1893.] ME. O. THOMAS ON MAMMALS FEOil NTASAI-AaT). 503 



16. Mtjs mxtsculus, L. 



a, b. Ad. sks. Zomba. 10/92. 



c. Ad. al. Milanji. 10/92. 



17. Mrs (Leggada) mixutoides, A. Sm. 

 a-c. Ad. sks. Zomba. 10/92. 



18. Ceicetomts GAMBLtNUS, Waterh. 

 a. Ad. sk. Zomba. 10/92. 



19. Saccostomus campesteis ', Pet. 



a. Ad. sk. Upper Shire. 11/92. 



b-f. Ad. & imm. al. Fort Johnston. 11/92. 



20. StEATOMTS PEATEXSIS ', Pet. 



a-e. Ad. sks. Zomba. 10/92. 



d. Ad. al. d . ^lilauji. 10/92. 



These rare and peculiar short-tailed African Muridae, with their 

 doubtful affinities, are of the greatest interest, and it is probable that 

 a clue to some of the most difficult problems of Murine phylogeny 

 will be gained by a study of their teeth. For these reasons 

 specimens of Saccostomus and Steatomys are most acceptable, and 

 it is to be hoped that Mr. AV^hyte will also come across examples 

 of the still rarer Malacothri.v undi Mystromys, approximately similar 

 in size and proportions, but with very different teeth. 



Although similar to each other superficially, these four last- 

 named genera appear to belong to three different subfamilies — 

 Mysti'omys seeming to be strictly Cricetine ; Steatomys, as Alston 

 has stated ^, being closely allied to Bendromys, as also are Mcdacothrix 

 and Bendromys ; while Lophuromys, Saccostomus, and Cricetomys 

 ought, I believe, all to be placed in the restricted subfamily Murine. 

 Of these last three the first was placed by Alston in the Dendro- 

 myinse and the other two in the Cricetinae. 



21. DEXDEOilXS ilESOilELAS, Bts. 



a. Ad. sk. Zomba. 10/92. 



b~f. 5 in al. Milanji. 



These specimens, although 1 place them provisionally under the 

 oldest name, D. mesomelas, certainly seem to differ constantly from 

 the ordinary form of that species in being entirely without the 

 black dorsal line and in being considerably smaller. In these 

 respects they agree with two out of the previous collection, 

 while two others are of the usual large-striped character. They thus 

 correspond with Wagner's B. pumUio, described in 1841% but after- 

 wards * considered by him as a synonym of B. mesomelas. Larger 

 series from different localities are, however, needed before one can 



* For nomenclature see Ann. Mag. N. H. (6) x. p. 264 (1892). 



^ P. Z. S. 1876, p. 82. 



^ Miincb. Gelehrt. Anzeiger, xii. p. 437. 



^ Schr. Saug. Supp. iii. p. 464 (1843). 



04: 



