110 



DR. ST. G. MIVART ON 



[Mar. 18, 



premolar is very small, and the two upper molars compared with 

 the fourth u|)pcT premolar are only as 122 to 100. But in the speci- 

 men at the British Museum now referred to, the upper molars are 

 very large, bearing to the fourth upper premolar a proportion of 160 

 to iOO, or just about the proportions exhibited by skulls of C. vetulus 

 and Burmeister's figure of the dentition of that species. It cannot 

 therefore be C. brasUieiisis. 



Lund's C. brasilieiisis much more resembles C cancricorus, as 

 Burmeister (in his ' Fauna ') took it to be, though later (Archiv f. 



Fis. 6. 



Surfaces ol' iipp;)r uiulni-!; 

 Canis iirostirtus. 



Surliioes of molar leetb of lower 

 jaw of Cani» urostictiis. 



Natur. ii. vol. i. p. 1 20) he was more inclined to regard it as distinct 

 from that species. Certainly the aspect of the figure given by Lund 

 differs considerably from any skull of C. cancrivonts I have noticed, 

 and its outline reminds one a good deal of that seen in the genus 

 Cijon. In the proportions of the teeth, however, it is like 0. can- 

 crivorus. In four skulls of the last-named species I find the average 

 length of !'■ -1 is 1 -27 and that of M. 1+M. l' is 1-57, or as 100 to 123, 

 with which Lund's C. brasiliensis almost perfectly agrees. 



It is interesting to note that the C. microtis of Sclater (fig. 7, 

 p. 1 1 1)^ shows its affinity in this res]iect to V. cancrivorus, although I 

 believe it to be a distinct species, P- -1 being 1-30 and M. 1+M. 2 being 

 1-65, or as 100 to 12(i. 



As for the skin and skull (1033 e) in the British Museum, which 

 cannot be C brasilie/isis, it is also destitute of the characters ascribed 

 to either C. vetulus or C . fulvicaiidus of Burmeister, while its skull 

 and dentition are so peculiar that it demands to be marked off as at 

 least a distinct variety, possibly a species. The most distinct 

 external mark about it is a longitudinal black stripe along the 



' r.Z. S. 1S82, p. (531, pi. xlvii. 



