EAST AFRICAN MAMMALS IN NATIONAL, MUSEUM. 87 



the exact status of R. tana. My own belief is that the form repre- 

 sented will prove to be a geographical race offumatus. On the other 

 hand skulls of certain specimens of fumatus are almost indistin- 

 guishable from skulls of some of the multimammate rats of the 

 subgenus Mastomys, and this is particularly true of specimens in the 

 "young adult" stage like the type skull of tana. The type of tana 

 is recorded as a female, but no trace of mammae can be found in the 

 dry skin. 



* RATTUS COUCHA ISMAIU/E (HeUer). 



Plate 23. 



1914. Epimys concha (sic) ismailix Heller, Smithsonian Misc. Coll., vol. 63, 

 No. 7, p. 9. June 24. (Gondokoro, Uganda; type in U. S. Nat. Mus.) 



Specimens.- — Thirty-six, from localities as follows: 

 Uganda: Gondokoro, 18, including 4 in alcohol (Loring); Ledgus, 

 1 in alcohol (Loring); Lombeki River, 1 (Loring); Mnyouri Jardin 

 1 (Loring); Nimule, 15, including 6 in alcohol and 1 odd skull (Loring, 

 Heller). 



The rats of this group, the subgenus Mastomys, known as the multi- 

 mammate species, are among the most difficult of small mammals 

 to arrange satisfactorily into species or subspecies. What appears 

 to be individual variation within a single form is usually far more 

 conspicuous than the characters marking valid geographical races. 

 The difference in size, and frequently in color, between fully adidt 

 examples and the older, aged individuals is very great; and the 

 change in the form and appearance of the skull tliroughout life is 

 extreme. It is only when large series are available and when the 

 skins and skuUs are arranged according to age, as determined by 

 the condition of the teeth, that the great variations are in a measure 

 explained. My .conclusions are tha't within the territory covered 

 by this report only a single species of Mastomys is found, and that 

 this species may be separated into several geographical races, no two 

 valid forms occupying the same region. With smaller series of 

 specimens one might easily be misled into describing several species, 

 based mainly on size, form of the skull, and the coloration of the 

 underparts. Size of body and size and form of skull are explained 

 in almost every case with our material by age; and the color of the 

 belly must frequently be accounted as purely individual variation, 

 except in cases when it is plainly a question of condition of pelage. 

 I have not seen authentic specimens of the two forms described by 

 Dollman from Lake Baringo^ which he found ranging together, and 

 one of which, the smaller effectus, he states is distributed over the 

 same territory occupied by Rattus coucha ugandse. I can not help 

 but beUeve, after my experience with the group, that these two 



1 Epimys effectus and E. evelyni Dollman, Ann. and Mag. Nat. Hist., ser. 8, vol. 7, pp. 524 and 526. 

 May, 1911. 



64952— 19— Bull. 99, pt 2—7 



