REVISIOlSr OF THE KING SNAKES. 237 



SUMMARY. 



Included in this group, on the basis of intergradation along the 

 common boundaries of their ranges, are, nelsoniy annulata, gentilis, 

 amaura, syspila, and triangulum; by reason of close similarity with 

 the first two, polyzona and micropJiolis; elapsoides and virginiana on 

 account of close resemblance to arnaura; because of apparently 

 closer relationship to this than to any other group in the genus^ 

 ruthvenij multicincta, and pyrrhomelaena . 



As already shown, micropJiolis can be derived only from polyzona. 

 The most evident reasons for tliis are (1) it is closely allied to this 

 form and to this one only; (2) it is decidedly isolated geographically 

 from all other forms of the genus; (3) it is inadmissable as a primi- 

 tive type on account of specialization in important structural and 

 color pattern characters, particularly {a) in reduction of anterior 

 temporals to one, (Jb) in the frequency of entire caudal plates, and 

 (c) in the white rings of the pattern being much widened and suffused 

 with red. Polyzona must be regarded as closely allied to annulata 

 and particularly, perhaps, to nelsoni, but its exact relationship is 

 imcertain, due to insufficient material. 



• The evolution of elapsoides from amaura has already been explained 

 under the discussion of the former, and we need not do more here 

 than repeat our belief that it represents an earlier and less extensive 

 migration than that which gave rise to the subspecies of triangulum; 

 that it underwent differentiation and specialization through an 

 isolation which may have been affected by a not very extensive 

 embayment of the Gulf of Mexico up the lower Mississippi Vuliey. 

 Its close ally, virginiana, was differentiated from elapsoides by a 

 joining of the adjacent black rings on the belly across the red spaces, 

 and their lateral migration, thus restricting the red to wide dorsal 

 saddles. This, as will be shown below, is almost exactly the same 

 method as that by which syspila was derived from amaura. 



If now we consider the subspecies of triangulum, we find that the 

 structural differences, separating any two adjacent forms are only 

 average and so quite insufficient to distinguish them, and even ex- 

 treme forms are but moderately distinct in structure. The differ- 

 ences are thus too slight to be a main reliance in determining the 

 center of origin of the series. This structural similarity is, however, 

 of much value in assuring us of the close genetic relationships exist- 

 ing between all these forms, and, if we are able by other means to 

 determine direction of relationship, structural tendencies will then 

 be of value in either strengthening or weakening those conclusions. 



Aside from the geographic demands which, must be met, our main 

 reliance must be upon color pattern. This, as well as the scutella- 

 tion, demonstrates the close relationship of the whole series and in par- 



