HYDROLOGY OF NEW YORK 



471 



river would prevent the broken ice from passing rapidly away, 

 and (3) that floods frequently come with little warning, means 

 that to certainly prevent ice dams the ice must be continually 

 broken during these months, a condition involving heavy cost for 

 furnishing and operating powerful ice-breaking steamers. More- 

 over, when we consider that ice dams sometimes obstruct the river 

 for several miles in extent, the impracticability of removing them 

 by mechanical means seems sufficiently obvious. 



While it is doubtless possible to mitigate floods in this section 

 of the river by isolating the flooded districts by dykes or levees, 

 which would necessitate intercepting sewers, as well as the pump- 

 ing of the surface and sewer drainage whenever there is a slight 

 freshet in the river, nevertheless the most rational treatment is 

 believed to be by storage of the flood-flow in the upper tributaries 

 of the stream. 



The opinion has been expressed that storage should l>e devel- 

 oped proportionate to the catchment area on the several tributaries 

 of the Hudson above Troy, and while this is theoretically true, 

 the writer imagines that the question of cost will finally come in, 

 very greatly modifying any purely theoretical deductions based 

 on this view. To show how material an element in the problem 

 cost may become, it may be mentioned that reservoirs have been 

 at various times considered costing from about $20 per million 

 cubic feet stored, to as high as |200 or $300 per million cubic feet. 

 The cost of such reservoirs as compared with the cost of reservoirs 

 for municipal purposes, even at these prices, is very low, the cost 

 per million gallons frequently rising higher than the cost per 

 million cubic feet here proposed. But nobody is likely to expend 

 a large amount of money in order to meet the theoretical require- 

 ment of equal storage in all parts of a catchment area when con- 

 siderably less money will build a storage reservoir of equal 

 capacity elsewhere. Practically, therefore, it is impossible to reg- 

 ulate as large a catchment as that of the Hudson river to anything 

 like a uniform flow throughout all of its tributaries. The mere 

 matter of cost alone will militate against such a conclusion. In 

 a few words the conditions are that in the upper Hudson, above 

 Glens Falls, very extensive storage reservoirs may be constructed, 

 either on the main river or its chief tributaries, Sacandaga. 

 Schroon and Cedar rivers, which were estimated in 1895, for the 

 whole system, to cost something like $60 per million cubic feet 



