Englac ia l Drift. — Crosby . 
227 
the progressive increase in the velocity of^the ice from the 
base upward, in obedience to the principle, already noted, that 
a current tends to force floating bodies from the lines of high¬ 
est velocity to those of lowest velocity. 
It is possible, however, that, as virtually pointed out by 
Upharn,* an important exception to this limitation of the 
range of englacial drift should be made for the case when ^ 
later sedentary ice-sheet is overriden by the readvance of an 
earlier sheet, the hight attained by the englacial drift depen¬ 
ding then upon the thickness of the overridden sheet. Again, 
a mountainous tract lying in the pathway of the ice-sheet 
may lead to the incorporation of drift at exceptionally high 
levels. In fact, the Malaspina glacier is a capital example. 
According to Russell, the drift which mantles the outer mar¬ 
gin of this great piedmont glacier, covers, even in parts of the 
area where it is forest-clad, not less than a thousand feet in 
thickness of ice; and where it is not forest-clad it rises to 
still greater hights. Russell states that this superglacial drift 
consists wholly of the lateral and medial moraines of the 
tributary glaciers descending from the Mount St. Elias range. 
All of the glaciers which feed the great Piedmont ice-sheet are above 
the snow line, and the debris they carry only appears at the surface 
after the ice descends to the region where the annual waste is in excess 
of the annual supply. The stones and dirt previously contained in the 
glacier are then concentrated at the surface owing to the melting of the 
ice. This is the history of all the moraines on the glacier. They are 
formed of the debris brought out of the mountains by the tributary AL- 
pine glaciers, and concentrated at the surface by reason of the melting 
of the ice. 
Probably this incorporation of drift at high levels would 
still occur, but on a grander scale, if the St. Elias range were 
completely buried in ice moving across it. Regarding these 
St. Elias glaciers as rivers of ice tributary to a sea of ice, the 
thought is naturally suggested that perhaps the drift-laden 
ice tends, on joining the piedmont glacier, to flow out across 
the ice sea regardless of its depth, and thus, while actually 
descending topographically, rise to greater hights in the ice- 
sheet. This principle would also, obviously, And application 
in the case of valleys or depressed areas of any form, trans¬ 
verse to the general direction of flow of the ice-sheet; and it 
*Bull. Geol. Soc. Amer., vii, 21, 22. 
