176 BULLETIN 99^ UNITED STATES NATION.VI- MUSEUM. 



British East Africa: Kakumega, 1 (Heller); Mem Road. 

 Laikipia, 1 (K. Roosevelt) ; Mount Kenia Forest Station, 1 skull 

 (Mearns) ; Naivasha, 1 skull (Loring) ; Suswa Plain, Sotik, 1 (Heller) ; 

 Telek River, Sotik, 1 (HeUer). 



There is considerable variation in tone of ground color and in the 

 pattern of markings within tliis small series and it is quite evident 

 that a large number of specimens of the East African serval must be 

 assembled before satisfactory results in the cUstinguishing of sub- 

 species can be assured. The characters used by Wroughton to sepa- 

 rate liindei from Felis capensis Icempi,^ described from Kjrui, Mount 

 Elgon, at 6,000 feet,^ are apparently not of much use in distinguish- 

 ing forms. Adult male skulls of Mndei in our collection are as large 

 as the dimensions given for Icempi, and it is plain that no great differ- 

 once in size between these two forms exists. I have seen no speci- 

 mens from Elgon; but can not distinguish our Kakumega specimen, 

 a young adult female, from other East African skins and skulls of 

 Mndei by any differences of subspecific value. 



The Gondokoro skin is slightly more cinnamon colored above than 

 are any of the British East African skins, and the spots and stripes 

 are more broken than usual; but, bearing in mind the great variation 

 known to exist in mammals marked as are the servals, it seems very 

 unwise to recognize a new form based on color alone without a suffi - 

 cient series of specimens to prove the constancy of the variation. 

 The skin from Meru Road has been mounted and therefore is not 

 comparable with tanned skins; the general appearance and size of 

 markings is much changed by the necessary stretching from the 

 shrunken state usual to tanned skins. 



Two specimens from Beira, Portuguese East Africa, and one from 

 Concession Hill, Mashonaland, in the collection, indicate that the 

 form described by Wroughton from Beira, Felis capensis leirse, is a 

 recognizable subspecies. The color differences are shght but the 

 teeth average larger and stouter. The general size of the animal is 

 not much, if any, greater than that of F. c. Mndei. 



The stomach of the Kakumega specimen is recorded by Heller to 

 have contained small rodents — 1 Dendi^omus and 3 Leggada. 



For the use of the specific name capensis Forster, 1781, in place of 

 serval Erxleben, 1776, see a paper in 1910 b}- Wroughton.^ 



For measurements of specimens see table, page 177. 



» Anu. and Mag. Nat. Hist., ser. 8, vol. 5, p. 206. February, 1910. 



' "Kiru villages are 6 miles below caves on the south slope of Elgon."'— E. Heller, MSS. 



3 Ann. and Mag. Nat. Hist., ser. 8, vol. 5, p. 205, 1910. 



