32 CALIFORNIA ACADEMY OF SCIENCES [Proc. 4th Ser. 



barred area at the tip, as shown in a considerable proportion 

 of specimens. 



In southeastern Arizona eremophilns is a common resi- 

 dent of both the Upper and Lower Sonoran zones, being 

 abundant in the Lower Sonoran river valleys. In eastern Cal- 

 ifornia this wren appears to breed only in Upper Sonoran, mi- 

 grating down into Lower Sonoran in the winter months. It 

 is, I believe, the only California wren of this genus that has 

 truly migratory habits. 



That individuals of the species should be found in the 

 valley of the Colorado River, at the western boundary of 

 Arizona, in winter only (Grinnell, 1914, p. 209), while the 

 species is resident the year through in valleys of similar zonal 

 character in the eastern part of the state, argues, it seems to 

 me, that these sets of individuals represent two entirely sepa- 

 rate aggregations, two distinct subspecies, if we wish to call 

 them so. Furthermore, observations so far made as regards 

 other species show that the valley of the Colorado does not serve 

 as a winter home for birds from the colder country to the 

 eastward. The transient winter population of this valley seems 

 to come almost entirely from the Great Basin region to the 

 northward, with its surrounding mountains, and it is fair to 

 assume that these winter visiting wrens are also from that re- 

 gion. The slight differential characters apparent, as enumer- 

 ated above, appear to bear out this assumption. 



Personally I believe that there are at least two separate 

 forms included under the term eremophilus as now used, one 

 occurring in the Great Basin region of southern Nevada and 

 eastern California, migrating over the Colorado and Mohave 

 deserts in winter, the other represented by the resident bird 

 of southeastern Arizona. Additional material is needed from 

 California, in the nature of fresh autumnal and early spring 

 specimens from breeding stations, more clearly to demonstrate 

 the presence of differentiating characters. In the absence of 

 such material, and bearing in mind the variability shown by the 

 M^rens of this genus where different subspecies meet, it seems 

 advisable to refrain from formally affixing a name to the pos- 

 sibly recognizable California race. It is significant in this con- 

 nection to note that in proportion as California specimens of 



