THE SOMALI LUK-DIKS 979 



who took as the type a specimen obtained by Mr. Lort Phillips 

 at Dobwein, 40 miles south of Berbera. 



This spot I have been unable to locate, so can only conclude, 

 after carefully examining the type specimen, that it was some- 

 where along the Golis foothills or, perhaps, just on the top of 

 Mirso, which I will show is the southerly limit of the coast 

 or Guban variety, to which I gave subspecific rank in July 1 909 * 

 and called M. phillipsi giibanensis. 



This subspecies is found all along the coast-belt of British 

 Somaliland, passing northwards into French Somaliland, certainly 

 as far north as Djibouti, but how much farther I have been 

 unable to ascertain, and eastwards towards the Mijertain country, 

 but how far I have not yet determined, owing to there being no 

 specimens available from that area. 



"What I propose to designate as the ti'ue M. phillipsi is the 

 brilliant rufous or dark cinnamon-flanked Dik dik found through- 

 out the interior of British Somaliland from, roughly speaking, the 

 Golis Range away to the south and west into the Hand, where, 

 especially in the west, the animal's flanks are so red that at 

 a short distance it looks rufous all over ; this is in marked 

 contrast with the coast variet}', which looks quite grey. 



As one passes through the hilly country from Jig-jigga to 

 Harrar and also to the west and south-west of the former, 

 one finds another and darker Dik-dik, which in its wild state 

 looks of a dark red-brown colour. This is the Dik-dik to which 

 Mr. Neumann gave the name 31. hararensis, and so distinct does 

 it appear in life from M. phillipsi that it fully deserves a name so 

 that it might not be confounded with M. phillipsi ; but in my 

 opinion it deserves only subspecific rank, and should be called 

 M. phillipsi hararensis (PI. LV. fig. 2). It is to be expected that 

 in localities where the soils are so distinct as in the coast-belt 

 and the Hand one would see some variation in the coloration of 

 the pelage, and here in this species we have this variation very 

 distinctly marked ; but this is not the only diffei^ence, another 

 equally potent factor, in the shape of altitude with its accom- 

 panying variations in temperature, also assists in increasing 

 the difierence between the type species and its subspecies, by 

 i-endering the pelage thicker. 



For instance, the hair in the Guban variety is scantier and 

 shorter than in the Harrar and neighbourhood specimens ; in 

 the former the individual hairs from the dorsal region of the 

 back measure from 13-23 mm., whereas in the latter variety they 

 are usually 30 mm. or more. This difierence renders M. phillipsi 

 guhanensis a much more sleek-looking animal (PL LV. fig. 1). 

 The pelage of the true M. phillipsi, which is the common Dik-dik 

 found all over the interior of the " Horn of Africa," is inter- 

 mediate between the two. 



The skull-measurements in all three are, as one would expect, 

 practically identical. 



* Ann, & Mag. N. H. ser. 8, Vol. 4. p. 49. 



