240 EODENTIA. 



was collected is not given, differs only from the last in the greater amount of pale 

 yellow on the tail, which is so profuse on its latter two-thirds as almost to obscure 

 the banding, but like all the other specimens it has a black tip. In another 

 specimen from Formosa, obtained in March, the upper fur is slightly paler than in 

 the dark-bellied specimen, and the under surface is almost concolorous with the 

 back, but sHghtly washed with chestnut, especially over the hinder half of the 

 belly, while it is entirely absent on the chest and on the anterior two-tliirds of 

 the mesial, ventral line, which is grizzled like the sides. Indeed, the whole of the 

 belly is covered with annulated hairs like the upper parts, with chestnut hairs in- 

 termixed. The tail is the same as in the other examples, but with little yellow, 

 the hairs on the posterior third being rather broadly black tipped. This form, then, is 

 in the grey-chested phase. Another and still more interesting example from the 

 same island, but without recorded date, has a broad, grizzled line dow^n the ventral 

 surface, the chestnut, which is dark and rich, being restricted to two narrow Lines 

 which do not extend anterior to the axiUa, so that all the chest and throat are grizzled, 

 Uke the upper surface. The tail is broadly black-tipped hke the foregoing specimen. 

 The last of the Pormosan examples appears to be identical with the last-described 

 squirrel, as they are exactly like each other, with this single remarkable exception, 

 that in the former there is not the slightest trace of rufous on the under parts, 

 which are grizzled exactly like the back. It is a male, but the date of its capture 

 is not recorded. These two specimens resemble in their coats the squirrel obtained 

 in March, with the rufous only faintly showing. The existence of red on the 

 beUy is not a matter of age, as the young and adults are both rufous on that 

 area. 



Two squirrels from the Island of Hainan, clearly referable to this species, 

 differ in this respect, that the specimen shot in February has, as in the female 

 type, the whole of the under parts rich chestnut, extending over the chest and 

 throat ; whereas, in the other, the throat and chest and a line down the middle of the 

 belly are grizzled without rufous, the sides of the belly only being rich chestnut. 

 The tails of these specimens are black-tipped, as in the types, but broadly washed 

 over their ends with yellowish, as in one of the Pormosan squirrels. 



Blyth's S. griseopectus was described from a living specimen, the habitat of which 

 was unknown, but it agrees with those squirrels which I have just described 

 from China, Formosa, and Hainan. 



This species appears to be confined to Western China and to the Islands 

 of Formosa and Hainan. 



* SciURTJS GORDONi, Andr. Plate XIX. 



Sciurus gordoni, Andr., Proc. Zool. Soc. Lond. 1871, p. 140; Blyth, Journ. As. Soe. Beng. 

 vol. xliv. 1875, ex, no., p. 37. 



I have collected about twenty-five specimens of this squirrel from various parts 

 of Burma, north of the capital, and have found only one type of variation, viz., the 



