THE LARGE INDIAN SQUIRREL. 305 



P. Eedii, p. 34). The only specimen of S. purpureus enumerated in the 

 Catalogue is thus entered : " A specimen of the male (from R. W.), 

 shoulders, loins, and tail black." On examination of the British 

 Museum collection I find that this specimen is the skin from Amarkantak 

 already noticed. It is identified by the Museum label, and it still bears 

 a paper ticket with the locality carefully recorded, the record being 

 initialled " R. W." This, I think, satisfactorily disposes of the supposed 

 Tevai locality. 



The specimen labelled " Manipur " is not so easily discredited. 

 Hume's collections were carefully labelled and the greatest pains taken 

 to preserve all records of localities. But so great is the improbability 

 of this species inhabiting Manipur, that I cannot but think it possible 

 that in this case also a mistake has occurred. A few of the labels, as 

 Mr. Thomas reminds me, must have become detached from the skinsof 

 the Hume collection, and although it is believed that in almost every 

 case the specimen and its label were correctly brought together again> 

 there was one case of a label having been attached to the wrong speci- 

 men, also a squirrel, that was noticed by me in 1878 (Journ. As. Soc. 

 Beng., XVII, part II, p. 162). I suspect a similar accident has occur- 

 red to the supposed Manipur skin of S. indictts. It should be recollected 

 that in numerous cases the labels of the Mammalian specimens had only 

 the locality and date marked on them, the name of the animal having 

 been left for future identification ; consequently if two or three labels 

 were found loose at one time there was always a possibility of tbeir being 

 reattached to the wrong skins. 



The accompanying figures are partly from specimens of the races 

 dealbatus and malabaricus now living in Mrs. G. W. Vidal's possession? 

 partly from skins in the British Museum. To facilitate reference, the 

 label of each skin from wbich the coloration was copied has been 

 marked, so that these specimens may, if necessary, be regarded as types. 



