VARIATIONS OF GARTER-SNAKES. 145 



California and oast central Oroijon and Wasliin^jton. At first sij^ht 

 the localities in whicli two preoculars occur seem to represent an 

 isolated re^iion, hut, as will he shown later, I helieve that the trait 

 outcrops all along the western houndary of the ran^je (in the Sierra 

 Nevada-Cascade ran<je) from the southern limit in the mountains of 

 southern California to the northern limit in British Columbia. 



EUgans is (piite constant in coloration over most of its range. 

 The following description holds for most of the specimens east of 

 the Sierra Nevadas: Ground color light brownish olive, relieved by 

 two rows of rather small black lateral spots that usually occupy 

 only the edge of the scales; stripes yellow, tinged with orange or green, 

 the laterals upon the second and third rows and frequently nearly 

 indistinguishable from the first row of scales which is lighter than 

 those above, the dorsal upon the median and more or less of the 

 adjacent rows being nearly always encroached upon to the median 

 row by the superior row of lateral s])ots ; narrow bars at the ])aso of the 

 ventral scutes that nuiy or may not be enlarged at the ends to form small 

 spots and in the midtUe to make an irregular median ventral band. 



In the Sierra Nevada-Cascade range the color tends to become 

 generally darker, although the pattern of eastern specimens is 

 retained. Such individuals have been distinguished by Cope (1S92, 

 654-655) as lineolata and brumua, but an examination of the types 

 as well as many other specimens from the same region shows that 

 they differ in no w^ay from tyi)ical elegans except in the darker ground 

 color, wdiich obscures the lateral spots on the scales. In his tlescrip- 

 tion Cope states that in the type of hrminea there is not the least 

 trace of lateral spots. This is not true, as they are easily seen on the 

 skin when it is stretched. Specimens do occur on the west slo])e of 

 the Sierras, however, in which the sj)ots are mostly fused on the 

 skin. This dark color is frequently accompanied by an increase in the 

 width of the dorsal stripe, a tentlency which in elegans occurs only in 

 this region, but, as I shall show later, ])ecomes of general occurrence 

 to the westward. vSuch a s|)ecimen was described as elegans by Baird 

 and Girard in lS53,but that thecoiul>ination of characters can not ])e 

 distinguished even as a subspecies is quite evident, as all intermediate 

 stages occur between it and the dark (lineolata or hnninea) specimens 

 of elegans in the Sierra Nevadas, while the scutcllation is identical." 



Occasionally the dorsal stri])e is nearly or quite obsolete, as in 

 hammondi. S>v.c\\ a specimen w^as (lescri])ed by Kennicott as E. couclii. 

 Some writers have considered these s])eciniens identical with ham- 

 mondi, but all variations in the width of the dorsal stripe occur, and 

 it is evident that its total or partial obliteration is merely a variation. 



o It is true that Baird and Girard gave the number of dorsal rows of the type speci- 

 men of elegans as 19, but these were evidently counted on the anterior part of the 

 body alone, for the formula is 19-21-19-17, as in many specimens of elegans (vagrans). 



