THE FLOW OF WATER IN IRRIGATION CHANNELS. 43 
gritty sand. As the canal was only about one-half full, no grass touched the water 
at the time of the test. Coefficient n=0.0373. 
No, 236, Expt. S— 12, Cochrane Ditch, Nevada. This ditch is constructed in gravelly 
■oi] with many cobbles. The test was made on a reach having one bend (see PL 
XVII, fig. 2). The bottom had many loose cobbles scattered on an otherwise hard 
gravel bed. The banks throughout nearly all of the reach tested were overhung with 
densely grassed sod. In addition about 20 per cent of the water section was occu- 
pied with moss. Coefficient n=0.0379. 
No. 237, Expt. S-25, Perrault Canal, Capital Water Co., Boise, Idaho. This ditch 
(PL XVII, fig. 3) constructed in gravelly loam soil, has a very hard cemented bottom. 
A dense growth of grass so kills the velocity for a distance of about 1 foot from each 
bank that the value of n is very much greater than would be the case if the canal were 
Jkept free from this growth. Coefficient n=0.0381. 
No. 239, Expt. S-46, Orr Ditch, in Reno, Nev. The reach tested passed in a horse- 
shoe curve around the small lake on the university campus. It was originally exca- 
vated in loamy soil with a little gravel. At time of test the bed was clean scoured, 
but near the sides a man wading sank about 4 inches in soft mud. Dense grass and 
willows retard the velocity at both banks, while the water section is very irregular 
throughout the reach. Coefficient n=0.0397. 
No. 240, Expt. S-21, a small ditch in Twin Falls, Idaho. This is a small ditch with 
grass arching across in many places. Originally constructed in hardpan, the reach is 
irregular with scattered debris such as is so often found in town ditches. Coefficient 
n=0.0399. 
No. 241, Expt. S-52, Capurro Ditch near Reno, Nev. As shown on Plate XVIII, 
figure 1 , this is a small ditch thickly fringed with grass and with scattered cobbles in 
the bottom. The banks are quite vertical, densely rooted, and very irregular. The 
bottom is covered with about 0.1 foot of soft mud through which the scattered cobbles 
project. Coefficient n=0.0403. 
No. 243, Expt. S^, New Rutner Ditch, Nebraska. This ditch follows a gentle 
contour line down a creek bottom. At time of test the ditch had a very hard bottom 
of medium-fine gravel, well packed, but a dense growth of grass killed the velocity 
for about one-half foot from each bank, and scattered patches of moss retarded that 
in the middle. The banks of the ditch are very irregular. Coefficient n=0.0436. 
No. 244, Expt. S-44, Sullivan and Kelly Ditch, Nevada. This test was made on a 
ditch excavated in a gravel and cobble hillside. At time of test the bottom was 
hard-packed gravel in the center with a slight deposit of soft mud at the sides. Scat- 
tered cobbles and a dense growth of grass retarded the velocity of the water. The 
reach follows a gently curving contour line with one right-angled bend near the lower 
end of the reach tested. Coefficient n =0.0436. 
No. 245, Expt. G-2, Roller Canal, Louisiana. This test was made on a straight 
reach of canal. The vegetation shown in Plate XVIII, figure 2, extended for about 
5 feet from each shore. Coefficient n=0.0461. 
No. 249, Expt. G-5, a small ditch in Louisiana. This ditch was chosen as repre- 
sentative of the small ditches in the rice country. Grass extended from one bank to 
the other across the bed of the ditch, occasionally growing to the water surface from 
the bottom of the ditch. The grass forms a dense mat in the bottom. The grass had 
been cut with a scythe about one week before the test. Coefficient n =0.0544. 
COBBLE-BOTTOM DITCHES. 
No. 251, Expt. H-31, Bitter Root Valley Irrigation Co.'s Canal, Montana. This test 
was made on a nearly straight reach 600 feet long, excavated in very gravelly ground 
with bowlders up to 2 cubic feet in size. The first third of the distance fairly smooth 
on the bottom with no stones larger than two-fist size; upper slope cobbly and with 
