12 
Colorado Experiment Station 
mic scale which presents the data in such a form as to be readily 
understood and applied. The use of the diagram can best be 
illustrated by an example. Assume a 2-foot Venturi flume with an 
upper gage height of 1.26 feet and a difference in head of 0.21 
feet. In Figure 15, find the difference in head, 0.21, in the column 
on the left of the diagram, midway between the two horizontal 
lines, follow across the diagram horizontally until the 1.20 upper 
gage height curve is reached ; estimate the fractional part and then 
read the discharge of 8.0 cubic feet per second at the bottom of the 
diagram vertically beneath this point. In determining the discharge, 
care should be observed in noting the change in the scale dimension 
in the different parts of the diagram. 
It will be observed that for each upper head there is a limiting 
point, beyond which the curve does not extend. This limit is rep- 
resented in the diagram by the curve designated as “free flow.” 
This results from the fundamental theory of flow through contrac- 
tions in channels. For every upper head as the difference in head 
increases, a point is finally reached beyond which the difference in 
head can no longer be increased. This is the point at which the 
maximum discharge occurs. This condition has been observed 
experimentally and it was found that it occurred when the throat 
gage registered from 50 to 60 per cent of the upper gage. This 
percentage varies with the size of flume and the upper head, increas- 
ing .as the size of the flume increases and decreasing as the upper 
head increases. After the free flow condition has been reached, 
the depth of water in the diverging section may continue to drop 
without affecting the discharge or the gage heights. 
Some idea as to the accuracy of the rectangular Venturi flume 
may be gained by inspecting the diagram. Figure 2, which is based 
upon comparisons of actual discharge with the quantity determined 
by computation, the basis of the computed values being the empirical 
expression developed from the experimental data. It will be noted 
that the per cent of deviation of the observed discharge from the 
computed quantity is largely confined to errors of less than 5 per 
cent; however, it is apparent that an occasional point is consider- 
ably more in error. This diagram is based upon 453' individual 
observations covering the calibration of the 1-, 2-, 3-, and 5- 
foot rectangular flumes. The distribution of observations is quite 
uniform over the various sizes of flumes, and the compilation of the 
data for the comparison was made without selection or elimination. 
