REVISION OF THE KING SNAKES. 251 



the latter rather than with the eastern that the closest relationship 

 lies. Of the TRIANGULUM group the outlying types. micropJiolis, 

 elapsoides, virginiana, and tnanguhim, are much more differentiated 

 from the other groups than its southwestern representatives, poly- 

 zona and annulata.. This criterion is therefore admirably fulfilled 

 by the Southwest, and by this region only, and its value as a criterion 

 may be regarded as second only to the sixth, 



4. Location of maximum size of individuals. — Apparently no greater 

 size is obtained by any forms than that reached by the soutliwostern 

 representatives of all the groups, with the exception of the purely 

 peninsular and derived forms, jioridana and hrooTcsi; and it will be 

 remembered that the forms which are most reduced in size and 

 scutellation occupy regions most distant from the southwest, namely 

 conjuncta of the Cape region of Lower California, and elapsoides, 

 and rhombomaculata of the southeastern States. Of ndsoni and 

 annulata we have too few specimens to know what the normal or the 

 maximum size may be but polyzona of southern Mexico may properly 

 be compared with triangulum of the northeastern States. Of 300 

 specimens of the latter the largest measured 1,085 mm., and of 61 of 

 the former the largest measured 1,610 and the next largest 1,580 

 mm., a decided difference in favor of the Southwest. The forms of 

 the GETULUS group are of approximately equal size; of the two 

 well-known forms of the CALLIGASTER group, the western repre- 

 sentative is decidedly the larger. This criterion therefore points 

 definitely to the Southwest. 



6. Continuity and convergence of lines of dispersal, — Both the 

 GETULUS and CALLIGASTER groups are represented in the 

 Southwest, and the isolated types, mexicana, alterna, and pyrrhome- 

 laena, are not represented outside of the Southwest. But it is the 

 TRIANGULUM group that illustrates this most admirably. In this 

 we see the convergence toward Mexico of the polyzonamicropholis 

 line from the south, the annulatorgentilis line from the north, the 

 annvlata-elapsoides line from the east, and the annvlata-triangvlum 

 line from the northeast. 



All genetic lines in the genus, therefore, either converge to the 

 Southwest, or are represented in the Southwest as well as elsewhere, 

 or ai'e not represented outside of the Southwest, and no genetic 

 lines converge to any region other than the Southwest. Thus this 

 most dependable of all criteria is satisfied only by the Southwest. 



7. Location of least deperulence upon a restricted habitat. — Too little 

 is known of the habitat relations of these snakes to discuss this cri- 

 terion profitably, but it may be noted that the only burrowing forms 

 (elapsoides, virginiana, and rhombomaculata) are located in the South- 

 east. 



186560— 21— BuU. 114 17 



