REGULARIZATIOX. 57 



The results of all works of regularization have sufficiently demonstrated this. The 

 slope of a river, whatever it may be, is never the same in all parts of its course; its 

 low-water surface, in particular, forms a broken line with the least slope in the deep 

 pools, and a much greater fall where there are bars. In lowering the level above the 

 bars new obstructions are developed, shoals which were unnoticed before, as they were 

 hidden under an ample depth of water. This, however, is not the result sought, and 

 to counteract such evils it is necessary to reduce the section of flow of the river above 

 the bar, and at the same time prevent the current from digging the new channel too 

 deep. For this latter purpose the bottom must be covered with fascines, riprap, etc. 



' ' Besides the fact that new bars are brought to light in the upper pool, daily expe- 

 rience shows that the alluvial matter which the current takes away from the contracted 

 section goes to form new shoals farther down stream. Thus the regularization of a 

 river, when only attempted at places where there are bars, does not ordinarily give the 

 desired results. After a certain time it becomes absolutely necessary to extend the 

 contracting dikes for the whole length of the river. We have numerous examples of 

 this in the rivers of western Europe, which have been subjected to works of systematic 

 regularization. Upon nearly all we find a continuous protection of the banks for their 

 whole length, and works designed to narrow the river-bed follow one another without 

 interruption. 



" In all cases the improvements that may be expected from these works cannot pass 

 certain limits which it is very difficult, not to say impossible, to predetermine; the 

 results must therefore be unknown. 



" It may also be asked if the form of the regularized bed established in low water 

 will resist great floods. It seems certain that works of this nature can be preserved, 

 but only by a vigilant maintenance. This maintenance is, moreover, the more necessary 

 and expensive as the works are of more recent construction or isolated, and not yet 

 consolidated into a system of complete regularization. Serious injuries happening 

 to these works may result in changes in the channel, especially where it is naturally 

 of a shifting character. It is only just, however, to observe that there are no works 

 which do not require repair, and that are not subject to injuries affecting more or less 

 widely the regime of the lines of communication upon which they are established. 



" While the authors of the first projects followed a uniformity of method which 

 is not compatible with the natural regime of streams (their object being to secure a 

 uniform normal profile, and uniformity of slope), the aim to-day in works of regularization 

 is to obtain a continuity as perfect as possible in the longitudinal and transverse 

 profiles as well as in the outline of the banks and of the center line. Every abrupt 

 change in the outline of the bed is considered as forming an obstacle to navigation. . . . 



" Submerged spur-dikes may often be employed with success in preventing or 

 correcting the inconveniences resulting from the construction of a bridge. Thus, on the 

 Saone at Lyons, the bridge of Ainay, now under reconstruction, the arches of which 



