1892. J MR. O. THOMAS ON MAMMALS FROM NYASSALAND. 549 



Firstly, they prove conclusively that the latter species has a 

 seasonal change of colour, the two coats being the one a grizzled 

 rufous and the other a grizzled grey. Peters's type, as figured by 

 him, was a changing specimen with the anterior half of the body 

 grey and the posterior rufous. Specimen a is in very much the 

 same state, and in my opinion unquestionably belongs to the same 

 species. From the relative lengths of the two sorts of fur, it is 

 evidently changing from the rufous to the grey phase, and the 

 latter is therefore obviously the summer form. In specimen b, 

 killed a month later, the change has gone a little further, the gre}'' 

 fur having become as long as tlie rufous, while the latter has much 

 decreased in area ; the hairs on the centre of the back are deep 

 black to their roots. 



Of the three co-types of S. shirensis. Gray, one, young, is in the 

 rufous state, except that its extremities are grey, but the two adults 

 are both wholly in their grey coats. They are, however, quite 

 uniform in colour, and have no black patches on their backs. All 

 the specimens examined have one premolar only in the upper jaw, 

 although Peters's type was said to have two, and on this account 

 S. mutabilis and 8. shirensis were kept apart in Dr. Jentink's 

 monographs of the African Squirrels ^. 



My own conclusion is that the new Nyassa specimens, with one 

 premolar, are certainly S. mutabilis, a determination which destroys 

 the importance of the presence or absence of the extra premolar in 

 this form, and then, this character being gone, that there is nothing 

 to separate the two forms but the black dorsal patches of tmitabilis, 

 and that as to these we must be content to wait until further 

 specimens prove them to be due either to individual variation, to 

 advanced age, or to genuine specific distinction. But the fact that 

 the types of S. shirensis came, as their name implies, actually from 

 the River Shire is strongly in favour of their specific identity with 

 Mr. Johnston's Nyassa examples. 



7. SciURus PALLiATus, Peters. 



a,b. J $ . Milanji Plateau, 6000 ft. 29/10/91. 

 c. 6. Milanji Plateau, 6000 ft. 2/11/91. 



8. Otomys IRRORATUS, Brauts. 

 a. Ad. al. ? . 



Like specimens of this species collected by Mr. Jackson in 

 Mianzini ^, Masailand, this individual has a molar lamina-formula 



3-2-7 

 0^4=2=2- 



9. Gerbillus (Tatera) afer. Gray (?). 

 a-d. 4 in al. 



The South African species of the subgenus Tatera are so little 



^ N. L. M. iv. p. 18 (1882). Dr. Jentink, however, implies that some doubt 

 exists as to the stall in Peters's type-skin really belonging to it. 

 2 See P. Z. S. IS'Jl, p. 184. 



