THE FLOW OF WATER IN IRRIGATION CHANNELS. 35 
pi gravel, and the sides are .slightly waterworn and have a slick deposit of silt. A 
sharp curve occurs about 50 feet above station 0, but the reach tested is practically 
straight. Coefficient ?i=0.0167. 
No. 93, Expt. G-3, Roller flume, Louisiana. This flume was about 14 years old, 
with the exception of a portion of the reach tested, about 75 feet long, which had been 
Repaired with new cypress lumber. The original lumber was unsurfaced, but at 
present is covered with a slimy growth. The lumber was parallel to the axis of the 
canal, but the surface was slightly irregular. Coefficient n=0.0191. 
No. 94, Expt, S-76, Riverside Water Co.'s Lower Canal near Riverside, Cal. The 
side boards of this flume were of vertical tongued and grooved lumber, making a very 
rough interior. The bottom was painted with a tar or asphalt product, and the posts 
were in such condition that the flume sagged and was out of line in numerous places. 
The value of n as found for this flume may be taken as that of a flume in exceedingly 
poor condition but without appreciable deposits of sand or gravel. It is understood 
the company expects to replace this flume in the near future. The reach tested is 
straight. Water enters from a cement-lined section about 50 feet above station 
and leaves the flume into an earth ditch about 50 feet below station 7 plus 46. Co- 
efficient ri=0.0196. 
No. 96, Expt. S-65, Fullerton Ditch of the Anaheim-Union Water Co., California. 
This flume has the bottom so covered with soft sand that it classes more nearly with 
an earth section. The lumber was unsurfaced, and a growth of moss retards velocity 
for about 0.1 foot from each side. The reach tested was straight, but there is a sharp 
curve about 50 feet above station where the earth canal joins the flume section. 
About 15 feet below station 6 plus 72 the water enters the earth section again. Coeffi- 
cient n=0.0202. 
MASONRY-LINED CHANNELS. 
No. 113, Expt. S-29, Jacobs Ditch, Boise, Idaho. This test is on a straight stretch 
of ditch. As shown in Plate VIII, figure 2, the sides are of first-class rubble masonry 
with all cracks smoothly plastered with cement. The bottom is smooth cement 
lining laid like a good grade of sidewalk. About 50 feet below the lower end of the 
reach the ditch passed through a vertical trash grating, where there was a loss of head 
of about 0.05 foot. This grating was kept clean of debris during the test. Coeffi- 
cient n=0.0130.« 
No. 115, Expt. S-27. This test was made on the same ditch as No. 113 above, but 
this reach is lined on both sides and bottom with dry laid, unchinked rubble, as 
shown in Plate VIII, figure 3. The bottom is quite irregular, with scattered loose 
cobbles. A clean grating came a short distance below the reach, as in No. 113. Coeffi- 
cient n=0. 0235. 
No. 116, Expt. S-28. This test also is on the same ditch as No. 113 above. This 
reach, one city squaie long, came between the reaches in No. 113 and No. 115 above, 
It appeared to have been originally like No. 113, except that the bottom was not 
lined. There were several cobbles throughout the bottom of this reach, and this 
probably accounts for the fact that the value of n is slightly higher than for No. 115, 
while to all appearances it should have been slightly lower. A grating similar to the 
ones noted above came below the lower end of the reach. Coefficient n=0.0250. 
No. 117, Expt. S^t5, Orr Ditch, Reno, Nev. As shown in Plate IX, figure 1, the 
sides are quite smoothly built rubble, with most of the cracks well plastered, but 
the bottom is covered with shifting sand and loose cobbles so that, if lined, the lining 
is completely concealed. The lined section ends shortly below the lower end of this 
reach, passing into an earth channel. This test exemplifies the need of keeping 
sand and gravel from a lined section if the low value of n that might be expected is to 
be realized. Coefficient ri=0.0298. 
