HEUDe's CHINESE MAMMALS. 21 



A look at the map of China will show that the places where 

 the above specimens were collected can be arianged in five areas, 

 more or less distant and separated from one another, i^anging 

 from the coastal regions to the border of Thibet, 



These are : — 



(«) Chekiang Province. 



(b) Kuang-tung (Canton) Province. 



(c) Tonkin. 



(d) North-eastern Ssuclman and South 8hensi. 



(e) West Ssuchuan and Moupin. 



Whether the last two are really separable is doubtful, for they 

 consist of much the same sort of country, i. e. high, precipitous, 

 and well-forested mountains. 



In examining the skulls, however, I found that they could be 

 divided up according to the above divisions. 



I first examined the lai'ge series of fine skulls from the Chen- 

 kou-ting district of North-eastern Ssuchuan, close to the borders 

 of Shensi and Hupei. They were all collected by M. Farges 

 during the year 1894. I found it impossible to separate them 

 one from the other, the whole series showing certain well-defined 

 characteristics. Most noticeable of these was the uniformly great 

 length of the horns, which sloped sharply back, were well curved, 

 and were closely set, the tips having a tendency to point inward. 

 Though the bases of the horns were annulated and there were 

 the usual narrow longitudinal grooves, yet these were not very 

 pronounced, and the horns had a smooth shiny appearance. 



Turning to the skins corresponding with these skulls, I was 

 at first struck by the great variation in colour, but on a closer 

 examination I found that this was due entirely to the quantit}^ of 

 under-wool present. Thus some pelts were of a grey-bufi' colour, 

 which was due to a great abundance of under-wool, through which 

 long coarse black hairs protruded, while other pelts were entirely 

 black, having no under-wool at all, the hairs being coarse and 

 short. There were all stages between these two extremes. It 

 is obvious from this that the vaiying colours as described by 

 Heude, and upon which he bases to a lai-ge extent his classifi- 

 cation, are due entirely to seasonal changes, mid-winter specimens 

 being light and mid-summer specimens dark. 



An examination of the manes, which were pretty well developed 

 in all the specimens, showed that in nearly all of them white or 

 cream hairs were present, in some greatly predominating over the 

 black and chestnut hairs, in others scai'cely noticeable on account 

 of their scarcity. There seems to be little connection between 

 the lightness or otherwise of the mane with the lightness of the 

 coat generally, and I came to the conclusion, finally, that the 

 white mane is a characteristic of age more than of season. At the 

 same time it may be a characteristic which varies in individuals. 

 However that may be, it seems pretty certain that the presence 

 or absence of the white mane is not a specific characteristic. I 

 also noticed that the rusty brown of the legs seemed a scniewhat 



