92 



Tiikiiij;- tlic results as «>ivcii in the prcc(Hliii<4- table, the average 

 weekly evaporation from May 1 to September 30, of a four-year period, 

 amounts to a trlHe more than 0.9 inch per week. This tlo-ure has])eeii 

 taken to compute the loss of water by evaporation from typical canals. 

 The following table gives the results of such computations for ten 

 canals: 



Amount of mthr evaporated from liiplcal amah in Montana. 



It has been considered desirable to give a reasonably close estimate 

 of the loss b}^ evaporation from typical canals in order to correct a 

 false idea which is prevalent regarding it. It is not unusual for 

 irrigators to hold that the greater i)art of the losses which occur 

 between the intake and their tields is due to evaporation instead of to 

 seepage. Now, the computations siiow that this loss, even in large 

 canals, is not worth reckoning with. A canal, for example, 12 feet 

 w^ide on top and carrying 1,600 miner's inches, loses less than a third 

 of a miner's inch per mile, while one 40 feet wide and carrying 12,000 

 miner's inches loses a trifle more than 1 miner's inch. 



SEEPAGE. 



In the measurements that have ))een made to determine the losses in 

 tlie transportation of water for irrigation [)urposes few attempts have 

 been made to separate the volumes lost in difl'erent ways. It has been 

 previously shown that the amount evaporated is small in comparison 

 wnth the total loss. In canals that are skillfully maintained and 

 operated the loss due to leaks is likewise small. It is apparent, there- 

 fore, that by far the greatest loss arises from percolation or seepage. 

 The results given in the tal)le which follows include loss from all 

 sources, the greater part of which is, as just stated, from seepage. 

 The canals tested were usually representative of the larger class in the 

 district. After one had been selected the necessary arrangements 

 were madewitii the superintendent before the date fixed for the begin- 

 ning of the test. A gauge rod of some k ind was in.serted near the intake, 

 and the discharge ascertained ])v means of a current meter. The person 



