58 BULLETIN 114, UNITED STATES NATIONAL MUSEUM. 



variant in respect to pattern in every way similar to the specimens 

 of californiae from Fresno County, California. The belly is almost 

 an immaculate white on the anterior half of the body, but on the 

 posterior half it is mostly black, mixed with some white. The dorsal 

 pattern of cross bands is broken up into spots of varying shapes, 

 chiefly elongate, with a conspicuous tendency to form a dorsal white 

 stripe. It is to be hoped that more collecting in this vicinity will 

 reveal other examples. It indicates the inherent potentiality in 

 getulus for the production of an Atlantic coast analogue of californiae. 

 Affinities. — It will doubtless already have been noticed that ven- 

 trals, caudals, labials, scale rows, and color pattern all indicate the 

 same thing, that is, that the region represented by Georgia is the 

 meeting ground of two diverging lines of evolution in getulus. That 

 one line has been southward there can hardly be a doubt. Floridana 

 is clearly an end form in all its characters, and it is certainly so in 

 geographic position. Similarly getulus^ as it occurs in the northern 

 portion of its range, demonstrates itself to be a reduced form of the 

 southern type. Consequently evolution has proceeded from the 

 Georgia-Alabama region northward along the coastal plain, and 

 southward into Florida. In central Alabama and northern Georgia 

 getulus and niger intergrade in the region indicated above under 

 Range. The material at hand indicates that this is the only region 

 where intergradation occiu-s between these forms. The relationships 

 may therefore be expressed as follows: 



(New Jersey) 



niger getulus 



(Georgia) 



floridana 



hrooJcsi 



