Etudes by boipion gras. 79 



should abide but little. And in like wise, the trees which be planted along 

 the mountains do much deaden the violence of the waters that flow from 

 them.' Lombardini attaches great importance to the planting of rows of 

 trees transversely to the current on grounds subject to overflow. — Esame 

 degli Studi sul Tevere, § 53, and Appendice, §§ 33, 34." 



In 1857 there appeared, in the Annates des Fonts et CkaussSes and in the 

 Annates des Mines, Etudes sur les Torrents des Sautes Alpes, by M. Scipion 

 Gras, ingenieur des mines. Of this work the following analysis is given by 

 M. Cezanne, in his supplement to the work of Surell : — " After having 

 defined torrents the author divides their course into four parts — hassin de 

 reception, canal de reception, lit de dejection, and lit d'icoulement. M. Scipion 

 Gras distinguishes amongst torrents at the bed of deposit four classes, 

 according to the character of the basins drained by them. The study of 

 the laws in accordance with which solid bodies are swept away by floods 

 leads him to the conclusion that there are two distinct modes of operation 

 — transport en masse and transport partiel — ^the former efiected by floods of 

 great body and strength, the second by floods of a medium character. 

 These different operations produce contrary effects upon the bed of deposit ; 

 the great floods, as they exhaust themselves, deposit over this a layer of 

 clay and gravel, over or through which the waters spread themselves in 

 thin sheets ; the lesser floods, on the contrary, dig down into the bed of 

 deposit and plough in it a channel for themselves, after having conveyed 

 thither the more comminuted materials referred to." 



Upon which M. Cezanne remarks that M. Gras does not occupy himself 

 much with the basin drained by the torrent, the special subject of study by 

 him being torrents the basins of which are not susceptible of being planted ; 

 and impressed with the evils resulting from the dejection of detritus, he 

 seeks to modify the natural advance of the bed of gravel, and discusses the 

 two methods generally employed to effect this, characterized respectively by 

 the employment of dykes and of barrages, or embankments and barriers. He 

 expresses himself very decidedly in favour of barrages, and he thus sums up 

 his opinion on the point :— " In short, the first proceeding (I'endiguementJ, 

 as a means of suppressing a bed of deposit, is often impracticable, or at 

 least the success of it is dubious ; when it does succeed, it only carries the 

 mischief elsewhere. It is, then, one which is very defective, and which ought 

 to he abandoned. 



" There remains the second course of proceedure ; we have demonstrated 

 its practicability, its applicability to all torrents, and its freedom from the 

 drawbacks attaching to the first. And here begins the most important part 

 of our task." 



This quotation (says C&anne) is characteristic of the method of procedure 

 of M. Gras, which is pre-eminently systematic ; he observes natural pheno- 

 nema with great accuracy and precision ; he then proceeds to distinguish, 

 to classify, and finally to bring aU his energy to bear upon a single and 

 exclusive system. " But," says M. C6zanne, " the consideration of nature 

 inspires one with a dislike and opposition to all systematic formulas. It 

 may be well to run down embankments, and extol barriers ; but it is not 

 less the case that there are circumstances in which the practical man wiU see 

 at a glance that there embankments are better adapted to meet the case 

 than are barriers. And it is necessary, in a study of this kind, to avoid all 



