CROCODILIANS, LIZARDS, AND SNAKES. 569 



Large; stripes more or less connected Avith spots which cut 



up the dark ground into spots and crossbars })08teriorly; 



femur without strip behind; femoral pores 20-21; loreal 



plate longer than high ; femoral scales 7-8 rows. 



C. graham ii. 

 yy. Post-antebrachial scales present. 



Median gular scales smaller than those of collar; femoral 

 pores 16-18; femoral scales in 8 rows; infralabials 6; 7 

 undulate black stripes on an olivaceous ground. 



G. septi7nritlatus. 



Median gular scales smaller than those of collar; femoral 



pores 18-23; muzzle elongate, loreal longer than high; 



dark bands interrupted by larger or smaller light spots or 



intervals C. (jularis. 



/3/?. Femoral pores fewer than 15. 



Femoral pores 12; 3 parietals; 3 supraorbitals; gray brown 



with 10 longitudinal stripes C. multilintatus. 



Femoral pores 9-11; 5 parietals; 4 supraoculars; anals forming 

 a triangle; olive brown with 6 stripes or some rows of spots. 



C. ocellifer. 

 aa. Anterior nasal plate reaching second labial. 



Femoral pores 13 ; one marginal anal plate; G white stripes; small. 



C. labial is. 



Ill this genus as in others, some characters which are constant in 

 one species are inconstant in another. The presence or absence of the 

 sixth iufralabial, and of the frenoorbital plates, are of this nature. 

 The number of femoral pores varies within a small range in all of the 

 forms. Anomalies in the division of the head plates are rare, but 

 sometimes occur in these genus. Such are the fusion of the symphys- 

 eal and postsymphyseal plates, the presence of an additional labial 

 plate, etc. 



The discrimination of the I^orth American species of this genus is 

 the most difficult problem in our herpetology. Nowhere are subspecies 

 more clearly defined than in Gnemidophorus, that is, definable geograj^h- 

 ical forms, which are not always true to their characters. 



The color markings ditfer in the same individual at different ages, and 

 the age at which the adult coloration is assumed differs in different locali- 

 ties. Some of the species, as for example, Cneniidophorus sexlineatus, 

 never abandon the coloration of the young of other species and subspe- 

 cies. The same condition is characteristic of the C. deppei of ^Mexico, the 

 C. lemniscatns of Brazil, and other species. The i)rocess of color modifi- 

 cation is, as I have pointed out,' as follows: The young are longitudi- 

 nally striped from two to four stripes on each side of the middle line. 

 With increasing age, light spots appear between the stripes in the dark 

 interspaces. In a later stage these spots increase in transverse diame- 

 ter, breaking up the dark bands into spots. In some of the forms 

 these dark spots extend themselves transversely and unite with each 

 other, forming black cross-stripes of greater or less length. Thus we 

 have before us the process by which a longitudinally strijied coloration 

 is transformed into a transversely striped one. 



Proc. Amer. Phil. Soc, 1885, p. 283. 



