220 Glacial Erosion in Norway. 
end was still upon it. Yet the sharpness of the edge had scarcely 
been blunted. 
Abundant examples were found to show that the flowing of the 
ice about loose obstacles was quite the rule. Both large and small 
(even an inch in length), angular and rounded masses, lying either 
upon the rock, or upon morainic matter, were sufficient to channel 
the bottom of an advancing glacier. No blocks of rock were seen 
in the act of being turn loose from the floor or sides of the valley, 
and certainly there were no loose or solid masses being picked up 
by the advancing glacier. 
Fie. 2.—Section of Tunsbergdalsbreen, a, bed rock; ¢, cavern under ice b, å 
boulder; ee, moulding in ice of the form of d. 
At Tunsbergdalsbreeen (Fig. 2), whose lower end is 1,600 feet 
above the sea, a modification of the above described phenomena 
was seen. A roughly rounded boulder (Fig. 2 d) of thirty inches 
diameter was enclosed in the convex side of the glacier, which rose 
above it from thirty to forty feet in height. It was resting upon a 
surface, sloping at a high angle, and was held in place by the ice 
itself. As the surface of the stone, bearing upon the rock, was 
small compared with that held in the ice, it should have been 
dragged along. But it was being rolled, as shown by the mould- 
ing (e e) of its form in the glacier which was advancing faster than 
the stone was rolling down the steep slope. The pressure upon this 
stone could not have been merely that of the superincumbent ice, & 
few feet thick, but also that of a powerful component of the weight 
of a glacier from 1,500 to 2,000 feet high descending more or less 
