Fisher — Glacial Origin of Denudation. 485 



work has been effected, simultaneously from its commencement, by 

 atmospheric causes on the sides and river action at the bottom, we 

 arrive at the conclusion that the abrading power of the atmosphere 

 far exceeds that of the river, which, when once the sides have 

 reached the angle of repose, seems very improbable. 



Sir Eoderick Murchison, in his address to the geological section 

 at the last meeting but one of the British Association, said: "I do 

 not believe that tlie wear and tear due to atmospheric subaerial 

 erosive agency could, even after operating for countless ages, have 

 originated and deepened any of the valleys and gorges which occur 

 in countries as flat as the tract in which we are now assembled" 

 (which was the neighbourhood of Birmingham). In this sentence 

 the argument rests upon the flatness of the surface. It is impossible 

 to conceive a very rapid flow of water caused by rain upon a moder- 

 ately flat surface of upland. A current of water of a given velocity 

 cannot move an ordinary pebble which is above a given size. A 

 current which can just move a given pebble can, roughly speaking, 

 move any smaller pebble of the same or less specific gravity. Yet, 

 in a mixed gravel, pebbles too large for the flow of water to move 

 will, nevertheless, progress by the washing away of smaller stuff in 

 front of them. A temporary inclined plane is thus formed, down 

 which the pebble rolls partly by its own weight. Thus it happens 

 that where the stream is too slow to move the larger pebbles, if 

 there be sufficient fall in the ground they nevertheless slowly pro- 

 gress, but the small stuff moves off faster, and the pebbles accumu- 

 late behind. Thus the gravelly bottoms of rivers are formed. If 

 then, the degradation of gravelly strata has been produced by rain 

 wash, we ought to have most of the larger pebbles corresponding to 

 the denuded material left behind. This is certainly not the case. 

 A still more obvious difficulty is the removal of large blocks from 

 escarpments. Take, for instance, such blocks as those which form 

 the upper portion of the Greensand in Dorsetshire, which are 

 notably seen at Eggerdon-hill and Bincombe-down. What has 

 become of those which must have fallen as the hills were cut back ? 

 "We see a few lying about on one side of Eggerdon, where a landslip 

 has occurred, but those which must have fallen as the Chalk was 

 degraded, if by atmospheric agency, ought to cover the spurs of the 

 hill and the bottoms of the valleys. 



Let us now suppose the land to have been covered with a sheet of 

 ice, what would be the character of the denudation ? I endeavoured 

 to explain, in a letter to the "Eeader,"i yihoi I conceived to be 

 the modus operandi of ice in the particular case of excavating a lake 

 basin, and Mr. Jamieson has very well illustrated the nature of the 

 motion of a sheet of ice by that of a quantity of corn heaped up 

 upon a floor.2 As its height is increased the grains in contact with 

 the floor move outwards from the centre. Provided, then, that a 

 sufficient thickness of snow were supplied, the ice would always 

 move outwards from some central region towards the sea. 



^ Reader, April 25, 1865. 2 Quart. Joum. Geol. Soc., Vol. xxi., p. 166. 



