122 THE PROBLEMS OF EVOLUTION 



midst of their area. If they extend ultimately 

 beyond the range of individual migration, their 

 extremes are as surely isolated as if an insurmount- 

 able barrier existed between them. 



There is a difference between these types of 

 spatial isolation, however. While organisms living 

 in Colorado cannot possibly mate with organisms 

 living in Maine, there is a gradual chain of associ- 

 ation connecting the two extremes. The individu- 

 als of a species are linked together throughout a 

 continuous area of distribution by indirect and 

 potential reproductive association, even though 

 not by unlimited possibility of direct contact. On 

 a group of islands like the Galapagos they may be 

 absolutely isolated in separate groups. In either 

 case certain characteristics within the range of 

 heritable variations of the species may originally 

 have determined the distribution of individuals, 

 but at the same time many incidental characters, 

 bearing no determinative relationship to the loca- 

 tion, may characterize the development of the 

 isolated or partly isolated groups. And even though 

 a tenuous indirect association may persist through- 

 out the range of a species, there is little actual 

 chance that the characteristics of the Colorado 

 individuals will affect the development of the 

 species in Maine to a noteworthy degree. 



An example of these types of isolation has re- 

 cently demanded my attention, namely, the cir- 



