1873.] 379 * — [Niles. 
vate or hollow out the rock into basins, which being filled with water 
on the retreat of the glaciers, would form lakes. Prof. Tyndall had 
objected to this, because at such places glaciers are almost stationary. 
I have seen no glacier which appeared to me to be excavating 
the rock at its termination in the manner required to form lake- 
basins. On the contrary the lower ends of most glaciers do not even 
extend into basin-like depressions. Many glaciers terminate on the 
steep slopes of mountains, and some at the edges of precipices. 
Certainly a glacier cannot be deepening a basin so long as it reaches 
only to its outer edge. In many instances the rock under the ice at 
the lower ends is convex rather than concave, and sometimes a mass 
of rock divides the glacier into two parts, as is the case with the Bies 
Glacier. 
In studying the manner in which glaciers now grind the rock be- 
neath them, we see what Prof. Agassiz has taught, that the greatest 
erosive force is exerted upon the convex or prominent portions of 
the surface. The Gorner Glacier furnishes superior opportunities for 
getting at its under surface in places, and of observing the contact 
of the moving ice and the rock. At such places the ice rests princi- 
pally on the eminences, and bridges the hollows between them, as so 
well described and illustrated by Mr. Edward Whymper. Under 
such conditions the prominent portions must receive the full erosive 
power, and it becomes evident that the valley cannot be much deep- 
ened by present processes till these elevated places are leveled. But 
suppose the action to go on till all such prominences are removed, 
the glacier would then move over such a smooth and even surface 
with less resistance, and the grinding process would go on stiJl more 
slowly than on the eminences. But the elevations still remain under- 
neath the larger glaciers, which have flowed over them for ages, and 
their existence shows that these valleys can have been deepened only 
to a limited extent by the erosion of these glaciers. 
Furthermore, such reches moutonnées are well known to be among 
the most common yroofs of ancient g!aciers, and they are found in 
larger as well as in smaller valleys. Sometimes still larger elevations 
standing in the valleys show how incompetent the glaciers have been 
to remove exposed masses of rock. A striking example of this is 
seen in the masses upon which the castles of Sion now stand, in the 
valley of the Rhone. As has been observed and stated by many 
others, it is at least difficult to understand how glaciers could have 
formed the Alpine valleys by excavation, when we consider that 
