
Parr IL. Suct. ii. § 3.] RIVER EROSION. 373 
~The effects of abrasion upon the loose materials on a river-bed are 
but a minor part of the erosive work performed by the stream. A 
layer of débris, only the upper portion of which is pushed onward by 
the current, protects the solid rock of the river channel, but is apt to 
be swept away from time to time by violent floods. Sand, gravel, and 
boulders, in those parts of a river channel where the current is strong 
enough to keep them moving along, rub down the rocky bottom 
over which they are driven. As the shape and declivity of the 
channel vary constantly from point to point, with, at the same time, 
frequent changes in the nature of its rocks, this erosive action is 
liable to continual modifications. Jt advances most briskly in the 
numerous hollows and grooves along which chiefiy these loose 
materials travel. Wherever an eddy occurs in which gravel is kept 
in gyration, erosion is much increased. The stones in their move- 
ment excavate a hole in the channel, while, as they themselves are 
reduced to sand and mud, or are swept out by the force of the 
current, their places are taken by fresh stones brought down by the 
stream (Fig. 107). Such pot-holes, as they are termed, vary in size 
from mere cup-like depressions to huge cauldrons or pools. As they 
often coalesce, by the giving way of the intervening walls between 
two or more of them, they materially increase the deepening of the 
river-bed. 
That a river erodes its channel by means of its transported 
sediment, and not by the mere friction of the water, is sometimes 
admirably illustrated in the course of streams filtered by one or more 
lakes. As the Rhone escapes from the Lake of Geneva, it sweeps 
with a swift clear eurrent over ledges of rock that have not yet been 
very deeply eroded. ‘The Niagara supplies a still more impressive 
example. Issuing from Lake Erie, and flowing through a level 
country for a few miles, it approaches its falls by a series of rapids, 
The water leayes the lake with hardly any appreciable sediment, 
and has too brief a journey in which to gather it before beginning 
to rush down the rocky channel towards the cataract. The sight of 
the vast body of clear water, leaping and shooting over the sheets 
of limestone in the rapids, is in some respects quite as striking a 
scene as the great falls. To a geologist it is specially instructive ; 
for he can observe that, notwithstanding the tremendous rush of 
water which has been rolling over them for so many centuries, these 
rocks have been comparatively little abraded. The smoothed and 
striated surface left by the ice-sheet of the glacial period can be 
traced upon them almost to the water’s-edge, and the flat ledges 
at the rapids are merely a prolongation of the ice-worn surface 
which passes under the banks of drift on either side. The river has 
hardly eroded more than a mere superficial skin of rock here since 
it began to flow over the glaciated limestone. 
Similar evidence is offered by the St. Lawrence. This majestic 
river leaves Lake Ontario as pure as the waters of the lake itself. 
The ice-worn hummocks of gneiss at the Thousand Islands still retain 
