Mr Hwjh Miller on River- Terracing. 303 



cattle. We satisfied ourselves that all the essential effects 

 of river-terracing had been produced by this streamlet : — 



1. Without having been induced to fall from terrace to 



terrace by repeated falls in the level of the river. 



2. Without having at any time filled its little valley u]> 



to the rim or base of its highest terrace, so tliat it 

 might fall from terrace to terrace in subsiding. 



3. Without having even flooded its valley to such an 



extent during the excavating of it, that its terraces 

 were indented flood-margins. 



4. Without having been sometimes glutted with sediment 



too abundant to remove, and at others permitted to 

 intersect the surplus deposits. 



5. But simply, by planation at different levels and to 



varying breadths, with a tendency to narrow the 

 field of its operations as it deepened its course. 



The existence of terraces is thus compatible with absolute 

 uniformity in the water supply. The mere breadth of terraced 

 valleys is no criterion of the volume of water that has occu- 

 pied them, any more than the mile-and-a-half's depth of the 

 Grand Canon is a proof that it once held that impossibly 

 vast column of river- water. The evidence of " pluvial " and 

 torrential periods must be sought in other directions. In the 

 valleys which we know best, the highest line of the river- 

 gravels, 60 or 70 feet above the present river-bed, is not a 

 terrace, but a shelving, shore-like slope.* It would appear 

 that, in these cases, the first flowing of the post-glacial 

 rivers was not attended by terracing. Their shores may 

 have been heavily plated over with ice, a sort of ice-armour, 

 warding off the attacks of the water. Or, more probably 

 still, the rivers may have been liable, like some arctic and 

 sub-arctic streams of the present day, to freeze almost en 

 masse, thus being glutted with such vast bodies of massive 



* It is necessary to distinguish carefully between the liighest river-flats and 

 approximate flats of some other kind -which the rivers found made to their 

 hand, and left here and there on either side of them. These, in this country, 

 consist either of boulder clay or of glacial gravel. When Gumselius speaks of 

 Norwegian terraces as being sometimes overlaid by moraines, it would appear 

 that the terrace is not a river terrace, but of some antecedent kind. 



