41« JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXIV. 



h^. General colour darker and 

 more olivaceous. Eastern 

 Sikkim... ... ... B.l. hhotia. 



If. Belly washed with bnfFy or yellowish. 

 A darker line on back, though 

 very faint in faded pelage. Mani- 

 pur and Chin Hills ... ... B. macmillani. 



h. Belly hairs whitish or creamy terminally. 

 Ear patches ordinarily ochraceous. 

 Chin Hills (Eastwards into China) ... B. jpernyi. 

 B. Cheeks, and middle area of tail below, 

 bright ferruginous. 

 a. Size larger. Colour darker ... ... B. rafigenis. 



a^. Upper Chindwin ... ... ... B. r. opimus. 



}f. Tenasserim ... ... ... B. r. rufigenis. 



h. Size smaller. Colour paler. Chindwin 



and Shan States ... ... ... B. r. adamsoni. 



In B. macmillani there appears to be a great difference in the 

 prominence of the dark dorsal line between specimens before and 

 after the spring moult, so that at one time I supposed the pre-moult 

 specimens to be examples of a local form of B. lokriah with yellow 

 bellies and a faint darker dorsal stripe. But closer examination 

 convinces me that the two apparent forms are really seasonal 

 changes of one and the same. The strongly lined post-moult form, 

 niai'Anillani as described, has its dorsal stripe sharply marked, and 

 has the hairs of its ear patches 'more or less ochraceous terminally. 

 Later on in the season the hairs of the stripe lose their distinct- 

 ness by fading, while the ends of the hairs in the postauricular 

 spots apparently fade or wear off so as to result in a more or less 

 whitish ear patch, similar to that of B. lokriah. 



In the skull there is absolutely no difference between B. lohriah 

 and 7}yxcmillani, so that it may be a question whether the latter 

 should not rather be considered as a local subspecies of the former. 

 But until intermediate specimens turn up it seems better to leave 

 things as they are. 



Bremomys pernyi, characterised by its whitish belly, is a new 

 addition to the fauna of British India, its previously known range 

 extending from Yunnan to Fo-kien and An-hwei, Eastern China. 



With regard to B. rufigenis I now merely repeat the characters 

 already used in describing the three subspecies, as it is evident that 

 further study of the group is necessary. Attention may, however, 

 be drawn to the really remarkable resemblance between the B. r. 

 opiiifhus of Hkamti, in the far north of the Chindwin and the true 

 rufigenis of Tenasserim, while all the forms between them, repre- 

 senting B. r. adamsoni, are so different that they should perhaps be 

 specifically separated. 



