VARIATIONS OF GARTER-SNAKES. 85 



scales arc a dull brick red, giving rise to a pattern that resembles very 

 much some specimens of panetaUs from the same region. This speci- 

 men is, however, the only one wliich we have observed with this 

 development of red pigment, and the trait must be considered as of 

 rather imcommon occurrence and not as t^^ical of specimens from 

 this region. 



The "haydeni^' type of color, as above described, may be considered 

 in general as tA'pical of specimens from South Dakota, Wyoming, 

 Nebraska, Colorado, and Kansas, but to the east and north of these 

 points the ground color becomes darker brown, and even black, some- 

 what or entirely obscuring the spots. The stripes in prairie specimens 

 remain much the same as in western ones, but the dorsal tends to lose 

 its golden tinge. Doctor Coues (1878, 277-278) states that Pembina 

 specimens are "olivaceous-blackish or obscure brownish black," and 

 that this color occurs as far westward as the Coteau de Mssouri on 

 the northern boundary, but that in the arid region of the upper Mis- 

 souri and Milk rivers it is replaced by a form whose principal char- 

 acter is seen in the increased breadth and intensity of coloration of 

 the dorsal band, especially on the anterior portion. To this western 

 form Coues and Yarrow (1878, 279-280) gave the name of radix 

 twiningii. Western Iowa specimens are as a rule darker than those 

 from Kansas and Nebraska, and show their close color affinitji mostly 

 in the golden A'ellow dorsal stripe, and it is evident that thef,( speci- 

 mens are to be considered as intermediate between the more pallid 

 western pattern and its darker eastern representative. That the color 

 tends to become darker to the northward as well as to the eastward is 

 shown by the fusing of the spots on the skin in specimens from Turtle 

 Mountain and Regina. 



The third color variety has been described by Cope (1888, 400-401) 

 on the basis of two specimens reported to have been taken at Brook- 

 ville, Indiana (see p. 80). The principal characteristic of these speci- 

 mens was the elongation and fusicm of the gastrostegeal spots to form 

 a broken ])an(l along each side of the abdomen; not an uncommon 

 occurrence in the darker eastern representatives of this form. 



None of these phases differ sufliciently to indicate well-marked 

 forms or to be given subspecific rank, and they have been dropped by 

 later writers. 



In the above discussion of the variations of radix 1 have purposely 

 excluded from consideration three specimens (Nos. 30872, 30873, 

 30874) in the U. S. National Museum from Milwaukee County, Wis- 

 consin. These specimens are typical hutlerl in coloration, and the lat- 

 eral stripe is up )n the third and adjacent halves of the second and 

 fourtli rows. In scutellation they agree both with hutleri and with 

 reduced specimens of radix. The scutellation is as follows: 

 33553— Bull 61—08 7 - 



