370 Mr. 0. Thomas on a 



Mandera, East Africa, it is distinguishable by its broadly 

 naked rump, that region in nigriseta being about as hairy as 

 in ordinary Petrodromus. 



Rhynchocyon. 



Since in Rhynchocyon there are three conspicuously diffe- 

 rent types of coloration — those of the cirnei, petersi, and 

 chrysopygus groups respectively, — it seemed likely that a close 

 examination of the skulls would indicate the advisability o£ a 

 triple subdivision of this genus also. Such, however, proves 

 to be only partially the case, for the natural grouping would 

 appear to be into two, as follows : — 



Rhynchocyon, s. s. 



Back with a chess-board pattern ; ground-colour some shade 

 of grey, rarely a little rufous. 



Outer upper incisor normally deciduous. 

 Genotype. R. cirnei, Peters. 



Rhinonax, subgen. nov. 



Back without chess-board pattern. Ground-colour deep 

 rufous or chestnut, the posterior back yellow or black. 

 Outer upper incisor normally permanent. 

 Genotype. R. chrysopygus, Giinth. 

 Also contains R. petersi and its allies. 



The fact that the character of the permanence or otherwise 

 of the upper incisor is not absolutely constant in either sub- 

 genus, a small proportion of the skulls of each providing 

 exceptions to the rule, induces me to treat these groups 

 merely as subgenera, and not full genera. 



I may take this opportunity formally to select B.M. 

 no. 80.11.30. 7, with its skull 1758 a, as a lectotype of 

 R. {Rhinonax) chrysopygus, the other two co-types mentioned 

 by the author thus becoming lecto-paratypes. 



XLIV. — A new Wild Dog from the Bogota Cordillera. 

 By Oldfield Thomas. 



(Published by permission of the Trustees of the British Museum.) 



The British Museum owes to the Rev. Prof. Apoll maris 

 Maria, of the Faculty of Medicine, Bogota, an example of 

 a wild dog from the eastern slopes of the Bogota* Cordillera, 

 and this appears to me to represent a species not hitherto 

 described. 



