The destructive work of rivers, including in that term all surface streams, is far less extensive, in the aggregate, than that of the atmospheric agencies, but because the work of a stream is concentrated along its narrow course, it appears much more striking and impressive.
Fig. 54. - The "Bottomless Pit," Arizona. The stream disappears in a limestone cavern and is not known to reappear. (Photograph by A. E. Hackett, Flagstaff, Ariz).
A certain amount of solution and decomposition is performed by rivers upon the rocks of the bed, and in limestones this may be considerable, especially if the water be charged with organic acids from a swamp or peat-bog. Limestone regions are characterized by a paucity of surface streams, most of which pass into caverns and underground channels which they have made by dissolving the limestone. Such subterranean streams may or may not reappear on the surface, according to circumstances.
The mechanical work of a river is much greater than the chemical, and is dependent upon the velocity of the current, varying directly as the square of that velocity. The velocity of a stream is the rather complex resultant of several factors, the chief of which is gravity; the steeper the slope of the bed, the swifter the flow of the water. A second factor is the volume of water, the velocity varying as the cube root of the volume. That is to say, if one of two streams which flow down the same slope has eight times as much water as the other, it will flow twice as fast. Other factors enter into the result, but slope of bed and volume of water are much the most important.
Pure water can do little to abrade hard rocks, though it can wash away sand, gravel, and other loose materials. When the Colorado River broke into the Salton Sink in southeastern California in 1905, it cut a deep trench with incredible rapidity through the soft alluvial soils. Streams also take advantage of the joint-blocks, into which all rocks are divided, and often loosen and carry down such blocks. This process is called plucking and is important in the destructive work of glaciers and the sea. As in the case of the wind, the stream merely supplies the power; the implement with which the cutting is performed is the sand, pebbles, and other hard particles which the water sets in motion. These abrade the rocks against which they are cast, just as the wind-driven sand does, but more effectively, because of the ceaseless activity of the stream, and because many rocks are rendered softer and more yielding by being wet. The cutting materials are themselves abraded and worn finer and finer by continued friction against the rocks and against one another. In the case of complex minerals this abrasion is accompanied by more or less chemical decomposition, as has been shown experimentally by rotating crystals of felspar in a drum half filled with water.
When the felspar was ground down to mud, the water showed the presence of potash and soda in solution. Angular blocks are speedily worn into cobblestones and these into pebbles of spheroidal or flat, discoidal form. A process of selection goes on, by which the softer materials are ground into mud, the harder remaining as pebbles and sand.
An example of exceedingly rapid wear of hard rock by running water, under favourable conditions, is given by the Sill tunnel in Austria, which is provided with a pavement of granite slabs more than a yard thick. Great quantities of debris are swept over this pavement at a high velocity and so rapid is the abrasion, that it was found necessary to renew the granite slabs after a single year.
A river which is subject to sudden fluctuations of volume, being now a rushing torrent and again almost dry, is a much more efficient agent, both of erosion and of transportation, than is one which carries nearly the same quantity of water at all times, or which fluctuates only slowly.
The'velocity of a stream differs much in its various parts, diminishing, as a rule, from the head waters to the mouth. In very many cases there are also local variations of speed, falls, rapids, and eddies alternating with quiet reaches. In eddies and at the foot of cascades the water acquires a rotary motion, which is transmitted to stones lying on the bottom. In a rocky bed these revolving stones excavate cylindrical holes, often of remarkable, regularity, called pot-holes, or giant kettles. The diameter and depth of pot-holes are determined by the volume and velocity of the water and by the length of time during which the eddy or fall remains at the same point.
Since the velocity of a stream is so largely dependent upon gravity, it is obvious that the deeper a stream cuts its channel, the less steep does its slope become, and that so long as the region is neither upheaved nor depressed, the river performs its vertical erosion at a constantly decreasing rate. Unless, therefore, the work is done under very exceptional conditions, as in the case of the Niagara, we cannot reason from the present rate of excavation to the length of time involved in cutting out a given gorge.