106 DR. ST. G. MivART ON [Mar. 18, 



vice versa, I cannot but look with much scepticism on the specific 

 distinctness of these forms. 



Burmeister describes (' Fauna Brasiliens,' p. 37, pis. xxiii., xxviii., 

 and xxix.) a specimen in his possession which he regards as identical 

 with the C. vetulus of Lund, which species he (as before said) also 

 identifies with the G. azarce of Wied. But his description and his 

 plates show that an important distinction exists between what he 

 calls C. vetulus and both the G. vetulus of Lund, and Wied's and 

 "Waterhouse's C. azarce ; for its fourth upper premolar is extremely 

 small, while the two upper molars are relatively very large. He also 

 gives their dimensions (' Reise durch La Plata,* p. 407) as follows : — 



Length of iVi=9 ; length of M. 14-M. 2 = 14, or as 100 to 155. 



But Lund gives an apparently careful and accurate figure of the 

 skull of his C. vetulus, the type of the species, and this shows a 

 well-developed fourth premolar and a small molar, which teeth 

 bear to each other the proportions of 100 to 130. 



Now my experience is that though the proportions of the teeth 

 are not constant, they yet afford better characters than do varia- 

 tions of tint in the fur — a condition often variable with the season, 



I do not think that the C. vetulus of Burmeister can be the same 

 as the C. vetulus of Lund ; and if it is not, it must be distinguished by 

 some other appellation. As to what the latter may be, Burmeister 

 identifies it with the C. azara of Wied, and therefore he ought not 

 to call it C. vetulus, but what he regards as its original denomination, 

 G. azarce ; and this, for all we can see, it may be, and I am disposed 

 to think that it is a pale variety of it, judging from Lund's represen- 

 tation of its external form. 



Now in the British Museum there are two skins and three skulls^ 

 from Brazil, which appear to me to belong to the same species as 

 that described by Burmeister under the name C. vetulus. Its 

 external characters fairly correspond with those of Burmeister's 

 form, but its dentition appears to me to weigh heavily in favour of 

 their specific identity. 1 find the P^J to be -7, and M. 1+M. 2 to be 

 1-20 in two skulls, and in the remaining one, P- 4is'75 and M. 1+M. 2 

 is ri5. The average of the three is therefore 7'1 and 1*18, or as 

 100 to 166. 



It is interesting to note that the three skulls referred to differ 

 amongst themselves in the form and development of the sagittal 

 elevation and in the shape and proportions of the frontal postorbital 

 processes. 



As to Lund's C.fulvicaudus, Burmeister remarks (' Fauna Brasi- 

 liens,' p. 40) that it seems to be very near Lund's C. vetulus, but is 

 distinguished by its smaller stature, blunter head, and proportionally 

 stronger build, clearer and more yellow ground tint, and rusty tail with 

 black end, ochre-coloured patch behind the ear ^ ; finally the front of 

 the arms and the hind legs, above the knees, are darker. 



^ Nos. 821 A, 821 B, and 821c. The first of these is extracted from the skin 

 No. 44. 3. 7. 3. 



^ The mounted skin of C. vetulus in the British Museum has an ochre patch 

 behind the ear, but has not the characters otherwise attributed to C. fulvicaudus. 



