No.412.] SMALLER NORTH-AMERICAN SHRIKES. 297 
the formation of numerous local variations, or family character- 
istics, whose recognition is a matter of power of discrimination 
on the part of the systematist. 
It seems highly desirable that the question of limiting the 
establishment of new subspecies or varieties by some generally 
accepted criteria be considered. 
I do not argue for the universal use of the method of the 
* Precise Criterion," but I believe that it is both desirable and 
practicable to employ it in certain problems of taxonomy, such, 
for instance, as the one just discussed. The ordinary work of 
classification, perhaps, does notat present require the precision 
in treatment furnished by purely quantitative methods, but 
problems of race distinction, I believe, need the precision of 
the “Precise Criterion." The contention that quantitative 
methods are less useful than those ordinarily employed because 
of the large amount of material required, is mischievous, for it 
argues that generalizations professing precision are possible by 
methods which are not precise. The problems of finer classi- 
fication can be properly settled only by the use of a large 
amount of material, whatever the methods used. 
V. SUMMARY. 
Quantitative methods have here been applied to the study 
of variation in the smaller shrikes of North America, and the 
following variable characters have been measured : 
. Length of wing. 
. Length of tail. 
. Length of bill. 
. Depth of bill. 
ulmen. 
. Color of dorsal surface of head. 
. Color of upper tail coverts. 
. Color of breast. 
The three color areas have been found to vary principally in 
the amount of melanism present. A series of 294 shrike skins 
from various parts of the United States, Mexico, and southern 
Canada have been studied and measurements of these skins 
have been classified. — 
on Am 2 CO Wh & 
AQ 
fa 
3 
E 
o 
S, 
o 
