STORAGE RESERVOIRS. 117 



not the physical difficulties, which stands in the way. It may be stated that as a general 

 rule a sufficient amount of storage can be artificially created in the valley of any stream 

 to rob its floods of their destructive character; but it is equally true that the benefits 

 to be gained will not ordinarily justify the cost. 



"The reason for this is plain. Floods are only occasional calamities at worst. 

 Probably on the majority of streams destructive floods do not -cur, on the average, 

 oftener than once in five years. Every reservoir built for the purpose of flood protec- 

 tion alone would mean the dedication of so much land to a condition of permanent 

 overflow in order that three or four times as much might be redeemed from occasional 

 overflow. One acre permanently inundated to rescue three or four acres from inunda- 

 tion of a few weeks once in three or four years, and this at a great cost, could not be 

 considered a wise proceeding, no matter how practicable it might be from engineering 

 considerations alone. The cost, coupled with the loss of so much land to industrial 

 uses, would be far greater than that of levees or other methods of flood protection. 



" In fact, the examples of natural reservoirs already cited, while they show con- 

 clusively the vast beneficial influence of large reservoirs upon the flow of streams, also 

 disclose the fatal obstacle to their successful imitation by man. In only very few places 

 has nature prepared sites where man can erect works which will create large bodies of 

 water, and even if she had done so the gain from utilizing them would not equal the loss. 

 The reservoir system of the Great Lakes involves the perpetual withdrawal from agri- 

 culture and industrial uses of an area nearly twice the size of the State of New York. 

 Were these areas not covered with water, but occupied as the surrounding country now 

 is, yet so fitted by nature that man, at slight expense, could convert them into great 

 lakes, as at present, the utter impossibility of such a measure is evident at a glance. 

 And so it will be found in general that the surface of the earth, where reservoirs could 

 be built on an extensive scale, is liable to be of more value in its present condition than 

 it ever could be if covered with water. 



" The construction of reservoirs for flood protection is not, therefore, to be expected, 

 except where the reservoir is to serve some other purpose as well, and inasmuch as such 

 purposes are not ordinarily extensive enough to develop systems of reservoirs, upon 

 which, rather than upon isolated works, the control of great floods depends, this large 

 control is hardly one of the possibilities of the future. The only probable exception 

 is that of a reservoir system on the watershed of the Missouri River, treated of in the 

 next section of this report. 



"For flood protection in isolated cases, however, and on a relatively small scale, 

 reservoirs will undoubtedly continue to be built, particularly when they serve other 

 purposes as well. From this point of view they will always be projects of public impor- 

 tance. The idea is well presented by the distinguished French engineer, P. Guillemain, 

 former inspector-general of public works in France, who holds that the creation of 

 reservoirs is of public utility in nearly all cases, either in flood prevention or in re-enforc- 

 ing low-water flow, and that whenever special interests, such as industrial uses, irriga- 



