296 THE AMERICAN NATURALIST. [VOL XXXV. 
were obtained from New England, whereas Illinois alone is 
represented by thirteen. This lack of equalization in the 
sources of material prevents certain desirable interpretations 
of tendencies towards minor variations. 
To my mind one of the most important results reached is the 
determination of the relative variability of different characters 
in a group of birds representing geographical areas of consid- 
erable size. The coefficients of variability (Tables VI-VII) 
indicate for the wing and tail a variability of less than 4 in. 
length. The bill is somewhat more variable, as is shown by a 
coefficient of 5.89 -- for the length of bill in a series of 164 
males. Color, as would be expected, is much more subject to 
variation. The upper tail coverts and breast are the most vari- 
able; but the coverts furnish only a very small part of a bird's 
coloration, and the color of the breast was the character which 
it was the most difficult to measure satisfactorily, especially as 
advanced age and the condition of the plumage are factors of 
possible importance which I have found it difficult to consider. 
Fortunately for this particular study, shrikes do not change 
much after the first winter plumage is obtained. 
I believe that migrans is as worthy of recognition as gam- 
belli. Whether it is profitable to encumber nomenclature with 
the names of these races, based on slight variations, is a ques- 
tion which is worthy of further consideration. 
The power of discriminating fine shades of color varies in 
different persons, and it can be highly developed by education. 
At the present time there is much activity among certain 
systematists in the production of new subspecies for geo- 
graphical varieties, which long experience and special adept- 
ness enable them to distinguish. A variation, no matter how 
slight, that can be correlated with geographical range is con- 
sidered to warrant an addition to nomenclature; but the dis- 
covery and description of geographical races can be carried on 
almost ad infinitum. 
Birds, because of their powers of flight, might be expected 
to be less subject to the factor of isolation than non-migratory 
animals, but the tendency to return in spring to the same 
breeding place must, in some species at least, be conducive to 
