Use of Current Meters in Irrigation Canals 



227 



discharge which would have been obtained had a less number of verticals 

 been measured. The different types of canal sections were grouped 

 into general classes. For each gaging, discharges using only every 

 other vertical measured were computed and also using only every fourth 

 vertical. Two computations of each gaging using the two sets of alternate 

 verticals were made, and also two sets for every fourth vertical. These 

 results were then compared with the discharge obtained by the use of 

 all the verticals measured, in order to determine the probable errors to be 

 expected when fewer verticals were used. The average number of 

 verticals observed in the experiments was 16; the number in the com- 

 parisons averages 8 and 4. In general field current-meter work, if only 

 8 or 4 verticals had been measured, the ones used might have been located 

 in the cross section differently from the arbitrary method used in this 

 computation, so that the selection of alternate verticals as used should 

 give errors larger rather than smaller than are to be expected. The 

 results of this comparison are given in Table V. 



Table V. — Effect on the accuracy of current-meter gagings of varying numbers of verticals 



. Table V gives both the average difference in percentage and the range 

 of variations in single gagings. Occasionally the use of a less number of 

 verticals may give a greater discharge than that obtained from a more 

 detailed gaging, owing to irregularities in the cross section or velocity. 

 Where an average of 4 verticals were used, less than 2 per cent of the 

 observations gave larger discharges than the use of all verticals, so that 

 the average difference is practically equal to the mean error. Where 8 

 verticals were used for all observations, one in each seven measurements 



