If the river channel is soft and if the sea water carries no silt, it may, in passing up and down the river, become charged with silt and return to sea still carrying it. It thus has a scouring effect on the channel, and may deepen or widen it. If, owing, for instance, to the flattening of the bed slope in its lower reaches, the river tends to deposit its own silt in its tidal reach, the sea water may prevent this deposit. Thus, as regards silting in the tidal reach of the river, the tidal water of the sea has little prejudicial effect if it is silt-laden, and a beneficial effect if it is not. Silt is likely to deposit in the tidal reach of a river of uniform width, only in a case in which the river water carries much silt, and the slope is flat or cross-section great compared to that of the upper reach. Sea water is heavier than fresh water by about 2·4 per cent., and this, to some extent, prevents their mixing. At all stages of the flood tide the tendency at the point where the fresh water meets the salt water is for the fresh water to accumulate towards the surface and the sea water towards the bottom. When the tide begins to flow up the river there may be a low-level salt water current moving landward and a high-level fresh water current moving seaward, but this is quite a temporary state of affairs. The surface slope is landward, and the water moving seaward is not moving in obedience to the surface slope. It is only moving as a result of momentum previously acquired. The low-level current may have some extra velocity and extra scouring power, but this cannot be much, because the mean landward velocity of the whole stream must, owing to the internal resistances caused by the two currents, be less than it would be if there were not two currents. Moreover, the state of affairs is temporary. The two kinds of water mix eventually, and their temporary separation has no considerable effect on the general tendency of the river in the tidal reach to scour or to silt. A body of water included at any moment between any two cross-sections of the tidal portion of a river may not reach the sea during the next ebb tide. In this case it will flow back up the channel with the next flood tide, and so be kept moving up and down, getting nearer, however, to the sea at each tide. De Franchimont has shown (_Min. Proc. Inst. C.E._, vol. clx.) how a diagrammatic route-guide can be prepared for any tidal river to show pilots or captains of vessels the best times for starting on voyages up or down the river, and for passing each point on it.