Formation of Alluvial Inlands in Rivers. 19 



breaches in the dyke, and (working backwards, as in all hori- 

 zontal water-courses), cut deep and long channels in the allu- 

 vium, perhaps leaving islands between them. In this state 

 things remain, until another flood covers all again, and either 

 fills the channels with a part of its load, or sweeps away the 

 whole, and deposites it lower down in the course of the river. In 

 the valley of the Rhine, about Bonn, and in the elevated plain 

 between that river and Juliers, now watered by the Rohr, sec- 

 tions of the gravel sometimes exhibit such a channel filled with 

 sand (See Fig. 16) ; and, wherever this appearance is found, 

 although the deposites be called diluvium, on account of their 

 elevated situation, or any other circumstance, it is evident that 

 they can only be the alluvium of a former age. 



There is another mode of action which modifies the even 

 slope of river deposites. Let us suppose a stream of any magni- 

 tude flowing between steep banks either of rock in place or of 

 alluvium. When the quantity of water is so great as to be sub- 

 ject to reverberation from the banks, it is driven with its load of 

 sand, stones, &c. from each side towards the middle of the cur- 

 rent, where, consequently, the water is deepest, and its action 

 most powerful. Hence the deposite will, previous to the sub- 

 sidence of the stream, be most considerable in the middle of the 

 bed, sloping from the middle towards the sides. On the other 

 hand, so soon as the subsidence commences, the middle will be- 

 come the most shallow and languid part, and the principal ac- 

 tion will be on the two sides. If a level part of the bed be suc- 

 ceeded by a declivity, it aids the removal of the sediment in such 

 a way that the central ridge of sediment is left terminating in a 

 slope both towards the two sides of the channel and towards the 

 declivity. (See Fig. 17). Over this tongue the water flows shal- 

 lower and shallower as it subsides, and at length the tono-ue 

 rises like an island between two branches of the river. Mean 

 while these two branches, to which the principal action is trans- 

 ferred, will, according to circumstances, either deposite their se- 

 diment so as to fill up, in part, the two channels between the 

 central ridge of sediment and the banks of the river, or they 

 will erode the base of that central ridge, and carry away its 

 materials. 



Thus It appears that the same increase of water may, accord- 



