Vol. VIII] VAN DENBURGH AND SLEVIN-GARTER-SNAKES 233 



Remarks.— Tht large series at hand shows that this sub- 

 species, which one of us formerly confused with T. o. elegans, 

 and which Brown and Ruthven confused with T. o. ordinoides, 

 really should be separated from both. From T. o. elegans it 

 differs in the smaller average number of its scale-rows and ven- 

 tral plates, as well as in coloration. The dorsal line usually is 

 wider than in T. o. elegans and there often is more or less red 

 in the coloration, which so far as we know is not the case in 

 the mountain snakes. 



T. 0. atratus differs from T. o. ordinoides in being of larger 

 size and in usually having a greater number of upper and 

 lower labials, scale-rows, and gastrosteges. The coloration 

 also is different, although a wide range in pattern and shade 

 is to be seen in both subspecies, and both often show some red 

 coloring. 



As regards scale characters, T. o. atratus may be considered 

 intermediate between T. o. ordinoides and T. o. elegans. 



The two specimens from Siskiyou, Jackson County, Oregon, 

 and two others (Nos. S4313 and S4434) from Anderson, 

 Shasta County, California, probably might best be regarded as 

 showing intergradation between this coast form and the T. o. 

 elegans of the Sierra Nevada, since they all have twenty-one 

 rows of scales and somewhat intermediate coloration. The 

 material is inadequate to make this conclusion a positive one 

 but it is in this region that one would expect to find these sub- 

 species merging. 



Five specimens (Nos. S4471, S4473, S4474, S4476, and 

 S4479) from South Fork, Coquille River, twenty miles above 

 Myrtle Point, Coos County, Oregon, are listed in this paper as 

 T. 0. biscutatus. They, however, are not typical of that form 

 in that they have only nineteen rows of scales. They thus 

 resemble T. o. atratus in this character and might well be re- 

 garded as intergrades. Additional specimens are needed from 

 this general region. The coloration of these specimens is simi- 

 lar to that of T. 0. couchii in the indistinctness of the dorsal 

 line and presence of dark pigmentation on the gastrosteges. 

 Two specimens from Gasquet, Del Norte County, California, 

 resemble these but are so puzzling that one (No. S4264) has 

 been referred to T. o. biscutatus and the other (No. S4266) 

 to T. 0. atratus. Both have more than nineteen scale-rows, a 



