SNAKES OF THE GENUS PITUOPHIS 205 



4. When an evident geographic variation is present it is in general 

 continuous from the probable center of distribution for the form to 

 the periphery of its range. 



5. In every form represented by a sufficient series of specimens the 

 variational tendency in scale rows, ventrals, and caudals is similar; 

 and in labials and oculars, as well, when these show a recognizable 

 geographic variation. When an evident geographic variation occurs 

 in the proportionate tail length, it is the reverse of that observable 

 in the scale characters. The number of tail spots varies in correla- 

 tion with the number of caudals, but the body spots show a con- 

 tinuous variation only in exceptional and contradictory cases. 



6. In all forms of the genus a general decrease in scale characters 

 from the probable center of dispersal to the periphery of the range 

 of the form is evident, and is accompanied by an increase in pro- 

 portionate tail length. 



7. Within a given form a general tendency to decrease the scale 

 characters is indicative of and correlated with a general dwarfing of 

 the form. Between forms, however, a general decrease in the scale 

 characters may be correlated with a general dwarfing, but is not 

 necessarily so, as is shown by the decrease from sayi to the subspecies 

 of melanoleucus, wliich latter forms are on the average the largest of 

 the genus. 



In the foregoing generalizations a knowledge of the center of origin 

 for the various forms is assumed. The first step in the task of attain- 

 ing this knowledge is the determination of the probable center of 

 dispersal for the genus as a whole. Various rules for the determina- 

 tion of the center of dispersal for a group of related forms have been 

 proposed by different authors. 



The ten criteria of Adams (1902, p. 115) have been given a wide 

 appUcation, and tbe repetition of them here seems unnecessary. 

 Only three of them can be considered to be of value in this study, 

 since of the others one applies only to the migration of birds, and of 

 the other six "none of them would necessarily hold true for the place 

 of origin after the time of origm" (Dunn, 1926, p. 8). These three 

 tests are: 



1. Continuity and convergence of lines of dispersal. 



2. Continuity and directness of individual variations or modifications 



radiating from the center of origin along the highways of dispersal. 



3. Direction indicated by biogeographical affinities. 



A brief discussion of each of these in relation to this genus may be of 

 value. 



1. Continuity and convergence of lines of dispersal. — The several 

 evolutionary lines leading to s. sayi and the forms of melanoleucus 

 in the east, to the deppei group in the south, to vertebralis in Lower 

 California, and to the three subspecies of catenifer in the west and 



