Nov. 8, i9ii 
Use of Current Meters in Irrigation Canals 
223 
secured and the cross section of the canal in order to give the discharge. 
In such measurements there are two principal sources of error: (1) The 
cross-sectional area is difficult to obtain except in flumes or lined canals 
of uniform cross sections and (2) mistakes may be made in choosing a 
coefficient to be used in reducing the maximum surface velocities as 
obtained from the floats to the mean for the whole canal. 
The following results relate to the proper coefficient to be used to 
reduce surface-float velocities to the mean velocity for the whole cross 
section. The average errors discussed are those arising from the de¬ 
terminations of float velocities and the choice of coefficients and do not 
include errors in determining the canal cross sections. For the other pur¬ 
poses of these experiments the areas of the canal sections were carefully 
determined. In the usual field use of float methods there may be a con¬ 
siderable error introduced due to errors in the approximate determinations 
of canal cross sections of variable dimensions, which would give larger 
probable errors for the discharge than would result from the probable 
error due to the choice of the coefficient to use with the velocity of the 
float alone. 
Various formulas have been derived for the relation of the surface 
velocity to the mean velocity. These have been derived both for the 
relation of the surface velocity to the mean velocity in any single vertical 
in the section and for the relation of the maximum surface velocity to 
the mean velocity of the whole channel. Ganguillet and Kutter 1 give 
a formula, deduced by Bazin, in which the ratio of the maximum to 
mean velocities in- a channel are made to vary with 
[RS 
V v 2 
As this term is equal to the C in Chezy's formula, a table is given for 
the value of the ratio for different values oTC. In this formula Kutter 
substitutes the values of n and R from his general formula and gives a 
table for the values of the ratio of mean to maximum velocity, depending 
on R and n. The formula derived by Bazin, which forms the basis of 
this table, was based on 61 series of gagings. 
In the canal experiments discussed in this paper in which float measure¬ 
ments were made several small floats would be started simultaneously 
at scattered points in the portion of the channel having the highest 
velocities. The time of the most rapid float was used to compute the 
maximum surface velocity. This gives lower coefficients than would be 
obtained by the use of the average of all floats. Small floats such as 
twigs or chips were used which would have both a small submergence 
and a small exposed surface above the water. It was found that there 
was little difference in the velocities of the floats thrown into the main 
threads of the canals unless some became caught in noticeable side 
1 Ganguillet, E., and Kutter, W. R. Op. cit. 
