80 WTEEATURB ON TOHRBNTS, 



special pleading in support of the absolute superiority of one system or of 

 another, and to confine remarks to showing clearly in what circumstancs 

 either of them should be preferred. 



"M. Gras being so decidedly in favour of the exclusive use of barrages, or 

 barriers, in reference to the two different categories of floods established by 

 him, recommends, according as it may be desired to effect a complete or a 

 partia} retention of gravel, the construction of submergible barriers in the 

 latter case, and insubmergible barriers in the former. 



"The latter, insubmergible — so designated, although actually overflowed 

 by the torrent, and expected and intended to be so at times, and it may be 

 frequently — belong to a class of embankments which have been long in use. 

 Erected in some favourable position in the gorges, they are designed to effect 

 a deposit of gravel directed up the river. If the reservoir designed for this 

 deposit be very considerable, if the transport be slow, it may tell effectively 

 for some distance below for several years. 



" Submergible barrages constitute, strictly speaking, the system which M. 

 Gras claims the credit of originating. This system is based on a very 

 delicate analysis of the effect of floods, which shows that high waters only 

 acquire their full force in a narrow channel in which they are confined. If 

 they be allowed or compelled to spread themselves out, their force is 

 diminished, and the larger materials which were being borne along by them 

 are deposited. To compel them to do this — to spread themselves out— it 

 is only necessary to raise, on a widening of the bed, a horizontal siU, which 

 cannot be washed away, worn down, or furrowed ; the waters, then, not 

 being able to concentrate themselves in any place at a lower depth than 

 that of the whole siU, spread themselves in a sheet over the sill, and a 

 deposit up the river follows as a consequence. 



" After a great flood, such as may be of ooccasional but comparatively rare 

 occurrence, floods of lesser magnitude, which are much more frequent, go 

 over this deposit anew, and do on a lesser scale what the greater flood has 

 done on a greater, excepting that such large blocks as could only be carried 

 along by a great flood will remain in the places above the barrier in which 

 they had been left. And the effect of the whole will be, that great floods 

 will be less disastrous, the work done by them being effected by a great 

 number of floods, the consequences of which are innocuous. 



" It is not necessary that these barrages should be of great height, nor, 

 consequently, of great solidity ; it is enough that their upper surface sustain 

 the friction of the pebbles carried down by the flood, and that their base 

 can sustain the slight water-fall which they occasion. 



" From this it may be seen," says M. Cezanne, " that the system of oper- 

 ations proposed by M. Soipion Gras is the very opposite of the course 

 formerly followed, in so far as formerly, when a dam or barrier was to be 

 erected, a narrow depth in the bed of the current was selected, that the 

 structure being short there might be given to it, at little expense, the thick- 

 ness necessary to enable it to resist the violent action of the water. He 

 recommends to select expansions in the bed of the current, and even 

 proposes to erect, on the cones of dejection, works of the same kind, which 

 he calls barrages radiers. To secure the plain of Bourg cPOisans, in the 

 basin of the former lake Saint Laurent, against ravages by the Bomanche 

 and the Veneon, which debouche each by a different gorge, he proposes the 

 erection of such barriers, spread out horizontally, the length of which should 

 be not less than 763 metres, or 2500 feet, upon which he supposes that the 



