REVISION OF THE KING SNAKES. 253 



throughout that environment. Thus it has come about that the 

 direct relative of any form is found in an adjacent environment. 

 The peculiar exception offered by California e (p. 94), where specia- 

 tion by mutation is indicated, affords opportunity for a most inter- 

 esting investigation. But in general, the oft emphasized rule holds 

 for this genus; namely, the occurrence of two different forms of the 

 same genus in the same locality is prima facie evidence that they 

 belong to different lines of descent, and, conversely, directly related 

 forms are not found in the same locality. 



The preceding might be detailed more fully, and more complete 

 information on the little known forms would doubtless result in 

 added illustrations of the fact that the environment seems to exert 

 a circumscribing influence on the species, that is, the range of the 

 species is determined by the environment, and, that when the latter 

 changes, the species is altered or dies out. 



The obvious differences in the relative ages of certain forms and 

 groups of forms in the genus Lampropeltis, as shown by specializa- 

 tion, structural isolation and diversification, geographic position, etc., 

 leads naturally to speculation as to the time relations of their spread 

 from the Southwest. Since triangulum is typically developed only 

 north of the southern limit of glaciers, we may assign it to a post- 

 Pleistocene migi'ation. The range of syspila and the northern por- 

 tion of the range of gentilis were doubtless likewise largely uninhabi- 

 table during the glacial epoch, and tliis together with the perfect 

 intergradation between these forms indicates that they belong to the 

 same "wave" of dispersal. The high specialization of elapsoides 

 and its distinctness from its nearest relative argues for isolation, and 

 greater age than that of triangulum and syspila. These demands 

 are satisfied by assuming it to have reached the Southeast at some 

 favorable time during or before the ice age, and to have been isolated 

 by an embayment of the lower Mississippi Valley during the Pleisto- 

 cene (see map in Willis and Salisbury, 1910, 277). The high degree 

 of specialization and structural isolation of pyrrJiomelaena suggests 

 that it is as old as any form in the genus. As it certainly has no 

 relatives outside of the Southwest, it may be regarded as indigenous 

 to this region, and its structural isolation is doubtless due to an early 

 differentiation from the primitive stock. In the CALLIQ-ASTER 

 group the apparently sharp separation and the specialization of its 

 eastern representative, make natural an assignment to a period 

 similar to that when elapsoides reached the Southeast. Consequent 

 separation during a greater or less portion of the Glacial Epoch 

 induced divergent differentiation. The GETXTLUS group may be 

 considered nearly or quite as recent as the varieties of triangulum, 

 with, perhaps, greater age indicated for the western representatives 



