280 THE AMERICAN NATURALIST. | [Vor. XXXV. 
properly appreciate the estimates of color given in this paper, 
a color mixer should be used to reproduce the combinations 
given. 
Ability in the discrimination of color I have found to be 
much developed by experience. The color determinations of 
my first three months’ work were rejected, as increased power 
of discrimination made more accurate results possible. 
C. Geographical Areas. 
The material described in this paper has been collected from 
a territory of great size and varying conditions, and it is there- 
fore desirable to compare individuals both as parts of a single 
group and in subdivisions corresponding to natural life areas, 
so that correlations between individual variability and geo- 
graphical variation may be made. I have adopted the life 
areas employed by Allen (93) for the territory covered by my 
material, which are four in number: 
I. Austroriparian Subprovince, embracing North and South Carolina, 
Georgia, and Florida. 
2. Appalachian Subprovince, embracing Maine, Massachusetts, Con- 
necticut, District of Columbia, Virginia, Ontario, New York, Pennsylvania, 
Indiana, Kentucky, Ohio, Illinois, Wisconsin, Minnesota, and North 
Dakota. ; 
3. Campestrian Subprovince, embracing British Columbia, Montana, 
Wyoming, Colorado, Indian Territory, New Mexico, Texas, Idaho, Utah, 
Arizona, and Nevada. 
4. Sonoran Subprovince, embracing California, Lower California, and 
Mexico. 
A comparison of variations for still smaller areas is desirable, 
but this, to be useful, would require more material than I have 
been able to obtain. The subspecies of shrikes have the dis- 
tribution given by Palmer (98), excepting gambelli (Palmer 
did not consider the western shrikes), which agrees very well 
with the above life areas. The inhabitants of each are: 
SUBPROVINCE. SUBSPECIEs. 
Austroriparian — ludovicianus 
Appalachian — migrans 
Campestrian — excubitorides 
[Sonoran — gambelli]? 
