REVISION OF THE KING SNAKES. 107 



Again, it will be noticed ihat Jloridana has changed from getulus by 

 an increase in number of ventrals. Since evolution by increase must 

 therefore be considered as a possibility, we can base no argument for 

 direction of evolution upon ventral plates alone. We may simply 

 note, for later reference, that the forms exhibiting the lowest scale 

 formulae also have the smallest number of ventral plates. 



Variation in number of caudals and labials is of little use in itself 

 in this discussion, but it is noticeable that the lowest numbers occur 

 in the forms having the fewest ventral plates and the lowest scale 

 formulae. Changes in the skull and dentition have been too slight 

 to be of much use in tracing relationships within the group. This 

 very constancy will, however, be of value later in connecting this 

 group closely with the calligaster group. 



The loreal shield is markedly different in the extremes of the group. 

 On the east coast it is square or decidedly higher than long, while on 

 the west coast it is usually longer than high, the dimensions here 

 rarely approaching equality. The intermediate forms are inter- 

 mediate in respect to the shape of the loreal, but TiolhrooJci is more like 

 getulus (fig. 4) and splendida (fig. 3) is more like hoylii. If the getulus 

 group is correctly associated with the other forms of Lampropeltis, 

 the shape of the loreal in the eastern forms is a decidedly speciahzed 

 instead of primitive condition. This is a very good reason for look- 

 ing west for the origin of the group. 



Like the loreal, the relative length of the posterior chin shields is 

 decidedly different in the extremes of the group. In the eastern 

 forms they are approximately as large as the anterior shields and 

 usually in contact with each other, while m hoylii and its allies they 

 are hardly more than half as large as the anterior ones and are 

 separated usually by two smaller scales. Splendida here occupies the 

 intermediate position. Its western individuals are like hoylii and 

 its eastern show a decided approach to the getulus condition. The 

 nearest allies to this group are calligaster and rhomhomacuJota, and 

 here the posterior chin shields are usually a little shorter than the 

 anterior and tend to be separated by about one scale, an arrangement 

 that is true as well of the triangulum group. Both coast forms 

 therefore represent extreme modification in the shape of the posterior 

 chin shields, and it is splendida that presents the closest approach in 

 this character to the other groups of the genus. 

 186550— 21— Bull. 114 8 



