1uvEns.] fall is formed ; a steep rocky declivity in the channel gives rise to rapids ; a flat plain allows the stream to linger with a scarcely perceptible current; while a lake renders the flow nearly or altogether imperceptible. Thus the rate of flow is regulated in the main by the angle of inclination and form of the channel, but partly also by the volume of water, an increase of volume in a narrow channel increasing the rate of motion even without an increase of slope. The course of a great river may be divided into three parts :— 1. The .l[ozmt¢'u'n T-racl',—where, amidst the clouds and s11ows it takes its rise as a mere brook, and, fed by innumerable similar torrents, dashes rapidly down the steep sides of the mountains, leaping from crag to crag in endless cascades, growing every moment i11 volume, until it enters lower ground. 2. Tim Valle}; Trac/r.—It now flows through the lower hills or undulations which traverse or flank a great mountain chain, and is found at one time in a wide fertile valley, then in a dark gorge, 11ow falling headlong in a cataract, now expanding into a broad lake. This is the part of its career where it assumes the most varied aspects, receives the largest tributaries, and fulfils most cl_1_aracteristically the various conditions which are present to our minds in the idea of a river. 3. The Plain "mc/.'.—Having quitted the undulating region, it finally emerges upon broad plaius, probably wholly, or in great part, made by itself. Here it winds sluggishly in wide curves, perhaps divides so as to enclose tracts of flat meadow or marsh, and finally, amid banks of mud and sand, passes out into the great ocean. In Europe the Rhine, Rhone, and Danube, in Asia the Gauges and Indus, in America the Mississippi and Amazon, in Africa the Nile, mo1'e or less fully illustrate this typical course of a great river. If we draw a longitudinal section of the course of any such river from its source, or from the highest peaks around that source to its mouth at the sea, we find that the line forms a concave curve. Steep at first, where it slopes from the mountain crests down into the valleys, the curve grows less and less through the middle portion, until it finally can hardly be distinguished from a horizontal line. Though characteristic of great rivers, this feature is not confined to their courses, but belongs to the architecture of the conti- nents. It is evident that a river must flow, on the whole, fastest in the first portion of its course, and slowest in the last. The common method of comparing the fall or slope of rivers is to divide the difference of height between their source and the sea—level by their length, so as to give the declivity per mile. This mode, however, often fails to bring out the real resemblances and differences of rivers, even in regard to their angle of slope. For example, two streams rising at a height of 1000 feet, an:l flowing 100 miles, would each have an average slope of 10 feet per mile ; yet they might be wholly unlike each other, one making its descent almost entirely in the first or mountain part of its course, and lazily winding for most of its way through a vast low plain, the othertoiling through the mountains, then keeping among hills and table-lands, so as to form on the whole a tolerably equable and rapid flow. The great rivers of the globe have probably a less average slope than 2 feet per mile. The Missouri has a descent of 28 inches per mile. The average slope of the channel of the Thames is 21 inches per mile ; of the Shannon about ll inches per mile, but between Kill-aloe and Limerick about 6 feet per mile ; of the Nile, below Cairo, 325 to 5'5 inches per mile ; of the Doubs and Rhone, from Besaneon to the Mediterranean, 24°18 inches per mile ; of the Volga, from its source to the sea, a little more than 3 inches per mile. Higher angles of descent are those of torrents, as the Arve, with a slope of 1 in 616 at‘Chamounix, and the Durance, whose angle varies from 1 in 467 to 1 in 208. The slope of a navigable river GEOLOGY 273 ought not, if possible, to exceed 10 inches per mile, or 1 in 6336.1 But not only does the rate of flow of a river vary at dif- ferent parts of its course, it is not the same in every part of the cross section of the river taken at any given point. The sides and bottom, being retarded by friction against the channel, move less rapidly than the centre. The central piers of a bridge have thus a greater velocity of river current to bear than those at the banks. It follows that whatever tends to dimin_ish the friction of the moving current will increase its rate of flow. The same body of water, other conditions being equal, will move faster through a narrow gorge with steep smooth walls than over a broad rough rocky bed. For the same reason, when two streams join, their united current, having in many cases a channel not much larger than that of one of the single streams, flows faster, because the water encounters now the friction of only one channel. The average rate of flow of rivers is much less than might be supposed, even in what are termed swift rivers. A moderate rate is about 1-,} mile in the hour; even that of a torrent does not exceed 18 or 20 miles in the hour? Mr D. Stevenson states that the velocity of such rivers as the Thames, the Tay, or the Clyde may be found to vary from about one mile per hour as a minimum to about three miles per hour as a maximum velocity.3 It may be remarked, in concluding this part of the sub- ject, that elevations and depressions of the land must have a powerful influence upon the slope of rivers. The uprais- ing of the axis of a country must increase the slope, and consequently the rate of flow which, on the contrary, will be diminished by a depression of the axis or by an elevation of the maritime regions. IV. GEOLOGICAL Ac'rIoN.—Like all the other forms of moving water, the streams which traverse a country have both a clae-2m'cal and meclzambal action. The latter receives most attention, as it undoubtedly is the 1nore important; but the former ought not to be omitted in any survey of the general waste of the earth’s surface. i. C’/Lemz'cal.—The water of rivers 111ust possess the powers of a chemical solvent like rain and springs, though; its actual work in this respect can be less easily measured,. seeing that river water is directly derived from rain and springs, and necessarily contains in solution mineral substances supplied to it by them and not by its own oper- ation. Nevertheless, it is sometimes easy to prove that streams dissolve chemically the rocks of their channels. Thus in limestone districts the base of the cliffs of river ravines may be found eaten away into tunnels, arches, and overhanging projections, presenting in their smooth surfaces a great contrast to the angular jointed faces of the same rock where exposed to the influence only of the weather on the higher parts of the cliff. The composition of the river waters of western Europe is well shown by numerous analyses. The substances held in solution include variable proportions of the carbonates of lime, magnesia, and soda ; silica ; peroxides of iron and manganese ; alumina; sulphates of lime, magnesia, potash, and soda; chlorides of sodium, potassium, calcium, and mag- nesium; silicate of potash; nitrates; and organic matter. The minimum proportion of inineial matter among the analyses collected by Bischof was 2'61 in 100,000 parts of water in a mountain stream 3800 feet above the sea. On the other hand as much as 545 parts in the 100,000 were obtained in the waters of the Beuvronne, a tributary of the Loire above Tours. The average of the whole of these analyses is about 21 parts of mineral matter in 100,000 of water, whereof carbonate of lime usually forms the half, its mean quantity D. Stevenson, Canal and Rircr Engineeriizg p. 224. Contjean, Géologie, p. 225. I.’ccIamat[nn of Lrzml, p. 13. FJIDD-‘
X- -- 35