294 THE AMERICAN NATURALIST. [Vor. XXXV. 
Because of limited time I was unable to obtain measure- 
ments of culmen curvature for more than forty-seven individ- 
uals (partly males and partly females) a number so small that 
a rather large probable error is found for coefficients of 
variability. 
The results obtained are given in the following table. 
TABLE XIII. — CuRvATURE OF CULMEN. 
23° | 24° | 25^ | 26° | 27^ | 28° | 29° | 30° | 31° | 32° | 33° | 34^ | 35° | 36° MEAN. 
Austroriparian | 1 t iiio? 28.66° + 
Appalachian 3 X*43419|/21311 29.32" + 
Campestrian sia iar 496 bse 1 azta | ee 
Sonora . rl I 32.00? 
To. = : | 1 214:2160|81613|/3/2|5] 141209000 
Austroriparian c = 274? +. Cl. = Dou Er i 0.534 +- 
Appalachian ¢ = 2.03? +. CV.= 6.924. P.E£.¢ = + 0.206 +- 
Campestrian o = 3.12° +. CV.— 10.06 +. P. E.e — + 0.372 +- 
Sonoran = 2160+. CV.= 6.75. P. E. a = + 0.504 ++ 
Total ¢=2.74°+. CV.= 9gaig—. P. E.c- 40.191 +. 
The shrikes of the Austroriparian subprovince, mostly Florida 
birds, are seen to have a mean curvature 3.34°—less than 
that of the Sonoran birds measured. With so few individuals 
precise statements as to curvature are not very reliable, but I 
believe that the figures given are approximately correct for the 
four subprovinces represented. The series of measurements of 
the curvature of the culmen when half the base line A B (Fig. 3) 
was used as the chord gave the following constants and coeffi- 
cients of variability for the forty-seven individuals measured in 
the first series: Mean, 26.34° +; Mode, 27^; o = 9.35 7) 
CV. =8.85; P. E.¢=+40.162 +. 
The angles obtained from the second series of measurements 
were never greater than those from the first series. The 
greatest deviation from the first set of measurements was 7^, 
which occurred in four cases, All the deviations are indicated 
in the following table: 
o? rÉ 29 Lu 4? 5° 6° 7? 
4 2 6 to IO 7 4 4 individuals. 
