REVISION OF THE KING SNAKES. 243 



Here the black failed to split on the midline, or, if it did split, it failed 

 to recede. The red failed to increase in area at the expense of the 

 black, so that in this form the black is usually wider on the middorsal 

 line than on the lower rows of scales and often excludes the red en- 

 tirely from here, as shown in figure 72 (a specimen from Provo, 

 Utah). Mottling of the white rings is variable in gentilis, but usually 

 present to some degree at least. 



Examination of the available material indicates that the split in 

 these two courses of evolution is not a clearly defined one, but that 

 the tendency to develop the pattern shown in figure 68 is gradually 

 intensified to the east. A fairly well-marked division is, howevei', 

 noticeable in the region of the ninety-seventh meridian. Here there 

 is a distinction in head pattern as well as in body pattern. To the 

 east the head is mostly red, and alternating spots are developed on 

 the lower rows of scales; to the west these are not developed and the 

 head is mostly black. The situation would be met if the ancestral 

 form spread from northeastern Mexico at the same time northeast, 

 north, and northv/est. 



The relationships of fyrrliomelaena and midticincta to each other 

 and to the other forms of the genus are still in doubt. There is little 

 if anything in the penial characters to ally them closely with the 

 getuluR group. It is true that the calyces are but few and slightly 

 fringed, but the spines are much more numerous and more slender, 

 and extend more than half way to the base of the organ, as is char- 

 acteristic of the triangulum series. Muliicmcta was described by 

 Yarrow as a subspecies of getulus and more recently has been regarded 

 by Stejneger and others as a subspecies of vyrrhomelaena. 



Cope confused it with both of these forms. It can hardly be a 

 close ally of hoylii, since their ranges overlap and their structural 

 differences are too great, particularly in respect to scale rows, penial 

 characters, and color pattern. Its relation to gentilis may be closer 

 than has been supposed. Typical patterns of gentilis and multicincta 

 (figs, 72 and 75) show that the white rings in gentilis are but little 

 widened in the most western specimens, perhaps an approach to 

 multicincta in which they are practically uniform in diameter; and 

 it is noticeable that in the latter form, as in gentilis, the black on the 

 belly is mostly concentrated o])posite the dorsal red areas. The 

 color pattern of multicincta could without violence be regarded as an 

 intensification of that of gentilis; that is, the black has still further 

 encroached on the red and has completely excluded the latter color 

 from the snout. The scutellation would be considered as a wide, but 

 perhaps not impossible, step. Thorough collecting in Nevada, eastern 

 California, and Arizona should be of much assistance in deciding this 

 matter. Scale rows and ventrals average at least as high as may be 

 expected for annulata in the plateau portion of its range, and higher 



