254 BULLETIN 114, UNITED STATES NATIONAL MUSEUM. 



to allow them to reach Cape San Lucas and Santa Catalina Island. 

 The Central American forms are less easy to account for. Nelsoni 

 assumed its specific identity before the Tres Marias Islands were 

 separated from the mainland. This is likely to have been pre- 

 Pleistocene. Poly zona shows too great variability to be easily 

 regarded as an ancient type, but it is old enough to have produced 

 the well-marked micropTiolis in Colombia and Ecuador, and is there- 

 fore without doubt older than the northern varieties of triangulum. 



The preceding paragraph may be summarized as follows: The 

 genus Lam'pro'peltis underwent a differentiation in the southwestern 

 portion of the North American continent in the late Tertiary, pro- 

 ducing a type that favored the mountains and uplands, the ancestor 

 of the TRIANGUIUM group, and another type that favored the low- 

 lands, ancestor of the GETULUS and CALIIGASTER groups. The 

 latter early differentiated into the progenitors of the two last groups, 

 and of these it was the CAIIIGASTER section that first spread north- 

 east and east. This must have occurred previous to the Pleistocene, 

 and at about the same time that the migration took place which 

 brought the progenitor of elapsoides to the southeast. During the 

 Pleistocene, the latter and the ancestral type of rhombomaculata 

 suffered an isolation from their direct relatives west of the present 

 position of the Mississippi River and were thus free to undergo con- 

 siderable differentiation. Following the Pleistocene, elapsoides ex- 

 tended its range south into Florida, there undergoing still further 

 specialization during the Recent Epoch, and west to Louisiana, 

 here meeting the form amaura, the post-Pleistocene representative 

 of its ancestral stock. If rliombomaculata extended its range into 

 southern Florida, it is yet to be found there. Calligaster, after the 

 retreat of the glaciers, was able to extend its range throughout the 

 prairie region northeastward, and north to the limit of its endurance 

 of cold. The western section of the GETTJITJS group may have 

 begun its evolution during or perhaps preceding the Pleistocene, but 

 it was only following that period that the spread east and northeast 

 occurred. And it was following the Ice Age, also, that the latest 

 forms of the TRIANGULUM group reached their present ranges 

 {amaura, syspila, triamgulum, and at least the northern section of 

 gentilis) , 



This historical outline corresponds with the present distribution 

 and structural relations of the king snakes, and with probable Pleisto- 

 cene geography. The genus as a whole seems to be relatively recent, 

 but the question of its origin is beyond the limits of the present study. 

 In this connection, however, we may say that there are apparently 

 no nearly related genera in South America, in the West Indies, nor 

 in Asia, and that it is entirely likely that its nearest relatives a.re to 

 be found in the genus Coronella of Europe and northern Africa. 



