i84 



SCIENCE. 



[Vol. XIX. No. 478 



the proposition that every farm may have a reservoir, and 

 see how it will figure out. For an average annual depth of 

 four or five feet the water will spread over several acres, 

 certainly not less than five acres. On a farm of 160 acres 

 the catchment basin cannot be more than 32 times as large 

 as the reservoir. Drawing from the lands of one's neigh- 

 bors cannot be counted upon. Your neighbor below will be 

 as likely to draw from your land as you are to draw from 

 your neighbor above. The chances are even, and, in the 

 general summing up of catchment areas, each can only count 

 upon his own. Indeed he cannot count upon all of his own 

 land, for, if it is all devoted to gathering and storing the water, 

 where is the Geld to be irrigated ? That must lie below the 

 reservoir, as the catchment basin must lie above it. This 

 simple matter of levels imposes another rigid limitation upon 

 successful storage. Tillage of the catchment basin, causing 

 greater absorption of the rainfall — possibly complete ab- 

 sorption of it — is another contingency which may defeat 

 storage. 



If thefai'mer owns a half section, 328 acres, and if we make 

 due allowance for irrigated fields, and for slopes which face 

 away from the reservoir, he may possibly get a ratio as high 

 as 50:1. This is not enough for successful storage. On a 

 section, 640 acres, it might be as high as 100:1, if the slopes 

 were happily disposed. Instead, therefore, of a possible res- 

 ervoir on every farm, it is clear that only very large farms 

 having a favorable topography can enjoy this luxury. 

 The ratio 100:1 probably represents the maximum of favor- 

 able conditions which can ordinarily be realized on the 

 plains. Hence we need not consider the possible results of 

 any higher ratio. Nor need we go below the ratio 50:1, 

 since that is already below the requirements of successful 

 storage. 



It appears then that, instead of the ratio of catchment to 

 the storage area being a matter of choice, it is subject to quite 

 narrow limitations. 



We set out to seek quantitative results. By using data 

 given above for evaporation, run-off, and seepage, which are 

 believed to be fairly good approximations to the actual values 

 of those factors, we may construct' the following table: — 



Table showing the annual average depth of water for ratios varying from 

 50;1 to 100:1, and for rainfall varying from one to two feet, the annual evapora- 

 tion being Ave feet, seepage two feet, and the run-oft 7.5 per cent. 



This table must not be taken to mean more than was in- 

 tended. "None" does not mean that a reservoir under the 

 given conditions would not contain water at any time in 

 the whole year. It might be full after a storm, yet the aver- 

 age expectation of finding water there at any date when it is 

 needed for irrigation is correctly expressed by zero. 



I The formula for computation is t^ (e-|-s) = D, in which B = 



rainfall, — = run off, )■' = ratio of basin to reservoir, e = evaporation, s = 

 seepage, and D = annual average depth of water resulting frcm the given 

 conditions. * 



The table is intended merely for a quantitative expression 

 of results which will follow if the assumed data are fairly 

 correct. And, if they are somewhat erroneous, whoever 

 knows a more accurate value for any factor can readily in- 

 sert it, and correct the table. Quantitative expressions, even 

 when based upon assumptions and hypotheses, are more in- 

 structive than vague and speculative generalizations. This 

 table, for instance, shows certain limitations of water storage 

 so narrow and rigid that any errors which are likely to be 

 detected in the assumed data will not overcome them. 



To specify some of these limitations, take the first column 

 of the table. It means unmistakably that no storage can be 

 made from a rainfall of one foot. The highest ratio, that 

 of 100:1, a ratio which can seldom be realized, gives only 

 six inches as the permanent average depth of water in the 

 reservoir. None of the assumed data can very well be so 

 far astray that its correction will raise the amount to a re- 

 liable irrigation head of water. Possibly full at one time, 

 but dry as a powder-horn at other times, such a reservoir 

 would be useless, because it would be unreliable. Certainty 

 — that most valuable feature of farming by irrigation as 

 opposed to an enforced dependence upon the fickle goddess 

 of weather in the rain-belt — would be lost. The farmer 

 must have the water just when he needs it, not just when it 

 happens to come. The figures for average annual depth 

 show the maximum which can be relied upon with certainty 

 at any given date. "While it might sometimes be greater, 

 there is no rational assurance of it. 



The seasonal distribution of the rainfall is so far favorable 

 to a speedy use of stored waters, without serious loss by 

 evaporation, as to make the case somewhat better than ap- 

 pears in the table. But over against this is the neutralizing 

 consideration that the greater rainfall of spring and summer 

 is more fully absorbed than the lighter precipitation of win- 

 ter upon frozen ground. Melting snows yield a greater run- 

 off than summer rains. This increases the average period 

 of storage before use, and correspondingly diminishes the 

 chances of success. 



These changes are still too slender to be at all reliable if 

 the rainfall is fifteen inches. Indeed, it is not until we come 

 to the column headed " 18 inches" that we find any encour- 

 agement. One result at the bottom of that column looks 

 hopeful, but that calls for a catchment surface one hundred 

 times as large as the reservoir — a condition which, when 

 coupled with the further limitation of enough good irrigable 

 land under the reservoir, not one farm in a hundred can 

 fulfil. 



The promising figures are twice as numerous in the next 

 column, and three times as numerous in the last. But even 

 with two feet of rainfall the chances of failure and success 

 are about even. The ratio must be at least 75:1, or a mean 

 between the lowest and highest in the table. 



For areas having a greater rainfall than two feet, where 

 the impounded waters might be useful for other purposes, 

 but would hardly be needed for irrigation, the possibilities 

 of storage may be easily discovered by extending the table. 



Water storage upon the high mesas of the treeless belt 

 is, if not wholly a delusion, at least somewhat delusive. 

 More hopeful is the expedient of deep tillage, which is also 

 a sort of storage. Hidden from sun and winds in the loose 

 soil and sub-soil, the moisture will thus be preserved at the 

 very spot where it is needed to sustain vegetation. 



L. E. Hicks. 



Dr. Baillon's " Dictionnaire de Botanique," the publication of 

 which was commenced in 1869, is now completed. 



