THE FLOW OF WATER IN IRRIGATION CHANNELS. 31 
found. This sand was about 0.03 foot deep. The clear water emerges quietly from 
the lower end of a siphon under a road about 100 feet above station 0. The canal turns 
at right angles, without curvature, about 100 feet below station 6. However, the 
depth at the various cross sections remains quite constant, showing that the water was 
not appreciably checked up by this turn. As shown in Plate III, figure 3, there was 
a very slight retarding effect due to grass and weeds dragging on the surface of the 
water near the edges of the channel. Coefficient n=0.0188. 
No. 44, Expt. S-67, Arroyo Ditch & Water Co.'s main canal, near Whittier, Cal. As 
shown in Plate IV, figure 1, this rough-finish concrete section has accumulated a 
deposit of rough mossy growth that greatly retards the velocity of the water. In a 
few places throughout the reach tested the lining was irregular and not in true align- 
ment, which also tended to increase the value of n. The reach was on tangent, with 
a sharp angle about 50 feet above station and a gentle curve about 100 feet below 
station 10. See discussion under Nos. 35, 40, and 42, above. Coefficient n=0.0188. 
Nos. 45 to 50, Expts. S-31, S-30, S-32, Central Oregon Irrigation Co.'s North Canal 
near Bend, Oreg. These experiments were conducted with varying discharges, on 
consecutive days, in identical reaches; (a) is on a tangent 240 feet long between a 
15° curve above and a 14° curve below; (6) embraces 157 feet of tangent, then 154 feet 
of 14° curve to the right, then 90 feet tangent, then 109 feet of 15° curve to the left, 
then the tangent that includes (a) 240 feet long, then 178 feet of 14° curve which passes 
into a very rough lava-rock cut about 200 feet below station 10 plus 26. 
This lining is a clean-scoured, very rough, and deeply pitted concrete made in a 
rough lava-rock cut. As shown on Plate IV, figure 2, the cross sectional form is even 
and the filaments of current are not disturbed except by the curves, but the inherent 
roughness of the lining accounts for the high values of n. The grade of the bottom of 
this canal was constructed 0.001 feet per foot. 
This lining was a 1 : 4 : 5 mixture, deposited behind shiplap forms against a hand- 
laid rock wall, filling the cavities in a rough rock cut. Expansion joints of \ by 4-inch 
lumber were placed on sides and bottom every 12 feet and left in the concrete. 
No. 51, Expt. S-38, main canal, South Orland project, United States Reclamation 
Service, California. As shown in Plate IV, figure 3, this lined section comes between 
two earth sections of the canal. The concrete is quite rough and pitted, with slight 
growths of moss, but not nearly sufficient to account for the high value of n found. 
The writer can only account for this value because of retarding influences due to the 
earth channel below the lined reach. Coefficient n=0.0211. 
No. 53, Expt. S-68, small ditch from pumping plant, California. Although con- 
structed with a smooth-finished cement wash, this ditch shows a high value of n 
because a dark, crinkly deposit has changed the condition of the walls. Vegetation 
on the banks dragged in the water and retarded velocity to a slight extent (PI. V, fig. 1.) 
This test is not given full weight because the ditch is too small to give a first-class 
current-meter measurement. The mean of three measurements was used. Coeffi- 
cient n=0.0220. 
No. 54, Expt. S-75, Riverside Water Co.'s Lower Canal, Riverside, Cal. This experi- 
ment gives a good example of a cement-wash lining in which under favorable con- 
ditions in southern California a friction factor of about 0.018 might be expected with- 
out eradicating the sand which appears to be ever present in the canals in this vicinity. 
If the sand were removed by the addition of numerous sand sumps and gates this 
factor would be reduced to 0.016 or thereabouts. At the time of making the tests on 
this canal, as shown in Plate V, figure 2, the lining had been broken in scattered 
spots, allowing vegetation to root and grow. In the bottom of the channel were 
scattered deposits of loose sand, covering possibly 10 per cent of the bottom area. In 
some of these deposits moss and water grasses flourished. A gentle curve about 100 
feet above station had little or no influence on the flow in the low velocities en- 
