PT. IK. OR POISONED BY VEGETABLE GROWTH? 161 



for what distances, it has been projected. But when 

 the torrent is turbid with the wash of rain, we can 

 hear its huge cannon-ball s ratthng down, and grinding 

 each other and tlieir rocky bed and banks till what 

 has started from the mountain's brow as a huge rock 

 arrives at the sea in the form of pebbles, or of sand. 

 For although, as the flood of rain subsides, tlie flow of 

 boulder-stones ceases, this is only for a time : each rain 

 sets them on a stage on their journey, as, in lower 

 levels and gentler gradients, I have said of soil, and 

 the more minute particles formed by disintegration and 

 vegetable chemistry. 



A very slight difference in hardness of surface, or 

 thickness of vegetation, at the brow of the hill may 

 concentrate the wash of rain into a stream ; this forms 

 a channel, which is fed by rain from its sides. And 

 though all possible natural accidents of this sort might 

 have been supposed to have taken place long ago in all 

 but volcanic, or newly raised regions, Lyell, quoting 

 Sir T. D. Lander's account of the great floods in 

 Morayshire, August 1829, says; 'Some new ravines 

 were formed on the sides of mountains where no 

 streams had previously flowed,'* and ancient river 

 channels which had never been filled from time imme 

 morial gave passage to a copious flood.' 



And again, Lyell, in giving an account of the 

 formation of new ravines by heavy rains on the 28th 



* ' Quodqne fait campus, vallem decnrsus aquarum 

 Fecit.' 



M 



