346 ON MOUNTAINS AND MANKIND. 



will find the scale increased ; Siniolchiim, Jannu, and Kangchenjnnga 

 are all j)ortenti()iis •■giants. To put it at a low average figure, the cliffs 

 of their final peaks are half as high again as those of ]\Ionte Kosa and 

 the Matterhorn. 



In all these chains you will find the same feature of watershed or 

 partings l3'ing not in but behind the geological axis, which is often 

 the line of gi'eatest peak elevation. This is the case in the Alps at 

 the St. Gothard. in the Caucasus for some 40 miles west of the Dariel 

 Pass, in the Ilinuilaya, in Sikhim and Nepal, where the waters (low- 

 ing from the Tibetan Plateau slowly eat their \vay back behind 

 Kangchenjunga and the Xepalese snows. The passes at their sources 

 are found consequently to be of the mildest character — hills " like 

 Wiltshire Downs," is the description given by a military explorer. 

 It needs no great stretch of geological imagination to believe in the 

 cutting back of the southern streams of Sikhim or the Alps, as for 

 instance at the Maloya; but I confess that I can not see how the 

 gorges of Ossetia, clefts cut through the central axis of the Caucasus, 

 can be ascribed mainly to the action of water. 



I turn to the snow and ice region. Far more snow is deposited on 

 the heights of the central Caucasus and the eastern Himalaya than 

 on the Alps. It remains plastered on their precipices, forming 

 hanging glaciers everywhere of the kind found on the northern, the 

 Wengern Alp, face of the Jungfrau. Such a peak as the Weisshorn 

 looks poor and bare compared with Tetnuld in the Caucasus or Siniol- 

 chum in the Himalaya. The plastered sheets of snow between their , 

 great bosses of ice are perpetually melting, their surfaces are grooved, 

 so as to suggest fluted armor, by tiny avalanches and runnels. 



In the Aletsch glacier the Alps have a champion with which the 

 Caucasus can not compete; but apart from this single exception the 

 Caucasian glaciers are superior to the Alpine in extent and pictur- 

 esqueness. Their surfaces present the features familial- to us in the 

 Alps — ice falls, moulins, and earth cones. 



In Sikhim, on the contrary, the glaciers exhibit many novel fea- 

 tures, due, no doubt, mainly to the great sun heat. In the lower i)()r- 

 tion their surface is apt to be covered with the debris that has fallen 

 from the impending clifts, so that little or no ice is visible from any 

 distance. In the region below the neve there are very few crevasses, 

 the ice heaves itself along in huge and rude undulati(ms, high, gritty 

 mounds, separated by hollows often occupied by yellow pools which 

 are connected by streams running in little icy ravines, a region excep- 

 tionally tiresome, but in no way dangerous to the explorer. In steep 

 places the alpine icefall is replaced by a feature I may best compare 

 with a series of earth pillars, such as are found near Evolena, in 

 Canton Valais, and elsewhere, and are figured in most text-books. 

 The ice is shaped into a multitude of thin ridges and spires, resem- 



I 



