36 BULLETIN 194, U. S. DEPARTMENT OF AGRICULTURE. 
EARTH CHANNELS. 
No. 119, Expt. S-2, Farmers' (Tristate) Canal, Nebraska. This test and also No. 
120 are on long, straight reaches of a large canal, constructed in Brule clay. An 
engineer connected with the original construction states that in the original design 
the value of n was estimated as 0.025, but, as shown in Plate IX, figure 2, the canal is 
running to but partial capacity, and the mean velocity is almost sufficient to scour 
the material. It had one riffle midway of its length, due to old bridge approaches 
jutting into the canal. The values of n in Nos. 118, 119, and 120 are comparable, as 
the Interstate Canal is in the neighborhood of the Tristate. A fringe of grass extends 
along the edge but retards only a very small part of the flow. The bottom is extremely 
even, smooth, and hard, and with the addition of a coating of sediment from the murky 
waters of the North Platte River, appears to be very efficient. Gentle curves adjoin 
both upper and lower ends of the reach. For further notes see No. 120. Coefficient 
n=0.0130. 
No. 120, Expt. S-l, Farmers' (Tristate) Canal, Nebraska. This reach (PI. IX, 
figs. 2 and 3) was perfectly clean cut throughout its length, and in the opinion of the 
writer gives the better value of n than reach in No. 119. In both tests the fall is 
slight, and the mean value of the results of several tests with the level was accepted. 
The reach is a tangent between two gentle curves. The bottom was as described in 
No. 119. Coefficient n=0.0164. 
No. 126, Expt. S-83, Maricopa Canal, Salt River project, United States Reclama- 
tion Service. This reach was part of a long stretch of canal with a clean, sandy bot- 
tom and a slight fringe of grass along the edge, but the influence of the latter was 
practicably negligible (PI. X, fig. 1). The slope is very gentle, so that a very slight 
error in the levels would appreciably affect the value of n. The measurement with 
current meter was made above the section tested, where the velocity was greater. 
Coefficient n =0.0166. 
No. 127, Expt. S-3, Winter Creek Ditch, Nebraska. This test was made on a long, 
straight reach of ditch, with very clean, hard bottom, in a cemented material. A 
fringe of grass bordered both edges. A stiff wind was blowing directly down stream 
during the test, and a value of n of 0.0180 is probably better for this canal under normal 
conditions than that actually found by the measurement made. Coefficient n= 
0.0170. 
No. 130, Expt. H-7, Billings Land & Irrigation Co.'s Canal, Montana. This test 
was made on a straight reach of canal excavated in clay loam soil. The little grass 
at the edges is of slight consequence, as the bottom is slick, though roughed by cutting 
in places. The mean velocity, 2.45 feet per second, was about the limit, as cutting 
was taking place where the bottom was not protected with a deposit of gravel. A 
downstream wind probably reduces the value of n from about 0.018, making it quite 
comparable with No. 127 above. Coefficient n=0.0174. 
No. 132, Expt. H-5b, Cove Ditch, Montana. This test is on the same ditch, with 
same channel conditions as No. 133. This reach is all curve, the first 300 feet being 
on a 30° curve, while the ensuing 300 feet is on a reverse 30° curve. While the slopes 
and cross sections are quite different, the value of n for this test and for No. 133 agree 
quite closely. Coefficient n=0. 0180. 
No. 133, Expt. H-5a, Cove Ditch, Montana. This reach is part tangent (station 
to station 2) and part on a 20° curve (stations 2 to 4). The ditch is 6 years old. 
Originally excavated in sandy loam soil, the bottom is now covered with a silt deposit. 
A fringe of grass retards the velocity at the edge, but not the main flow. Coefficient 
71=0.0186. 
No. 134, Expt. H-19, Billings Land & Irrigation Go's, main canal, Montana. This 
reach of canal follows a gentle hillside contour, although practically straight. There 
is a little sand and fine gravel scattered over a general bottom of clean soil. The 
