STATESIENT BY Jl. StTKELL. 239 



energetic to destroy torrents already formed. One sees also that the 

 injurious result of the removal of woods is not only to open every 

 where the soU to new torrents, but that it augments the violence of 

 those which exist, and resuscitates those which appear completely 

 extinct. We may then sum up the influence which forests exercise 

 on torrents already formed on two facts, parallel to those which sum 

 up their influence on lands where the torrents have not yet appeared. 

 (1.) The presence of a forest on a soil prevents the formation of a 

 torrent there. (2.) The destruction of forests leaves them subject to 

 become the prey of torrents. Nor is there in this any thing for 

 which we may find it difficult to account." 



He proceeds then to explain the modus operandi whereby forests 

 produce such effects : " When the trees fix themselves in the soil the 

 roots consolidate this, interlacing it with a thousand fibres ; their 

 branches protect it, as would a buckler, againet the shock of the 

 heavy rains ; and their trunks, and at the same time suckers, 

 brambles, and that multitude of shrubs of all kinds which grow at 

 their base, oppose additional obstacles to the currents which would 

 tend to wash it away. The effect of all this vegetation is thus to 

 cover the soil, in its nature mobile, with an envelope more solid and 

 less liable to be washed away. Besides, it divides the currents and 

 disperses them over the whole surface of the ground, which keeps 

 them from going off in a body in the lines of the thalweg and meet- 

 ing there, which would be the case if they flowed freely over the 

 smooth surface of a denuded ground. Finally, it absorbs a portion 

 of the water which is imbibed in the spongy humus, and so far it 

 diminishes the sum of the washing away forces. 



" It follows from this that a forest, in estabUshing itself on a 

 mountain, actually modifies the surface of the ground, which alone is 

 in contact with atmospheric agents, and all the conditions find them- 

 selves then modified as they would be if a primitive formation had 

 been substituted for a formation totally different. Whence it is not 

 more astonishing to see the same soil alternately cut up or free from 

 torrents, according as it is despoiled or clothed with forests, than it 

 is astonishing to see torrents cease when we come to primitive 

 formations, or reappear suddenly on friable limestone. 



" In accordance with this we find — first, the development of forests 

 brings about the extinction of torrents ; second, the destruction ot 

 forests redoubles the violence of torrents, and may even cause them 

 to reappear. And nothing is more easy than to explain these new 

 actions. It will be remembered what are the causes which call forth 



