differentiation of the numerous taxonomic races 

 which are found. 



The variations shown in the northern races of Alaskan birds have 

 no obvious adaptive value. Nevertheless, they clearly mark off geo- 

 graphic populations readily distinguishable from other populations 

 of the same species elsewhere. The distinctions, whether of large or 

 small degree of difference, appear to be maintainable only by prev- 

 alent inbreeding among the members of each race. Segregation of 

 a species into recognizably distinct populations cannot take place 

 without a sorting process, by means of which the homogeneous species 

 becomes fragmented into races. Rather than surmise that some in- 

 visible adaptation in form or in function favors the orderly segre- 

 gation of populations within a species, it seems preferable to seek 

 out the conspicuous behavioral activities of birds that can keep these 

 groups apart and intact as reproductive segments of the whole. This 



