VARIATIONS OF GARTER-SNAKES. 149 



tains, California, on the south to southern British Cohinibia on 

 the north. I have examined specimens from the following locali- 

 ties: Monteiw, Oakland, Fresno, San Francisco, Nicasio, San Pablo 

 Creek, Eureka, Mendocino, Santa Cruz County, Crescent City, 

 California ; Willamette Valley, Chilowynck, Gold Beach, Fort Umpqua, 

 Astoria, Portland, Oregon; Tahot plain, Shoalwater Bay, Chala- 

 hapa, Attlapootl, Port Orchard, Fort Townsend, Olympic Mountains 

 (30 miles from Port Angelus), Puget Sound, Seattle, Fort Steilacoon, 

 Tacoma, Washington; and Comox Lake, Victoria, Vancouver 

 Island. While its distribution is quite definitely limited on the north 

 and west, I believe that it is impossible to fix the exact eastern 

 and southern boundary of ordinoides for the reason that it inter- 

 grades with elegans throughout the entire length of its range. If 

 I am correct in this opinion, there remains to be established the 

 region of intergradation of the two forms. Van Denburgh (1897, 

 209, 211) has recorded specimens from as far east as Tuolumne 

 Meadows (Tuolumne County), Yosemite Valley (Mariposa County), 

 Lake Tahoe (Placer County), and El Dorado County, California, 

 and writes that vagrans (elegans) "is known to live on both slopes 

 of the Sierra Nevada throughout the whole length of the chain," 

 in which case the ranges of the two forms would overlap. As pre- 

 viously stated (pp. 140-141), I have examined specimens similar to 

 those usuall}" referred to elegans from El Dorado, Mariposa (Tenaya 

 Lake Meadows), Tuolumne (Tuolumne ]\Ieadows) , Placer (Lake Tahoe), 

 and Shasta (Baird) counties, California, and Dalles, Oregon, and 

 prefer to consider them all as belonging to elegans ivagrans); while 

 true ordinoides only occurs in material from the base of the moun- 

 tains and to the westward. It is impossible, however, at present 

 to draw the exact line along which these forms intergrade, both 

 owing to the lack of specimens and to the uncertainty of the exact 

 locality from which the specimens that we have were taken. As 

 lias been elsewhere stated, collectors are too prone to label their 

 material with the name of the nearest conspicuous landmark, 

 although the latter may be miles distant. Thus, a series of speci- 

 mens "'from a collector at Oakland" (Brown 1903, 289) are evidently 

 intermediate, but they cannot be used, as we do not know the exact 

 part of the county from which they came. 



For the present, therefore, we can onl}" say that ordinoides meets 

 elegans somewhere on the lower level of the western slope of the 

 Sierra Nevada-Cascade range, but if the former be distinguished 

 from the latter, as we have indicated above, by a more reduced 

 scutellation, and the frequent increased width of the dorsal stripe, 

 we may consider specimens from Stockton, Fort Reading, and 

 Dalles, as intermediate between the two forms. 



As might be expected, the range of ordinoides is nearly divided 

 in northern California and southern Oregon b}^ the mountainous 

 33553— Bull. 61—08 U 



