The 3fargiii of the Cornell Glacier. 149 
the sea, instead of pushing them to a new margin. The rea- 
son for the ridge form is that debris comes to the ice ledge, 
and. loosened by melting, rolls to the bottom, accumulating 
there. In most cases the moraine was evidently much less 
thick than it appeared to be because of the ice foot below ; Init 
some ridges twelve to fifteen feet in hight, were evidently en- 
tirely composed of rock fragments because separated from 
the ice by ah interval in some cases as great as 150 feet. The 
ridge form is necessary because the ice along the land advan- 
ces and retreats only very slightly. The only way in which 
an extensive deposit of hummocky moraine could be formed 
would be for the ice to stay nearly in one place for a long- 
time. If the glacier retreats, the debris from the ice is dis- 
tributed as a thin covering; if it advance the moraine is car- 
ried away toward the sea. 
The material brought by the glacier to this terminus was 
very stony, and hence the surface features are those of very 
rocky moraines. They are often piles of boulders with very 
little clay,, though at times the clay element formed one-half 
the bulk of the moraine. The rocks were frequently rounded 
and scratched like ordinary ice transported fragments. 
Among these boulders gneiss predominated, but porphyritic 
granite, slate, and quartzite, derived from some hidden areas 
beyond the ice margin, are abundant, and the size of the 
bl(jcks varies up to ten or fifteen tons. 
In some places along the land iiiargin the moraine on the 
border of the ice is increased in bulk by the addition of de- 
bris falling from the land. But this is rare, and is found on- 
ly where the ice movement is such as to press it against pre- 
cipitous hills. Around the nunataks this condition is not 
uncommon, for there the rapid glacier movement presses the 
ice against the land. At mount Schurnian there are two such 
bands, one on either side, caused partly by glacial debris and 
partly by masses falling from the clilfs; and tiie union of 
these two bands forms the medial moraine described above. 
Professor Barton showed me exactly such a moraine around 
the base of the Umenak mountain near the town of Umeiiak. 
It had been gathered there at a time when the glacier jiad 
passed out into the Umenak fjord, and the mountain had ris- 
en as a nunutak above the valley, in exactly the same way 
that mount Schurman now rises above the Cornell glacier. 
