Amount of Water Used in Irrigation 209 



stated canal is actually applied is so uncertain, and 

 the amount of water lost by seepage and evaporation 

 from the canal and its distributaries before the land to 

 which it is nominally applied is reached, is so variable 

 and indeterminate that the best which can be said 

 regarding most available data is that they should be 

 looked upon as only rough approximations. Further 

 than this, it must be constantly borne in mind, when 

 dealing with the problem of how much water is re- 

 quired for irrigation, with all the variations of weather, 

 climate, crops, soils and degrees of skill in applying 

 water which exist, that were sufficiently exact data at 

 hand covering a wide range of conditions, it would 

 still be impossible to combine them into averages not 

 requiring wide marginal allowances to be made when 

 specific application is desired. But, notwithstanding 

 all this, general statements may be helpful if only 

 they are rightly considered. 



Referring, first, to Italy,* where irrigation has long 

 been systematically practiced, it is generally calculated 

 that in Piedmont one cubic foot of water per second 

 will serve satisfactorily 55 acres of land ; but on ac- 

 count of loss by evaporation and seepage, this is 

 reduced to 51.4 acres, this providing sufficient for 

 4.63 inches of water every 10 days during the irri- 

 gation season. 



Under the canal of Ivrea, where a large amount 

 of rice is grown, which is given more water than ordi- 

 nary crops, one second -foot serves but 42.75 acres, or 

 at the rate of 5.668 inches every 10 days ; and under 



*Baird Smith, Italian Irrigation, Vol. I. 



N 



