124 BULLETIN 17 5, UNITED STATES NATIONAL MUSEUM 



the back, a row of much smaller blotches on each side, transverse bands of jet 

 black upon the tail, tail short, abdomen and tail thickly maculated with black, 

 thirty-one rows of carinated scales, abdominal scuta 221; subcaudal 64. 



Dimensions: Length of head 1 inch 2 lines; greatest breadth 8 lines; length of 

 body 2 ft. 5 inches, of tail 5 inches 5 lines; greatest circumference 2 inches 2>^ lines. 



Another specimen was received of the same species as the above, but which 

 presents a remarkable deviation in the form and arrangement of the plates upon 

 the head, which is no doubt abnormal. Thus there are seven plates upon the head 

 instead of six, as in Piiyophis; these are arranged in three rows, two in the middle, 

 and three in the posterior; on each side of the middle row is a small quadrangular 

 plate lying immediately above the loral, constituting as it were a superior loral; 

 there is but one antorbitar and four posterior orbitars on the right side, and three 

 on the left; there are nine superior labials; abdominal scuta 227; subcaudal 7L 



Habitat: New Mexico. 



Systematic notes. — The name affinis proposed in 1852 by Hallowell 

 for specimens from New Mexico has never since been recognized by 

 other authors. However, as both the description and the type 

 locality would identify his specimens with this form, it must be 

 retained as the earliest name applied to the form. 



The name bellona was given by Baird and Girard in the same year 

 to a specimen from Presidio del Norte, Chihuahua, Mexico, and was 

 for many years used rather generally to apply to specimens from 

 Arizona and New Mexico as well as to specimens of typical sayi. 

 Of the name bellona Stejneger (1893, p. 206) says: "There can be no 

 doubt that Baird's and Girard's original Churchillia bellona, which 

 came from Presidio del Norte, Chihuahua, Mexico, was a typical 

 P. sayi. The type appears now to be lost, but I have before me a 

 specimen from the identical locahty (U. S. N. M. No. 1542) with a 

 most pronouncedly narrow rostral and agreeing with P. sayi in all 

 other respects also." Thus, although this name has frequently been 

 applied to specimens of affinis, it is originally a synonym of P. s. sayi 

 and must be discarded. 



The name P. catenifer rutilus was proposed by Van Denburgh in 

 1920 for the Arizona gopher snakes. Since Arizona and New Mexico 

 specimens are identical, rutilus becomes a synonym of affinis. 



P. s. affinis is here considered as a subspecies of sayi rather than of 

 catenifer, since it intergrades with s. sayi in northern Mexico, where 

 the two forms can be distinguished only with the greatest difficulty, 

 while it is evidently quite distinct from all the subspecies of catenifer 

 throughout its range. Furthermore, affinis is obviously much closer 

 to sayi than to any of the subspecies of catenifer in both scale and 

 pattern characters. 



Diagnosis. — From the forms of the melanoleucus group, affinis may 

 be separated at a glance by the shorter rostral, which is only slightly 

 longer than broad, while in the subspecies of melanoleucus it is at 

 least twice as long as broad. It may be distinguished from the sub- 

 species of melanoleucus also by coloration. Thus in affinis there are 



