182 



THE AMERICAN NATURALIST [Vol. LII 



latter point, tests which I have made of the standards of 

 measurement employed by several competent collectors 

 show clearly that the differences due to "personal equa- 

 tion" are sometimes at least as great as those which char- 

 acterize quite distinct local races. 



Accordingly, we might feel justified a priori in enter- 

 taining some skepticism as to the homogeneity of these 

 races of animals throughout such great areas. Further- 

 more, I already have a certain amount of direct evidence 

 which renders this contention improbable. Such evidence 

 will be considered later. 



An extremely desirable undertaking would be to run a 

 series of trapping stations through the territories of two 



A 5 



adjacent subspecies, at right angles to the supposed line 

 of demarcation. This the author hopes to do in the 

 course of time, though the task is not as simple as might 

 perhaps be anticipated. Theoretically, a number of pos- 

 sible conditions might be revealed by such an investi- 

 gation. 



In the first place, it might be found (Fig. 1) that each 

 of the two races was, in reality, "absolutely identical' ' 



A B 



throughout its own range, while the transition between 

 the two might be fairly abrupt. 



Secondly, there might be an unbroken intergradation, 

 in respect to the differential characters, throughout the 



