76 LITEBATUEE ON TOERENTS. 



banks faced with stones. I have shown in another memoir that one may 

 obtain the same results more economically ; but wishing to occupy myself 

 at present only With secondary torrents, I confine myself to this. 



"Before proposing means of preventing or of repairing the ravages 

 which the secondary torrents make, it is necessary to know these torrents, 

 and for this purpose to take them up at their birth, to examine them in 

 their course, and in following them in the increase of their bed of deposit 

 year by year to show the enormous extent of the damage which they may 

 have occasioned. It is certain that a secondary torrent does but little or 

 no evil so long as it is shut up between steep banks. It is when it leaves 

 the lateral mountains to enter into the valley that it begins its ravages. 

 Let us examine how this comes about. 



" So long as the waters of a torrent are confined within steep banks they 

 roU on in a great body, drawing on with them not only gravel, but even 

 enormous rocks. Scarcely have they left the mountain, when, not being 

 sustained and kept together by the banks, they divide themselves into a 

 thousand little currents ; and then, so far from drawing on rocks, they scarcely 

 roU gravel along, and as their force diminishes more and more they scarcely 

 bear along to the principal torrents some grains of sand. This explains 

 'perfectly the form taken by the deposits formed by secondary ton-ents. 

 At the departure from the mountain this form is that of a portion of a cone, 

 the summit of which corresponds to the point where the torrent comes out 

 from the mountain. In effect, the waters, in quitting the mountain, have 

 still an acquired force which permits them to roll the rocks on to some dis- 

 tance ; in the second instance, this force being diminished, they deposit the 

 rocks and carry foi-ward only stones ; in the third instance, their force being 

 stiU further diminished, they abandon the stones and then carry on the gravel. 

 Thus, then, is formed a first deposit, which wiU be less and less considerable in 

 proportion to its distance from the mountain. In a second flood of the 

 torrents the waters get freely away, and the deposits of sand and of gravel 

 will increase less, always in this following a slope. In fine, the increase 

 may become so inconsiderable that the sides of the cone recede from the 

 mountain; then the torrent divides itself into two cun-ents, and soon there 

 comes to pass, at each of these two currents, what had occurred with the 

 prmcipal currents. Thus the fertile lands of the valley may disappear 

 under the heaps of stone and of sand ; as these torrents are greatly multi- 

 plied, there will come a day that their deposits, spreading out till becoming 

 conjoined, a whole vaUey will become sterile, and wiU not be able longer to 

 support its inhabitants. 



« We have seen that the secondary torrents do not deposit the gravel and 

 stones which they carried from the mountain ; but when their waters are 

 no longer confined by the banks— when they enter the valley— they spread 

 themselves over a great surface, and thus lose their force; they cannot carry 

 further the stones and the gravel, and these they abandon at a greater or less 

 distance from the mountain. This indicates to us the course to be followed 

 in order to control these ton-ents at their embouchure, and to prevent them 

 covering the land with gravel. 



" I would propose, then, in accordance with this principle :— First to dicr 

 a bed for the torrent in the deposit which has penetrated to the exit from the 

 mountain; second, to give little breadth to this bed, but great depth in 

 order that the waters may be there confined as they are in the natural bed 

 which the torrent has dug for itself in the mountain, and that they may 



