86 BULLETIN 99, UNITED STATES NATIONAL MUSEUM. 



(Loring, Heller); Southern Guaso Nyiro River, 32, including 6 in 

 alcohol (Loring, Mearns); Telek River, Sotik. 2 (Heller). 



This is the subspecies oifamatus with the gray undercolor to the 

 hairs of the otherwise white underparts. Around Naivasha the 

 form is well marked, but specimens from the Southern Guaso Nyiro 

 and Sotik sometimes have very little of the gray undercoloring on 

 the belly, and resemble the Taita HiUs and Ukamba skins. There 

 are no pronounced differences in size or any other characters of 

 consequence beyond this coloration of pelage to separate the two 

 subspecies. 



Never found except among rocks; we always found it where there were cliffs or 

 stony koppies. Lives in crevices in the rocks and along the ledges of the cliffs. Noc- 

 turnal. (Roosevelt, African Game Trails, p. 478.) 



RATTUS TANA (True). 



Plate 22. 



1893. Mus tana True, Proc. U. S. Nat. Mus., vol. 16, p. 602. October 25. (Along 

 the Tana River, between the coast and Hameye, British East Africa; 

 type in U. S. Nat. Mus.) ' 



1909. Mus tana Lyon and Osgood, Bull. 62, U. S. Nat. Mus., p. 155. January 28. 



Specimen. — One, the type: 



British East Africa: Tana River, between the coast and Hameye 

 (Chanler and von Hohnel). 



The type-specimen of Rattus tana was originally preserved in 

 alcohol, but was made up into a skin after about one year of immer- 

 sion. The coloration is evidently considerably modified by this 

 treatment and satisfactory comparison with other skins is impossi- 

 ble. True described it as "brownish-gray above, hoary below; 

 feet white," which is about as accurate an account as can be given. 

 There is a distinct undercolor of gray below, much as in many skins 

 of R. f. suhfuscus, but all the hairs of the underparts are broadly 

 tipped with a light color which was presumably white before its dis- 

 coloration. The tail in the dry skin measures 85 millimeters in 

 length, but was recorded by True as 93, in the alcoholic specimen 

 before skinning. The skull shows the type to be an adult specimen 

 and has the teeth moderately v^^orn. The measurements as given 

 by True are probably of little value, as the specimen is much shrunken 

 and was probably preserved in spirits much too strong for the pur- 

 pose. The tail in the dry skin has been separated from the body 

 and sewed in place, so that it is obviously considerably shortened 

 from its original condition. 



A careful comparison of the type of R. tana with all the forms of 

 Rattus from East Africa in the collection proves it to be almost 

 certainly a member of the subgenus Myomys closely related to 

 Rattus fumatus. Until a satisfactory series of rats of this group is 

 collected from the Tana River country it will be impossible to settle 



