the Genus Otomys. 271 



weak tail, the presence of faint grooves on the lower 

 ineisors, &c. 



nil is composed of four complete laminae, while in^ is quite 

 the same as that tooth in nnisuicutus. 



Section II. 



The forms of this Section arc easily distinguished from 

 those of the last by the grooving of the lower incisors, and 

 the two groups of the section from one another by the 

 character of this grooving. Group 1, with forms which 

 have only one deep groove in the lower incisors, comprises 

 the larger number of the individuals in the genus. As usual 

 in the classification of the mammal fauna of S. Africa, the 

 species {irroratiis) representing Group 1 can be broadly 

 divided into two forms iuiiabiting opposite sides of the 

 Zambesi River. Individuals from localities south of the 

 river are found to have an m^ made up of six laminre, but 

 in about 5 per cent, of the specimens in the Natural History 

 Museum are found seven laminae ; in specimens from north 

 of the Zambesi exactly the converse is found to be the case, 

 except only in the case of the very high-level form from 

 IMount Kenya, which at the northern limit of the species 

 has uniformly six laminae, and the Ruwenzori form Denli, 

 which has only five. Some such geographical division of the 

 species is also indicated by the colouring, the southern forms 

 being black speckled with white, while the northern sub- 

 stitute a brown for the white of the southern forms. It 

 is in dealing with the S. -Zambesi form of irroratus that I 

 have found that, notwithstanding the very considerable 

 quantity of material available for examination, it is all too 

 small for any really satisfactory result to be arrived at. So 

 far as skull-characters go, the whole species seems to be in an 

 unstable condition. I have failed to find in the southern 

 specimens a single series in which any one distinctive 

 character is really constant. I have already said that the 

 laminae composition of m^ presents exceptions to an other- 

 wise general rule. In size there is similar variation : 

 specimens (quite mature) from Cape Town, De Kaap, Trans- 

 vaal, &c., show a greatest skull-length of 36 mm. ; others, 

 from King ^Villiam^s Town, Kurumau, &c., show 44 and 

 even 46 mm. ; while the normal size is 40—41 mm. Similar 

 in-and-out variation could be shown for almost any character. 

 Under the circumstances I have decided to leave all these 

 forms under irroratus, only distinguishing, south of the 

 Zambesi, a couple of colour-forms as subspecies. 



