The yatlierlng of tlie Glaciers. 29 



would not suspect that the visible portion is 3 2 miles long. West 

 of and separated from Girdled glacier only by a narrow ridge is 

 Granite canyon, a deep gorge with precipitous sides, running 

 about eight miles into the heart of the mountains.* The ice 

 slopes downward into the canyon, whose drainage, however, must 

 be back under the ice ; for although I was unable to see ever}^ 

 point of the ridge which closes in the further side of this valley, 

 I could see sufficient of it from diiferent jDoints of observation to 

 convince me that no part of it is less than a thousand feet above 

 the floor of the vallej'. This curious condition seems to be due 

 to the fact that the valley once contained a tributary glacier, 

 which on account of the present smaller supply of ice and the 

 reflection of the heat from the northern side of the canyon has 

 melted down more rapidly than the surface of the main glacier, 

 so that now (although this I could not see) the glaciers draining 

 into this valley are probably entirely separated from the ice 

 enterii^g at its mouth. The tributaries so far mentioned supply 

 none of the ice which forms the ice-front in Muir inlet ; all the 

 ice coming from them that does reach the end of the glacier is 

 compressed into about 800 yards between the ice-front and the 

 mountain on the east. If a line were drawn from the nunatak 

 H to the eastern side of the first northern tributary and a second 

 line toward the northwest at right angles to the first, the sources 

 of all the ice w^hich reaches the ice-front would lie in the quadrant 

 between them. 



The first and second northern tributaries and the main glacier 

 present no striking peculiarities (see plate 7). These are immense 

 streams of ice, fed by innumerable small glaciers. The moun- 

 tains Avhich rise between them and through them are deeply 

 laden with snow, and toward the northwest seem to raise only 

 their summits through the icy sea. The extremities of these 

 branches could not be clearly determined, although they all seem 

 to connect by low divides with valleys beyond. The northwest- 

 ern tributary heads in two beautiful white conical mountains, 

 which we called the Snow cones. A part of its ice flows over 

 the divide between /. and Z^, and joins a large glacier which is 

 probably identical with the one entering the head of Glacier bay. 

 The western tributary supplies no ice to the ice-front ; moreover, 



*This was named from the crystalline nature of the rock, which, how- 

 ever, according to Professor Williams' report (supplement ii), is not a true 

 granite. 



