REVISION OF THE KING SNAKES. 289 



The races of triangulum form a continuous chain from southern 

 Mexico to southeastern Canada. Some form in the series must be 

 closer to the ancestral type than all the others. Let us consider 

 the northern end of this chain. 



If we start with typical triangulum, as it occurs in western New 

 York, and examine a series of representative specimens from localities 

 between here and Central Florida, we will notice a very gradual 

 change in pattern in Pennsylvania, West Virginia, and southward; 

 the half collar behind the neck very frequently replaces the chevron 

 mark; the latter, when present, is more imperfectly shaped than in 

 the north, and the other characteristic head markings are often 

 very badly formed. Throughout this wide extent of latitude we 

 will find no distinctive change in triangulum and no closely allied 

 form — nothing closer than elapsoides, getulus, and rlioinbomuculata. 

 If we now examine a representative series of specimens from northern 

 Indiana and Illinois southward into Louisiana, we mil find a very 

 different situation. In southern Illinois and southern Indiana 

 triangulum intergrades perfectly with syspila. The latter is spotted 

 like triangulum but the spots are typically longer, wider and fewer; 

 normally they are red, but occasional examples have them brown 

 or gray as in triangulum. The head pattern of the latter (fig. 54) 

 rarely appears in full in syspila (fig. 52) but portions of it are usually 

 present, as, for instance, the anterior black border of the half collar 

 behind the neck is sometimes indented, indicating the position of 

 the arms of the chevron mark of triangulum; the half collar may 

 be more or less interrupted by the approximations of its black borders 

 on one or both sides of the median line; light superciliary spots v/ith 

 black borders are often present over the eyes, and these are some- 

 times joined, as in triangulum; the light postocular band is often 

 represented by lateral indentations of the anterior black border of 

 the half collar and sometimes it is present in entirety. As we go south 

 and west from southern Illinois the frequency and perfection of these 

 markings decreases. To one who has examined considerable num- 

 bers of these specimens the geographic change is most striking. 

 Typical triangulum (figs. 54 and 69) occurs about as far south as 

 Vigo County, Indiana; at this point there is a relatively rapid change 

 through intergrades to syspila (figs. 52 and 68); then south into 

 Arkansas we find a gradual, progressive change which leads directly 

 into amaura (fig. 67), in the southern part of the state. That syspila 

 is the only close relative of triangulum, and that the former passes 

 by perfect intergradation into amaura, can not be denied. One of 

 these forms must then be ancestral to the others. 



It has been said that the color pattern of triangulum must be re- 

 garded as primitive, because its distinctive features are repeated in 

 other forms in widely separated groups, as, for instance, the calli- 



