Cli ff s - 535 



When we think of the meaning of this, it at once 

 explains the whole history of these constantly recurring 

 forms, in all winding rivers that flow between rocky 

 banks higher than broad alluvial plains and deltas. 

 Take the history of the curve, fig. 107, as an example. 

 On a high tableland the river, r, at an early period 

 of its history, flowed where it is marked in fig. 110, 

 the beginning of the curve, c, fig. 107, having al- 

 ready been established, but without any high cliffs. 

 Then the stream, being driven with force against the 

 concave curve, c, by degrees cut it back, we shall sup- 

 pose, to c 1 , at the same time deepening its channel. A 



cliff was thus commenced at c 1 , and, as the river was 

 changing its bed by constant encroachment in the same 

 direction, a gentle slope, s, began to be established, 

 facing the cliff c 1 , and so, on and on, through long ages, 

 to c 2 , c 3 , and c 4 , where the present cliff stands, itself as 

 temporary as its smaller predecessors. This is the 

 reason why in river curves, the concave side of the curve 

 is so often opposed by a high rocky bank, while the con- 

 vex side so generally presents a long gentle slope, s s, 

 often more or less covered with alluvial detritus. In 

 countries free of glacial debris, these effects are often 

 best seen in their perfect simplicity ; and in this way 

 the Moselle, and the Seine near Eouen are, so to speak, 

 model rivers. In many a British river it is clearly 

 seen on the Wye in South Wales, in many a river and 



