of which is constantly increasing toward the source of the

stream.

The regularity of this curve is of course affected by many natural conditions, especially by the different resistingpower of the rocks over which the water runs, but if it traversed only one kind of rock, and the tributaries were nearly equidistant, there would be a near approach to a regular curve.

If we regard a river as consisting of as many lengths or sections as there are important tributaries, we may consider the volume of water passing through each section when the stream is full to be a certain average quantity, and the slope to which the current is reducing (or has reduced) its channel in that section to be an approximately straight line with a continuous fall of so many feet or inches per mile. This slope has been called by Professor Powell the base level of erosion,' but as it is not a level, the base line of erosion is a better term, and the curve produced by the combination of the several lengths or sections may be called the curve of erosion (see fig. 196).

We see, therefore, that if a region remain stationary for a sufficient length of time, the streams will reach a limit in the work of erosion, and though extraordinary floods may occasionally enable them to continue the work, the ordinary full stream will have ceased to deepen the channel through which it flows. Its force will be expended partly in the transport of detritus, and partly in the work of lateral

erosion.

Let us now consider the effect which a movement of upheaval will have on such a region. It is not difficult to perceive that its general effect will be to accelerate all processes of erosion, if only because it will, as a rule, increase the rainfall on the watersheds, and by increasing the volume of water in the streams will thereby give them a greater velocity and power of erosion. Apart from this, however, the upheaval may increase not only the absolute height of the land, but the declivity or proportional slope measured in feet per mile between a given watershed and the sea; and wherever the declivity is increased the streams

1 66 'Exploration of the Colorado River, 1875,” p. 203.

will regain their power of vertical erosion. This result, however, must depend on the local slopes of the tract of land which has been added to the country by upheaval.

Let us suppose that the movement has been fairly rapid, and that a tract of varying width has been added to the former coast-line: in some places this tract may be so broad and flat that its general inclination is only equal to, or perhaps less, than the slope of the previously formed valleys; the river channels will then simply be continued across it, and the curve of erosion will only be prolonged without being otherwise altered. In other parts of the region, however, the slope of the newly-added tract may be greater than of the valleys which open on to it, and in this case the streams occupying those valleys will recover their excavating power, and will readjust the slope of their channels to a deeper curve of erosion. This process will

a

h

Fig. 197. Effect of Upheaval on Valley Erosion.

take place by degrees, rapids will be formed at the old mouth of the valley, and if the rocks are of unequal hardness, it is probable that a succession of rapids will eventually be produced, the site of each gradually retiring up the river and up its tributaries till every part of the valley system has been cut down to a lower base-line.

The diagram (fig. 197) will assist the reader to understand the final result of this action, as viewed along the line of the main stream or river. In this the line a b represents the slope of a watercourse or river-channel when the sea-level coincided with the line s', and the coast was at the point b. If now the country was raised till the sealevel came to correspond with the lower line s2, and the slope of the newly formed land was represented by the line bs, the stream would immediately begin to deepen its channel at b and along the slope b 82. This erosion would be carried back along the whole length of the valley until the stream had completely accommodated itself to the

altered conditions, and had deepened its channel by an amount proportional to the increased height between 8' and 8. The amount of rock which might be removed along the bottom of the valley is indicated by the black part of fig. 197.

If instead of being raised the region we have imagined were depressed, so that a portion of it sank below the level of the sea, the ends of the valleys in this portion would be submerged and converted into tidal estuaries, and would be partly filled with estuarine deposits; while even above the tidal influence the velocity of the streams would be lessened and erosion checked. If subsidence also diminished the amount of the rainfall on the watersheds, the velocity of the streams would be still further decreased, and they would begin to deposit their detritus sooner than they had previously done; flood plains and alluvial levels would consequently be formed throughout the lower parts of the valleys. It may in fact be stated as a general rule that wherever we find an alluvial level by the side of a stream, we may be sure that the stream is not now deepening that part of its valley, but has reached the local base-line of erosion.

As nearly all rivers flow over rocks of different hardness and cut their way back by a series of rapids and waterfalls, it has been suggested that the position of falls and rapids is a guide to the age of the river,' its age being great in proportion as these are near to the sources of the river. With certain limitations and exceptions this is doubtless cerrect; but, as we have seen, an upheaval of the sea-board may cause a new set of rapids to be developed: while, therefore, the presence of rapids in the lower part of a rivercourse does not prove the valley to be of late origin, their limitation to the higher parts of the tributary streams is certainly an indication of the ancient establishment of the drainage system.

It should also be noticed in passing that the lowering of the water-level of a large lake will have the same effect on the streams running into it as the lifting of a whole country has on the streams running into the sea.

1 "Quart. Journ. Geol. Soc.," vol. xxxv. p. 116.

From the above considerations it is evident that the influence which movements of subsidence or upheaval exercise upon the power of rivers to erode their channels is most important, and the author believes that it is mainly because the advocates of river-erosion have neglected to insist upon this point that many persons have been unable to see the cogency of their arguments, or to believe that rivers have had any share in the excavation of their valleys. It is said by these sceptics that the rivers are not now deepening the valleys in which they run, and, therefore, that the valleys cannot have been made by the rivers some have answered this by supposing that the volume of water in the rivers was much greater formerly than it is now, but the real answer is that these streams have reached their base-lines of erosion, and consequently that observations on their present action will not enable us to realize. how they have excavated the valleys in which they flow. This seems to be the case with many of the large European rivers, especially the Rhine and Moselle, whose valleys were described on p. 147.

;

The southern part of England is another region where the valleys are not now being deepened. At no remote period of geological time the whole of the British Islands stood at a much higher level than they do now, and during this time erosion was very active, and the whole valley system of the country was gradually elaborated. After this period came a succession of movements the details of which are not yet fully known, but it is fairly certain that the last movement was a tilt from north to south, causing a slight elevation of Scotland and a depression of southern England, which finally became severed from the continent. The result of this differential movement was to check the velocity of the English streams, and to cause the formation of those alluvial levels or water-meadows which are such a characteristic feature in the valleys of southern and central England. The north of England experienced little change of level, but in Scotland the rivers have cut channels through the old alluvia and estuarine deposits, and are probably still in many places deepening the upper parts of their valleys among the hills and mountains.

We have hitherto assumed that the only movement