84 BULLETIN 61, UNITED STATES NATIONAL MUSEUM, 



be seen that there may be 8, 9, or 10 in this series, but that 8 infra- 

 labials have only been found in the prairie region. In the case of 

 both the superior and inferior labials we readily admit the possi- 

 bility that larger series of plains specimens may destroy this appar- 

 ent tendency toward a decrease in the number of plates to the east- 

 ward, but we do not think that it is probable, for, as in the number of 

 dorsal scale rows, an examination of all of the plains specimens has 

 failed to turn up a single specimen with 6 superior or 8 inferior labials, 

 so that if these numbers exist in this region it must be in small pro- 

 portion. 



If the tables represent the true state of affairs, then radix evidentl}/ 

 tends to become slightly reduced in the number of dorsal scale rows 

 and inferior and superior labial plates in the prairie region. (The 

 mean number of subcaudals and ventrals, subject, as they are, to a 

 considerable range of sexual and individual variation, can not be 

 determined in the small series at hand.) However this may be, one 

 thing at least is certain, and that is that radix, as a form, does not 

 have a constant number of scale rows and labial scutes. This is 

 shown not only by the frequency of the variations that occur, but 

 also by the fact that the dorsal scale formula is so frequently 19-21- 

 19-17 (which is a transition stage between the formula 21-19-17 and 

 19-17) and the fact that generally when the labial formula is 8/10 

 the third superior and the corresponding inferior labials are notice- 

 ably reduced. This is significant, as it indicates, we believe, that the 

 form is a dwarfed offshoot of a stock with a larger scutellation. 



The variations in color have but thrice been thought of subspe- 

 cific value. In general, as previously described, the ground color is 

 brown, with three stripes ( the lateral being on the third and fourth 

 rows and the dorsal on the median and halves of adjacent rows), 

 and three rows of black spots on the skin and involved scales. In 

 western specimens, from Kansas and Nebraska, the ground color 

 is light brownish olive, and the black spots are consequently very 

 distinct. The dorsal stripe is very conspicuous, being usually bright 

 orange yellow, and often covering more than the median and halves 

 of the adjoining rows, the lateral stripe being usually pale yellow and 

 less conspicuous. The interspaces on the skin between the spots are 

 usually whitish. This coloration was described as a species (haydeni) 

 by Kennicott in 1860 on the basis of a specimen from Fort Pierre, 

 Nebraska, that was further characterized by having the interspaces 

 on the skin between the lateral spots red instead of the usual whitish. 

 This color in the type specimen has now faded, but a specimen in the 

 possession of the Academy of Natural Sciences of Philadelphia (No. 

 16619) from Peabody, Kansas, shows the same coloration. In this 

 individual the interspaces on the skin and the edges of the involved 



