38 BULI^ETIN 114, UNITED STATES NATIONAL MUSEUM. 



Specimens from Kansas are more like those from New Orleans in 

 respect to the distinctness of the dorsa bands and the prominence of 

 the spots between, but the spots are yellow (in alcohol) — Branson says 

 (1904, 398) pearly white or yellow. 



The pattern of young individuals is deserving of attention. In all 

 the specimens examined the white crossbars are prominently devel- 

 oped and the scales between are unspotted, or the spots are small, 

 or present only midway between the bands. From the material at 

 hand it seems probable that, as suggested by Branson (1904, 398) 

 and by Hurter (1911, 185), the spots between the bands develop as 

 the young snake grows older. 



The variation in scutellation and proportions needs very little dis- 

 cussion. In these characteristics the differences between this form 

 and its allies, splendida, niger, and getulus, are slight or not apparent. 

 This may be said of the labials, the caudals, the temporals, and the 

 proportionate tail length. The number of specimens is entirely inad- 

 equate for the demonstration of geographic tendencies in these char- 

 acters. It is probable, however, that geographic variation occurs in 

 the number of ventral plates. The average for this species is 210, 

 extremes 201 and 221; for splendida, 217, extremes 208 and 226; 

 iov niger, 209, extremes 200 and 217; for getulus, 215, extremes 204 and 

 223. While it is apparent from the slight differences in the averages 

 for these species that but little reliance can be placed on variation in 

 the number of ventrals, it is nevertheless a fact that the lower numbers 

 of ventrals in TiolhrooTci are more common in the northern and north- 

 eastern parts of its range, and the higher numbers more frequent in 

 the western and southwestern portions. This means approach in this 

 character to niger and to splendida, respectively. 



The scale rows are subject to considerable variation, as the table 

 shows, but the great majority of the specimens have one of the three 

 formulae— 21-19, 19-21-19, and 19-21-19-17. 



This form can thus be compared in lowness of scale formula only 

 with niger and getulus; the former is fully as much reduced as liol- 

 hrooki, and the latter nearly a3 much so. Reference to the table will 

 show that the higher formulae characterize the western and south- 

 western parts of the range, and the lower formulae the northeastern 

 and the extreme southern. If Jiolbrooki had a southwestern origin, 

 these regions would be the last reached, the former because it is the 

 farthest removed from the southwest, and the latter because it is but 

 recently elevated above the sea. 



Affinities. — Ilolhrooki is without doubt very closely allied to splen- 

 dida a nd to niger, and with both of these forms intergradation takes 

 glace along the common boundaries of their respective ranges. We 

 will, however, leave a fuller discussion of affinities to the section on 



