258 MESSES. OLDFIELD THOMAS AKD M. A. C. HINTON ON 



the presence of an extra cusp in 711 *' — wliich is far from constant, 

 and many specimens of Taie?-illus have a rudimentary cusp in 

 the same position. Major Graham has sent examples of Taterillus 

 from S. Kordofan, which we may take as representing both 

 ..kadugliensis a,nd lorentzi, and these quite agree with the type of 

 hutleri from Bahr-el-Ghazal. 



32. Taterillus clivosus, sp. n. 



6 . 509, 516, 522, 525, 539, 645, 707. $ . 519, 520, 521, 524, 

 647, 698, 699, 708, 1032. Jebel Marra. 4000-5500'. 



6. 770, 830, 847, 860. $. 820, 848. 850, 862, 864, 865, 

 869. Ivuhne, Wadi Aribo. 3300'. 



$. 726. Zalingei. 2800'. 



<S . 473. 35 miles W.S.W. of El Fasher. 



$ . 1102, 1106. Jebel Maidob. 



Like T. hutleri, but with longer and rather more bushy tail. 



Size about as in butleri, or a little larger. General colour 

 buff}^ brown, a little lighter than in hutleri, but decidedly darker 

 than in the plains form to be described below. Face without 

 special white markings. Ears of medium size, bufFy brown, 

 scarcely different from the back. Hands a.nd feet white, soles 

 blackish, generally with well-marked and often with very broad 

 hair -bands. Tail decidedly larger than in hutleri, and more 

 heavil}^ pencilled, its terminal two-fifths with blackish hairs 

 upv/ards of 13 mm. in length, its upper surface brown, its lower 

 buffy whitisli. The type has a Avhite tail-tip, but this is 

 ■'evidently abnormal. 



Skull -without special peculiarities, a.bout 35-37 mm. in length. 



Dimensions of type (measured in the flesh) : — 



Head and body 126 mm. ; tail 182 ; hind foot 33 ; ear 20. 



Skull : greatest length 37'5 ; condylo-incisive length 34 ; pos- 

 terior palatine foramina 4'2 ; bulla 10 ; upper molar series 5*4. 



Hub. Slopes of Jebel Marra below about 6000', and region of 

 Wadi Aribo. T3''pe from Jebel Marra, south, 5300'. 



Type. Adult female. B.M. No. 23. 1. 1. 127. Original number 

 '698. Collected May 5, 1921. 



* We are quite agreed tliat in systematic work it is, at least for tlie pi-esent and 

 for a long time to come, better to use the conventional notation of m', ni^, m^ for the 

 three cheek-teeth of Muridse, rather than anjr notation which attempts to indicate 

 their real homologies as compared with the cheek-teetli of other rodents. Mr. 

 Kellogg has recently, in a paper dealing with certain Californian Voles (Univ. Calif. 

 Publ. Zool. xxi. p. 245, l'.>22), described and figured the anterior tooth as the last 

 pi-emolarf^'*), the second tooth therefore as ni^, and the third jm^, a method which is 

 liable, in systematic work, to give rise to a good deal of confusion. All the more 

 that his revised nomenclature is by no means acceptable to other writers, as will be 

 seen from a paper by Hinton (Ann. & Mag. N. H. (9) xi. p. 162, 1923), who reviews 

 the whole question, and has reason to confirm the view that the anterior tooth is a 

 milk-premolar {mp^) and not a permanent one, a conclusion which Thomas also 

 thinks may be the true one. 



But in either case the conventional notation m^, m^, m^ would seem to be the best 

 for current sj'stematic work. 



