48 The. Colorado Experiment Station 
should not be used in very narrow or very shallow channels. 
4. The 0.6 method assumes the average velocity from top to 
bottom to be represented by the velocity at six-tenths of the depth 
from the surface. Actual experience has shown that this is not 
true in all cases, for the coefficient to be applied to velocities at this 
depth ranges in deep channels from 0.94 to 1.04, and in shallow 
channels from 0.97 to 1.04*. In some cases the mean velocity does 
occur at exactly 0.6 of the depth, but the error may be at least as 
great as six percent. 
With either the 0.2 and 0.8, or the 0.6 method, approximate 
results will probably be obtained, the degree of accuracy varying 
with the cross-sectional factor and the roughness of the material. 
However, for large streams, and a swift current, they are to be 
preferred to the multiple point or vertical integration methods. 
Obviously many sections of canals are unsuited to current 
meter work and quite unreliable results would be obtained by what¬ 
ever method used. Very often when the water is not clear, what 
appears to the eye to be a good place for taking a rating has 
some hidden obstruction which exerts an influence on the lower 
portion of the section. This would be detected by the multiple 
point method, somewhat compensated for in the vertical integra¬ 
tion method, but might cause serious error by the 0.2 and 0.8, or 
0.6 method. 
Due regard should always be paid to securing a favorable sec¬ 
tion for rating. It should ■ be free from debris, moss and other 
aquatic growth, and the channel should be straight, or fairly so, 
both above and below for a distance varying with the quantity of 
water and general conditions. For accuracy of measurement it is 
essential that the water flow in as nearly straight lines as possible. 
Just as it is necessary to choose a suitable section for rating a 
channel, so is it necessary to suit the method to the conditions which 
obtain therein. In an ideal section each of the four methods will 
give reasonably accurate results, but a perfect section is rarely en¬ 
countered. The value of a current meter measurement often de¬ 
pends as much upon the hydrographer’s judgment in choosing the 
method and place of rating, as upon his kill in manipulating the 
instrument. The novice should use the more simple methods. 
Acknowledgment.—To the managers and engineers of the 
canal systems upon which these experiments were conducted, 
acknowledgments are here made for the assistance rendered in 
permitting access to original records, and otherwise facilitating 
the work. 
*“l\iver Discharge.” Iloyt & Grover, p. 50. 
