92 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [DeC. 18, 



have done so before the valley was filled with water, because seve- 

 ral of the lateral streams had cut back the slopes to a considerable 

 depth before the formation of the lake, so that the " roads " wind 

 into the recesses thus produced. 



Given, then, a valley with sides sloping at the angle at which the 

 matter of which they are composed will stand, what will happen 

 if the lower end is dammed up, and a lake formed ? 



The deep still water below the surface would effect no change ; but 

 although in a narrow mountain-locked valley no very great dis- 

 turbance of the water would take place, still the waves would have 

 a certain amount of rolling-power. Now, suppose a stone at the 

 edge of the lake moved by the water a Kttle either to one side or 

 the other; being on a slope, of course it would have a tendency 

 downwards ; and of the stones once moved, a large proportion would, 

 when once disturbed, roU down at once to the bottom of the vaUey. 

 A.ny one may try this on a small scale, as I have done, and satisfy 

 himself by experiment that this would be the result. 



It is evident that in this way two fresh angles of slope would be 

 produced — one immediately below the water at a less inclination 

 than before, and the other above the water at a steeper one. But 

 the original angle was that at which the material would stand, and 

 the steeper slope above the water would therefore immediately begin 

 to crumble back until it reacquired the slope of equilibrium, which, 

 as the matter has had time to settle and become somewhat consoli- 

 dated, would naturally be slightly steeper than before*. As soon as 

 the slope of repose was reacquired, the crumbling would be checked, 

 the two slopes would have acquired a condition of repose ; and even 

 if there were a steeper part above, supplying fresh matter, the impe- 

 tus of fall would carry it in most cases over the road, which there- 

 fore would not be obhterated. 



A block is now shown which fell a year or two ago, and which, 

 far from stopping on the roads, has been driven by its impetus part 

 of the way up the opposite hill. 



Mr. Darwin remarks, as we have seen, that the shelves " contain 

 fewer weU-rounded pebbles at the greater heights than would be 

 expected on any theory of their origin." If, however, my idea is 

 correct, the stones remaining on the shelves are those which have 

 not been much moved by the water, and would therefore naturally 

 not show much trace of wear. 



No suggestion, I think, has yet been made to account for the 

 uniformity in the breadth of the roads. This also follows, however, 

 as a consequence of my theory ; for as the angle of the roads neces- 

 sarily varies within narrow hmits, and the depth to which the water 

 is disturbed in different parts of the valley does not greatly differ, 

 the breadth of the roads, which commence from the water-level and 

 reach down as far as the disturbed water reaches, cannot vary much. 



It vnll be observed that, according to Macculloch, not only is each 

 road pretty uniform in breadth, but the three different roads are also 



* I think it probable that this process would be facilitated by the absence of 

 vegetation, owing to the extreme cold. 



