474 TLEE OGL INNATE OF GLE OE OGN A 
of the glacier, heaped them up in cones at its base. Not only 
was this indicated by the structure of the cones themselves, but 
the process of their formation was observed in actual: progress." 
Its rapidity was something surprising. As the granules were 
brought down the ice-wall by the cascading rivulet they occa- 
sionally struck projections and bounded forth into the air and 
fell upon the slopes of the cone. In other cases they were seen 
to leap forth from the rivulet when it struck the summit of the 
cone; but in the main they simply gathered at the apex of the 
cone until they blocked up the little channel-way and forced the 
rivulet to seek another course only to be blocked in turn and 
thus forced to and fro, building up the conesymmetrically. The 
rapidity of action observed was doubtless somewhat exceptional 
for the day was unusually bright and warm. It is doubtful 
whether it was equaled by more than ten days during the entire 
season. 
The embankment between the cones had a well developed 
granular structure, but it appeared to be formed in considerable 
part from wind-drift accumulations. It was very notably solidi- 
fied into a half-ice. The surface was covered with dirt and to 
some extent with rocky and earthy débris that had fallen from 
the face of the glacier. From its solidity, dirtiness and other 
features it seemed scarcely possible that it was the product of the 
current season solely. As may be inferred from the illustrations, 
it lay close against the front of the ice-wall rising in some places 
half its height. Nowhere did I observe any disruption or crump- 
ling of the embankment as though the glacier had crowded against 
it since it was formed, nor, on the other hand, it is scarcely need- 
ful to say, was there any crevasse or interval between it and the 
glacier, indicating a retreat of the ice. Such retreat would only 
be produced by the melting of the glacier backward faster than 
it moved on, and such melting could scarcely be supposed to take 
place without involving the destruction of the embankment as a 
concurrent, if nota prerequisite, condition. Manifestly melting 
could not have counteracted the onward movement at the base of 
* PROFESSOR LIBBEY was the first to note the action in progress. 
