EAST AFRICAN MAMMALS IN NATIONAL MUSEUM. 83 



otherwise white undcrparts. A subspecies, geographically inter- 

 mediate, but most closely resembling the Taita Hills form, was 

 named by Heller from the Ulukenia Hills (ulse). The identity of 

 this rock rat with the Mvs fvmatvs of Peters was naturally unsus- 

 pected, but when Heller examined the types of East African mammals 

 in Berlin he discovered that the type oi fumatus is a young example 

 of the East African form of vcrroxi, the niveiventris of Osgood. His 

 manuscript notes on this type-specimen are as follows: 



Mus fumatus Peters. Type No. 5277 Berlin Mus. Skin mounted. Skull with 

 occipital part and floor of hraincase lost. Very young; last molar just erupted. Lo- 

 cality Ukamba (probably Kitui). Hind foot, dry, from specimen, 20. Ears and all 

 of tail except basal 2 inches lost. Color peculiar broccoli brown, belly and feet pure 

 white. Interorbital region rounded, not beaded. Upper molars, 5; diastema, 5.7; 

 nasals, 9.5 by 3; interorbital breadth, 3. 7; condylo-incisive length of mandible, 16. 



In the original description of fumatus, Peters gave the length of 

 tail as 80 and the color of the underparts as "grauweiss." The 

 immaturity of the specimen would account for both these differences 

 from adult specimens of the East African representative of Battus 

 verroxi, but Heller has noted the bell}^ of the type as "pure white." 



I can not find characters by which to recognize three forms of 

 this species in British East Africa, and although Ukamba specimens 

 are somewhat intermediate in color between the two extremes, they 

 are clearly nearest to the white-bellied subspecies of the Taita Hills. 

 The skins from Mount Gargues average slightly grayer above than the 

 specimens from Ulukenia Hills and other localities, but since tho 

 series is small and the differences trivial, it does not seem advisable 

 to separate them from B. fumatus fumatus. They have tho hairs of 

 tho underparts as clearly white to the bases as do the skins from 

 Taita Hills. The range of the animal probably will be found to bo 

 practically continuous in rocky localities when more collecting has 

 been done east of Mount Kenia. 



Heller found four embryos in a female collected on Mount Gargues, 

 September 1 ; and two embryos in a fem.ale from Ndi. Novemxber 3. 



For detailed measurements of this subspecies and other forms of tho 

 subgenus Muomys see page 84. 



RATTL'S FUMATUS SUBFUSCUS (Osgood). 



1910. Mus niveiventris subfuscus Osgood, Field Mus. Nat. Hist. Zool. ser., a^oI. 



10, No. 2, p. 12. February. (Lake Elmenteita, British East Africa; 



type in Field Mus. Nat. Hist., Chicago.) 

 1910. Epimys nieventris ulae Roosevelt, African Game Trails, Amer. ed., p. 



478; London ed., p. 489. (Part; references to Sotik, Naivasha, and 



Rift Valley.) 



Specimens. — Seventy-one, from localities as follows: 

 British East Africa: Engare Narok River, 1 (Loring); Lake 

 Naivasha, south end, 5 (Mearns, Loring); Naivasha Station, 27, 

 including 6 in alcohol (Loring, Mearns); Oljoro O Nyon River, 4 



