378; Mr. 0. Thomas on the Dwarf Man joose. 



H. undulata ] but no opportunity had occurred of inakiivj; a 

 detailed examination of the skulls and other characters. 0;i 

 now making such a comparison with Dr. Atkinson's speci- 

 men, I find that, although undoubtedly nearly allied, it is 

 sufficiently distinct to demand specific separation. I would 

 propose to call it, in honour of its collector, to whom the 

 Museum is indebted for an interestins; series of Somali 

 mammals, 



Helogale Atkinsoni, sp. n. 



" Very similar in external appearance to IJ. laidulafa, and 

 especially to the darker-coloured individuals of that species. 

 It is, however, rather darker than the darkest, with less 

 I'ufous or fulvous in the general tone, the chest and belly are 

 browner, and the muzzle is blackish brown, as compared to 

 the rufous which occupies this part in all the Museum speci- 

 mens of H. undulata, and is also mentioned by Peters as 

 being present in tlie types. 



. Skull with the nasals ))eculiarly short and broad, extending 

 backwards barely 2 millim. beyond the level of the pre- 

 niaxillary bones, instead of, as in //. undulata, running back 

 as a long narrow wedge past the level of the anterior edge of 

 the orbit. Middle upper premolar {p^) with scarcely a trace 

 of an internal secondary cusp, such as is prominent in all the 

 available skulls of 77. undulatuy five in number, and is men- 

 tioned by Peters in his description. 



In size 77. Atkinsoni appears to be just about equal tO' 

 H.iwdulata, for the type (a rather immalurc male) has a skull 

 of just the same size as an equally immature male of //. undu- 

 hita, and also just equal to the old female of that species and 

 the old male of 77. parvula. There appears therefore to be 

 just the same differences between the sexes of 77. U7idulata 

 as between the species undulata and loarvula^ an old male of 

 the latter just equalling an old female of the former. This 

 question of sexual dift'erence in size was not discussed by 

 I)r. Jentink when he suggested * that 77. undulata and^ 

 H. parvula were really indistinguishable by size and should 

 not be specifically separated. 



Dimensions of the type (a slightly immature male), 

 measured in flesh by the collector: — 



Head and body 192 millim.; tail 153; hind foot 41 j 

 ear 17. 



Skull: basal length 41'7; greatest breadth 25'6 ; nasals, 

 length in middle line 5'1 ; interorbital breadth 10*2 ; breadth 



* N. L. M. xi. p. 31 (1888;. 



