80 BULLETIN 114, UNITED STATES NATIONAL MUSEUM. 



(Fresno County), Wild Rose Spring (Panamint Mountains); in Ari- 

 zona — Cave Creek (Maricopa County), Fort Verde (Yavapai County), 

 Cottonia (Mohave County) ; in Utali — Bellevue (Washington County) ; 

 in Nevada — Current school district in Nye Count}^; in Lower Cali- 

 fornia — -Cape Colnett, Cape San Quintin, San Salado, and the foot- 

 hills of San Pedro Martir Mountain. 



Published records for other localities are as follows: Arizona — 

 Date Creek, 50 miles south of Fort Whipple (Coues, 1875, 618), Gila 

 Desert (Baird, 1859, 20); California — Tujunga Wash, east of San 

 Fernando (Grinnell and Grinnell, 1907, 41), Lower Santa Ana 

 Canyon (Grinnell, 1908, 165), and the following records by Van Den- 

 burgh (1897, 171), Oakland, Mount Tamalpais, Camp Taylor, Santa 

 Margarita, Healdsburg, Applegate; Nevada — St. Thomas (Clark 

 County) (Van Denburgh, 1897, 171), Pahranagat Valley (Stejneger, 

 1893, 204), ''El Paso Creek'' and ''Bernicia" (Hallowell, 1859, 14). 

 In addition, the following localities in California are given on the 

 authority of Mr. Carl L. Hubbs: Bed of the Nacimiento River near 

 Joion, Monterey Count}"; road between San Luis Obispo and Avila, 

 San Luis Obispo County; and near Fall Brook, San Diego County. 



Variation. — In spite of similarity in general appearance throughout 

 its range, hoylii is very variable in all of its specific characters The 

 range includes very diverse conditions of topography and climate. 

 The long, high ridge of the Sierra Nevada extends from near its 

 northern limits southeastward for nearly 500 miles, splittmg the 

 range into two sections — an east and a west. That this is a natural 

 barrier to its east and west migration in this region seems fairly cer- 

 tain, and the Mohave Desert, just south of the Sierras, may perhaps . 

 continue the barrier where the mountains leave off — at least there 

 is no record of a hoylii from this desert, although it was found by the 

 Death Valley Expedition in numerous other places. 



Whether or not the Mohave Desert is or ever was a barrier to the 

 westward migration of hoylii, it is certain that this form must have 

 obtained entrance to the Great Valley of California, either from around 

 the south end of the Sierras or from the north. There is certainly 

 no reason for excluding its entrance from the soutli, as Imperial and 

 Riverside Counties offer no obstruction to yumensis from the lower 

 Colorado region, and, as has already been brought out m the dis- 

 cussion of the latter form, yumensis undoubtedly nitergrades with 

 hoylii in southern California. 



In the north the situation is different. The Sierras continue to be 

 a barrier to very near the northern limit of the form. That it is not 

 adapted by nature to range much north of here is indicated by com- 

 parison with the northward distribution of its allies, liolbroolci and 

 getulus. The latter is not known north of the forty-first parallel, and 

 that the former extends somewhat farther north is due to a more 



