H. R. Roivorth—A Great Post- Glacial Flood. 419 



namely, that the water in a river has a uniform motion. This is 

 true in one sense, but it is very misleading in another. The most 

 elementary knowledge of a fast river obtained in trying to force 

 a boat against the stream is enough to show that, if we take a 

 vertical section of it, at any given point, the water in all parts 

 of that section is not flowing with uniform velocity. The current 

 is very rapid in the middle of the river, but dies away as we near 

 the banks, and in many cases is barely perceptible there. The drag 

 of the banks is proverbial, and greatly assists navigation up stream ; 

 but the drag of the bank is really the drag of the river bed, and, as 

 a matter of fact, a swift river is a swift current flowing through a 

 cushion of water moving much more slowly. This is a well-accepted 

 fact in hydrostatics, and has been confirmed by numerous careful 

 experiments. In the case of deep rivers the cushion probably moves 

 hardly at all, just as the Gulf Stream is a more or less rapid current 

 of warm water flowing in a cushion of much colder and much more 

 quiescent water. It is this properly of rivers which in my view pro- 

 bably explains how in the case of the Neva, the Volga and other rivers 

 the freezing takes place first at the bottom, whence detached plates 

 of ice rise to the surface and get united together, forming a natural 

 bridge over a rapidly flowing current. The water in rapid motion 

 would probably never freeze, if that which is much more quiescent 

 near the bottom, which, therefore, freezes much more slowly, did 

 not furnish it with a constant stream of materials for making itself 

 an ice bridge. This has probably been suggested before, although 

 I do not remember to have seen it. It is only mentioned here by 

 the way. The real burden of my argument is that we must not 

 suppose that, because a river is flowing with a full current in 

 mid-stream, therefore it is denuding its banks and channel in a 

 corresponding fashion. It would seem that large rivers do hardly 

 any work in denuding the land, except in their head streams. It 

 is in these that the work of grinding down and of denudation 

 takes place, and the main stream does little more than bear along 

 the mud, etc., in suspension which has been poured into it by its 

 tributaries. All the great rivers known to me are, as regards their 

 main channel, conservative and recuperative, and not destructive. 



While such a rainfall as we have been discussing would not in 

 my view increase the denuding powers of the rivers, it ought 

 undoubtedly to have had a very potent effect in scouring the general 

 surface of the land. It would assuredly have washed the high 

 plateaux clean of their mantle of fine loam, and would have swept 

 it away into the rivers, and thence into the North Sea. It might 

 have left the larger blocks of stone there, but the fine loam would 

 certainly have gone; but this is just the reverse of what we in fact 

 find. As a matter of fact, the heavy debris is all in the valleys, 

 while the uplands are mantled and covered with fine loam. Again, 

 such a rainfall would no doubt have carved enormous ravines and 

 rents in the contour of the valleys at rapid intervals, when the 

 immense mass of water falling on the plateaux made itself dis- 

 charging channels into the main river, thus breaking the continuity 



