To understand our river systems we must go back to the time when strata formed by deposit of sediment in the sea were upheaved above the sea level. To take the simplest case, that of a single anticlinal axis fading off gradually at each end, we shall have a sort of turtle back of land emerged from the sea, as in figure 6, _aa_ being the anticlinal axis. From this ridge streams will run down on either side in the direction of the dip, their course being determined by some minor folds of the strata, or difference of hardness in the surface, or cracks formed during elevation. On each side of the dip-streams smaller ones will flow, more or less in the direction of the strike, and run into the main streams. Various irregularities, such as started the flow of the streams, will favour one or another. Consider three streams, _a_, _b_, _c_, and let us suppose the middle one the strongest, with greatest flow of water, and cutting down its bed most rapidly. Its side streams will become steeper and have more erosive force, and so will eat back their courses most rapidly until they strike the line of the streams on either side. Their steeper channels will then offer the best way for the upper waters of the streams they have cut to reach the sea; and these streams will consequently be tapped, and their head waters cut off to flow to the channel of the centre stream. We shall thus have for a second stage in the history a system such as is shown in fig. 7. The same process will continue till one river has tapped several others; and there will result the usual figure of a river and its tributaries, to which we are accustomed on our maps. We shall observe that tributaries do not as a rule gradually approach the central stream, but suddenly turn off at nearly a right angle from the direction in which they are flowing, and, after a longer or shorter course, join at another sharp angle a river flowing more or less parallel to their original direction. [Illustration: FIG. 6] [Illustration: FIG. 7] DEVELOPMENT OF RIVER SYSTEMS