4 
The Coeorado Experiment Station 
the Price rod meter. This instrument was rated at Berkeley, Cali¬ 
fornia, just prior to beginning the experiments, and upon their 
completion was rated by the Bureau of Standards, Washington, D. 
C. The former ratings were used for the season of 1912 and the 
latter for 1913. A high grade wye level, target leveling rod read¬ 
ing to thousandths of a foot, standard steel tape, and graduated 
hard wood stick with tapered edge, were used. 
In choosing the waterways upon which to conduct the experi¬ 
ments, particular care was exercised in selecting those typical of 
the class or group being investigated. Those possessing the 
slightest changes in the value of the hydraulic elements throughout 
the length of the section were best adapted for the purpose of ex¬ 
periment. Sufficient length was selected to accurately obtain the 
slope of the water surface. Other conditions being uniform, the 
length required varied inversely as the fall of the water surface. 
The distance required to develop the true fall was not so long, 
however, as to introduce a broken profile of the water surface 
gradient. Wherever possible, sections were selected on tangents, to 
avoid effect of curvature. 
As repeated experiments were made upon some of the con¬ 
duits as a check, or to determine any change in the value of the co¬ 
efficient with the variation in discharge, permanent bench marks 
were established at convenient points along the section, from which 
future elevations of the water surface could be determined, or the 
wetted areas could again be cross-sectioned. 
In all of these tests particular care was exercised in obtaining 
the correct discharge by current meter measurement, but at the 
same time the fact was not overlooked that measurement of the 
cross-sectional elements at the several stations was of equal import¬ 
ance in obtaining the coefficient of friction for any given water¬ 
way. In most cases the discharge was likely to fluctuate slightly 
while the test,was being made, hence the importance of a simul¬ 
taneous determination of the slope, cross-sectional elements and dis¬ 
charge, or where this was impossible, as but two men were in the 
field, slope and cross-sectional measurements were taken imme¬ 
diately after the current meter measurement near the head of the 
section, proceeding from the upper to the lower end. Hence, under 
such conditions it seemed inadvisable to soend time in making 
more than two determinations of the discharge, esoeciallv on a 
large channel where much time was consumed in doing so. The 
mtegration method was preferred for measurements in shallow 
streams. Independent meter measurements checked with an error 
seldom greater than t t; percent, and usually close to 0.; nercent. 
