I 
Nov. 8,1915 Use of Current Meters in Irrigation Canals 221 
depths, which is also true of the other methods. This is to be expected, 
as the smaller velocities and depths usually occurred in canals of small 
discharge, where the general conditions for the use of the current meter 
are not so favorable. The accuracy does not appear to be affected by 
the character of the channel or value of n. 
There is some indication that the correction to be used with the single¬ 
point method should be greater than 5 per cent for low velocities and 
less for the higher ones. This tendency is not marked, however, and it 
is doubtful if it is sufficient in amount or that it is sufficiently proved 
by these results to warrant the use of different corrections; also the 
correction seems to vary with the depth in a similar way. 
The integration method seems to give the closest average results for 
velocities from 2 to 3 feet. It also appears to be more accurate for the 
greater depths. This latter result is to be expected. In the use of the 
integration method the velocity in from 0.2 to 0.3 foot in depth must 
be either missed entirely or imperfectly determined both at the bottom 
and at the water surface. The velocity at the bottom is lower than 
the average. Therefore the measurements in the remaining portions of 
the depth would give results above the actual average velocity. As the 
proportion of the depth for which velocities are undetermined is larger 
in the shallow canals, the proportionate error would be greater. 
Another method sometimes used is that known as the three-point 
method, in which the velocity is measured at 0.2, 0.6, and 0.8 of the 
depth. This is more usually computed by giving the velocity at 0.6 
depth equal weight with the mean of the 0.2 and 0.8 depth velocities. 
As Table I shows the single-point method to be less accurate than the 
two-point, there is no apparent advantage in the three-point method 
over the two-point. In sections where the two-point method gave 
results too low and the single-point too high, their combination might 
increase the accuracy over that secured by the two-point method alone. 
Where both were of the same sign, the use of the three-point method 
would give less accurate results than the two-point alone. The two- 
point and single-point methods gave results having opposing signs on 
less than one-third of the total number of experiments, so that the three- 
point would seem to have little advantage over the two-point method. 
To definitely determine the relative accuracy of the three-point 
method, the discharge of each experiment was computed, using both 
the method by which the velocity at 0.6 depth is averaged with the mean 
of the velocities at the 0.2 and 0.8 depths, and also the method by which 
the velocities at the three points are given equal weight. This latter 
method would seem to be the more logical, as it has been shown that the 
two-point, or 0.2 and 0.8 depth method, gives results more accurate than 
the 0.6 point alone, so that in the use of the three points it would be 
preferable to reduce the weight given to the velocity at 0.6 depth. 
