220 BULLETIN 19 5, UNITED STATES NATIONAL MUSEUM 



about houses or bams ; but it is doubtful whether it would long remain 

 in any locality after the entire removal of the native vegetation. It 

 is easy to overestimate the abundance of these birds, not only on ac- 

 count of the plurality of nests built by each family but also because 

 their vociferousness usually insures one's consciousness of their pres- 

 ence whenever they are near. 



The geographical subdivision of this species has been attended by 

 considerable confusion and revision, pi-obably because of wide indi- 

 vidual variations, and the number of races ascribed to the United 

 States has fluctuated from one to three and back again. Dr. Edgar 

 A. Mearns (1902b) described the subspecies couesi from Texas, an- 

 thonyi from the interior deserts, and hryanti from the Pacific slope, the 

 second being distinguished from the others by generally lighter colora- 

 tion, and the last by broader white stripes on the back and more white 

 on the tail. In all these, according to Mearns, the throat is mainly 

 black, this feature separating them from affinis of southern Baja 

 California. However, a great degree of variation seems to exist in 

 the pattern of the black throat markings ; these are variously described 

 and illustrated as spots, streaks, and occasionally as coalescent spots, 

 but these differences have not figured as diagnostic characters. In all 

 the adult individuals known to the present writer in Los Angeles 

 County, Calif., the black spots of the throat have coalesced to form 

 a conspicuous throat patch, the upper part of which usually is almost 

 if not quite solidly black and is outlined sharply against the plain 

 white chin. The l^lack area extends slightly higher along the sides of 

 the upper throat and gradually breaks up into streaks or chains of 

 spots on the chest and the sides of the throat. In connection with this 

 tendency in coloration, possibly there may be some significance in the 

 apparent differences in nest-building and singing habits hereinafter 

 noted. 



Nesting. — The list of nesting sites utilized by cactus wrens is a long 

 one. Following a study of a large number of nests near the base of the 

 Santa Rita Mountains south of Tucson, Ariz., Mrs. Bailey (1922) 

 wrote : 



While the name Cactus Wron was justified iu this locality as in others by the 

 innumerable nests found in cholla cactus, here thorny trees and bushes especially 

 catsclaw zizyphus {Z. lycioidcs) or lote bush, were also used extensively, while 

 mesquite and tlie dense shrubby haclvberry or grenjefio were used occasionally for 

 nesting sites. It was interesting to note that zizyphus bushes containing nests gen- 

 erally stood under mesquite trees, so getting double protection. The pi'otection 

 afforded by tlie armament of thorns was often so complete that it was impossible 

 to reach a nest without cutting away the obstructing branches. Even that, how- 

 ever, did not always satisfy the nest makers, for such bulky, conspicuous nests 

 need to be safeguarded in every way from hawks, owls, and other enemies. 

 Thirty-five out of sixty-four nests examined were not only protected by the en- 

 tangling thorns of the surrounding branches but were built within clusters of the 



