19 14] Grinnell: Mammals and Birds of th( Colorado Valley 157 



were seen among cottonwoods. On the California side, from the vic- 

 inity of Potholes south all along the river to the Mexican line, the 

 species was numerous throughout the hottom lands. Here up to May 

 15 cowbirds were to be seen usually in pairs frequenting the densest 

 willow tracts, as well as foraging on the mud bars at the water's edge. 

 When prospecting in thickets near the ground they were usually quiet 

 and gave the impression of sneaking, as if intentionally planning to 

 avoid observation. But when affecting the taller tree tops and open 

 places, they kept flying about actively, often intently pursuing one 

 another, and .giving frequent voice to an attenuated squeal. The species 

 confines itself strictly to the river bottom : none were seen out on the 

 desert, even in flight overhead. 



Fifty-three skins of the dwarf cowbird were obtained, nos. 12956- 

 13008. 



In addition to these there are in the Museum two skins, one (no. 

 4347) taken by J. G. Cooper at Fort Mohave, April 4, 1861, and 

 another (no. 5963), of unknown collector, from "Mineral City" 

 (=Ehrenberg), April 20, 1864. 



The acquisition of so large a series of Molothrus ater obscunis (pre- 

 viously sparingly represented in collections) from so limited an area, 

 warrants a detailed examination of its characters. The fifty -three 

 specimens are all comparable, as far as age is determinable, and are in 

 comparatively unabraded plumage. The variations in measurements 

 are therefore intrinsic, except, of course, for an indefinite but small 

 amount of error which seems unavoidable. Although but four dimen- 

 sions are presented, it is believed that these are a close index to any 

 other correlated features that could be selected for measurement. And 

 the ones adopted would appear to be capable of more exact determina- 

 tion than, for example, length of body and wing expanse. 



The wing is measured, according to prevalent custom, from the bend 

 of the carpus in closed wing to tip of longest primary; but even here 

 a slight error may result from the mode in which the wing has been 

 "set" in drying, that is, according to the sharpness of the curve along 

 its outer edge. The tail is measured from the point between the inser- 

 tions of the median rectrices into the uropygium to the tip of the 

 longest rectrix in "closed" tail; and here too the proximal point from 

 which to measure is not quite satisfactory. The culmen is rather more 

 accurate of measurement in the cowbirds because it is a well-defined 

 ridge coming to an end rather abruptly mi the forehead. The depth of 

 bill is most unsatisfactory because of the liability to slight separation 



