470 PROCEEDINGS OF THE ACADEMY OF [April, 



have the color pattern of the adults, but more or less cross-banding 

 with narrow interspaces frequently replaces the spots on the forepart 

 of the body in young and some adult T. s. sipcdon, resembling those 

 which form the normal pattern on the whole length in eastern examples 

 of T. s. fasciatus — a fact which perhaps signifies the partial retention of 

 an ancestral pattern, now completely lost in the third subspecies, T. s. 

 transversus of the Louisianan fauna. It is probable that T. s. fasciatus 

 of the southeast represents the earliest form. Within the bounds of 

 the Ocmulgian, T. cyclopeum and T. taxispilotus have developed 

 from T. sipedon, while farther west, in the Mississippi valley, the way 

 to T. rhombifer was through T. s. transversus. The whole of the ill- 

 defined series grouped about T. compressicaudus and T. ustus is Flori- 

 dan and shows the wealth of differentiation, still unstable as to fixity 

 of character, which has resulted in the Ocmulgian under conditions 

 highly favorable to the genus. There seems no clue to the direct ori- 

 gin of the group of species represented hj T. leberis and its allies, with 

 seventeen or nineteen rows of scales and longitudinal stripes, but it is 

 worth noting that a similar distribution of color occurs m a number of 

 eastern Palsearctic and Oriental Tropidonotus. 



Separated structurally from the foregoing genus only by its entire 

 anal plate is Eutcenia, which covers the whole Nearctic to the northern 

 limit of snakes, and perhaps equals in number of individuals all other 

 species combined. From its cosmopolitan extension Tropidonotus 

 must be the parent. It is suggestive that the eastern form of the com- 

 mon garter snake, known as E. sirtalis ordinatus, with dark spots on 

 a greenish ground and no stripes, much resembles some color phases 

 of Tropidonotus natrix of Europe, and has like it nineteen rows of 

 scales and seven upper labials. The yellow collar of T. natrix is not 

 known in any race of E. sirtalis, but it possibly reappears in the pale 

 postoral crescents of some Sonoran species, as E. marciana and E. 

 hammondi. 



E. s. ordinatus grades into E. s. sirtalis, whose western representative, 

 E. s. parietalis, on the Pacific coast rims into E. elegans through such 

 examples as those which Cope regarded as E. infernalis,* and in the 

 southwestern Sonoran into E. marciana, E. hammondi and E. eques. 



The species of the great plains, E. radix, has commonly twenty-one 

 scale rows and seven upper labials, but often presents the scutellation 

 of E. sirtalis and probably is derived from it, an important change 

 being that the lateral stripe has moved up one row of scales. And it 



^ A. E. Brown, Proc. Acad. Nat. Sci. of Phila., 1903, p. 288, etc. 



