502 



ME. G. A. BOrXENGEU ON THE 



[May 18, 



number, one or t\po, of canthal scales as a specific character, I 

 have divided accordingly the specimens in the following table. 

 One, which is here figured, defies the system in being referable 

 to the first category on the right side, to the second on the left. 



Kg. 3. 



Head of Sccloporus formosus from Costa Eica, 



I have placed it last in the first division. It will be seen that the 

 character is not constantly correlated with any other, and for that 

 reason I must refuse to attach any special importance to it. Of 

 S. irazuensis, Giinther says : — " Two canthal scales, but owing to 

 the shortness of the snout the anterior is not always fully 

 developed." I cannot account for such a statement ; there is no 

 correlation between the number of cauthals and the length of the 

 snout, as may be seen from the specimen figured above, one of the 

 types of S. irazuensis. Giiuther himself does not seem to have had 

 o-reat faith in the value of the character, as he has in several 

 instances associated in the same species individuals with one 

 canthal and others with two, although those exceptions are not 

 alluded to in his descriptions. Curiously, Giinther's S. smarag- 

 dinus is described as with one canthal scale (p. 69), and yet 

 S. tceniocnemis, Cope, is referred to it as a synonym in spite of 

 Cope attributing to it two canthals. 



A. One canthal shield. 



(1) Referred by Giinther to S. irazuensis. 



(2) Referred by Giinther to S. lun<si. 



(3) Referred by Giinthei- to S. sntaragdinus. 



