EVILS FOLLOWING DESTRUCTION OF FORESTS. 91 



water and retain the matter which the water was taking 

 with it in its course." The chopped straw which we 

 knead into the clay, is it not designed to augment the 

 solidity and consistency of the clay masonry ? What else 

 does art do here than copy nature ? Let the engineers who 

 think we exaggerate cast an eye over their accounts, and 

 compare the value which they assign to rubbish under 

 woods, and to rubbish under uncovered land. They will 

 find in the considerable augmentation of the former the 

 measure of the action which we are seeking to demon- 

 strate.' 



The action of trees in preventing avalanches is two-fold. 

 The interlacing of the roots tends greatly to consolidate 

 even arable earth ; and while most landslips are 

 occasioned by the undermining to a short depth of the 

 superficial layer of a slope, this is generally occasioned by 

 a rush of water produced by the rainfall in some more 

 elevated locality ; and the unvarying effect of reboisement 

 in the basin of reception as to prevent the formation in 

 this way of torrents. 



In regard to erosions thus produced, M. Surell writes in 

 his Etude sur les Torrents des Hautes-Alpes : 



' The torrent which dashes a great body of water over 

 very steep slopes undermines and eats away with fury the 

 base of the banks. These fall in, and little by little pull 

 down towards the bed the adjoining property, which is 

 finally engulfed by the waters. As the banks are generally 

 very deep, their fall brings in its train effects the results 

 of which extend far from the spot. All the surrounding 

 land is disturbed. Some portions undermined subside, 

 others slip, others break away, leaving deep crevices. 

 Along the two banks of the torrent may be seen large 

 chinks or rents running parallel to the bed. These sub- 

 sidences, these rents, and this disturbance spread from 

 place to place, repeat themselves to incredible distances, 

 and end by including the whole sides of the mountain 

 within the range of the effects. There are many quarters 



