Frictional Resistance in Artificial Waterways 13 
water entering and leaving the section. Usually the condition was 
that of eddies and whirls and the banking of the water on the out¬ 
side. Hence usually as much as 30 feet above or below the turn 
was allowed for the water to regain its flow in an undisturbed 
condition. 
Several experiments were made at heads varying from 60 to 
230 second feet. The unevenness of the water surface necessi¬ 
tated the taking of two readings from the bench marks to the 
water at its lowest and highest stages, and the mean of these was 
used in the computations for the slope. Table 5 shows the results 
of 75 experiments on 22 sections of this canal. 
Of the 22 sections experimented upon, 17 show a higher value 
of (n) for the highest than for the lowest discharges. Of these 
iv. 14 have a greater slope of water surface for the highest than 
for the lowest heads. Only eleven, however, indicate a gradual 
increase of the value of (11) with that of the hydraulic radius. At 
the bottom of Table 5 is given the mean values of the elements ob¬ 
tained for each discharge. As will be noted there is no apparent 
change in the value of (n) with that of the hydraulic radius. 
However, to make the change apparent, of the ten mean values ob¬ 
tained, arranged in order of corresponding discharge from least to 
greatest, an average of 1-4, 4-7 and 7-10 indicate (n) to* be .0117, 
.0121, and .0124, respectively. This set of experiments indicates 
the possibility of a variation in the value of (n) with the change in 
hydraulic radius for a rectangular flume of this type at least, and 
that this value increases with the discharge. These data are not 
positive proof of such variation, but they are presented here merely 
as a suggestion. Even if this change in the value of (11) does 
occur, it is probably not of great practical concern. 
An average value for the coefficient of roughness of this 
flume is 0.0122. 
There follows a brief description of other timber flumes in¬ 
vestigated, and Table 6 gives the hydraulic elements as determined. 
The results are of value merely in illustrating the condition of 
some flumes after a period of use. Many of these may not orig¬ 
inally have been built upon the proper grade to> have the same carry¬ 
ing capacity as the channel leading to or away from them, or if 
formerly well designed and constructed, subsequent conditions 
may have caused sagging or irregular alignment. 
REDLANDS MESA FLUME, GRAND JUNCTION. The 
flume is in excellent condition as regards the character of the 
material of which it is constructed, the lining being tongue and 
