14 
Colorado Experiment Station 
smooth trowel finish. The most important element in the con- 
struction of these flumes is the dimension of the throat section. 
If concrete, it is recommended that the forms be set so that the 
distance between the two vertical faces will be from ^ to 1 inch 
over size. After the forms have been removed, a finish coat of the 
required thickness should be put on to bring the width of throat to 
the correct dimension. The finished length of the throat measured 
along the axis of the flume is to be one foot, and if a plaster coat 
is added after forms are removed it will be necessary to make the 
cast face from 2 to 3 inches longer. This method of constructing 
the throat will insure a greater degree of accuracy in obtaining the 
true size. A trowel finish on the sides or walls of the flume is not 
absolutely necessary, but in most cases will add to the appearance 
and general utility of the structure. For wooden structures, see 
description of method given for V-notch flumes on page 23. 
The observation wells should be cast in place and made a part 
of the structure. The tubes leading to these wells should be one- 
inch pipe, a single tube being sufficient, and should be placed two 
or three inches above the floor. These inlet tubes should be set 
flush with the inside surface of the flume, and it is important that 
they be normal or at right angles to this surface. The depth of 
the well should be sufficient to permit sand or silt to deposit and 
only require cleaning at long intervals. The well should extend 
somewhat above the general ground line, thus avoiding storm water 
washing trash into the well. Covers for the wells are recommended 
as a protection against foreign matter being deposited within them. 
At least one of the inside faces of the well should be made vertical 
and be provided with stud bolts for the purpose of anchoring a 
plank to support a hook gage or other measuring device. 
Experience indicates that a graduated scale fastened to the 
inside face of the Venturi flume is not sufficiently accurate to 
determine the correct elevation of the water surface. The size of 
the wells would need to be comparatively large in cross section to 
permit of accurate reading of the scale if placed in the still wells. 
For the determination of the depth of water in the still wells above 
the floor of the flume, a depth scale, Fig*ure 3, has been proposed 
which is simple in construction and operation, and when set will 
give the upper gage reading and the difference in head. The design 
of this scale is such as to eliminate the possibility of erroneous 
