124 LITERATURE Oy TORRENTS. 



Embrun, to which I shall afterwards have occasion to refer ; and he proceeds 

 in the next division of the work to the consideration of the torrential 

 phenomena in great rivers. 



He says, that with the knowledge which has been attained in regard to 

 mountain torrents by observation and experience, the question, How can 

 thev be controlled and stifled? may be considered as settled. But the 

 same cannot be affirmed in regard to rivers, which throughout their course 

 are governed by the same laws — both those governing or regulating the 

 flow of water and those governing or regulating the torrential phenomena — 

 any apparent modification being attributable to the greater quantity of 

 water, its greater fluidity or lesser viscosity, the lesser rapidity' of its flow, 

 and to the more extended reaches throughout which this maintains a 

 uniformity. All that has been done hitherto in regard to rivers has related 

 to the delivery or quantity of water in flow. Attention has not been given 

 to the perturbations of torrentiality and to effects produced by these. 



In the regulation of river currents it is desireable that the delivery or 

 water in flow should be equalized, and all perturbation in that flow reduced 

 as much as possible ; and he says, — " This double result is obtained by the 

 reboisement of the mountains, but it may be brought about in two difierent 

 ways. When the object to be accomplished by the planting of forests is to 

 equalize the delivery or quantity of water it is necessary to extend the 

 boisement over extensive areas, comprising the greater portion of the basin. 

 If, on the other hand, it is the perturbations in the flow upon which it is 

 desired to act, it is necessary to concentrate the reboisement on properly 

 selected points, and, it may be, to strengthen the action of these by a series 

 of the artificial accessory works employed for the extinction of torrents. 



" It is this latter system which is alone efficacious and practicable in acting 

 on a great water-course. 



" In a basin such as that of the Loire, for example, there might be planted 

 a hundred thousand hectares of land without perceptibly modifying the 

 regime of the river, if the lands were not selected with intelligence with a 

 view to the consolidation of the soil and to the accomplishment of the 

 end desired. A study of the whole course of the river, and a comprehensive 

 plan of operations founded thereon, is absolutely requisite as a preliminary 

 measure. 



" It is necessary from the first thoroughly to know the regime of the parti- 

 cular water-course, and to ascertain its torrentiality. This may be accom- 

 plished by a general reconnoissance. 



" All the affluents should be classified in a hydrological chart, according to 

 their degree of torrentiality. 



" Most frequently an inspection of the state of the confluence suffices to 

 reveal the regime of ihe affluent. When such is torrential it will be found 

 to straggle oyer an extended bed before flowing into the main stream. 



" By subjecting, then, every one of the affluents to such an examination, 

 and following out this in all the upper ramifications of the river, it is easy 

 to determine what are the main centres of production of the stone or clay 

 materials borne along by the river, which are the causes of the perturbation 

 which have to be fought. 



" By this procedure the evil is localized, determined, and circumscribed ; 

 and it is often astonishing to find how limited in extent, compared with the 

 area of the basin, are the whole of the sources whence the gravel is obtained. 



" By such a procedure the operation is not left to chance. All is done 



