THE FLOW OF WATER IN IRRIGATION CHANNELS. 33 
ated to capacity. The flume has been in use nine years, and the inside relined with 
1-inch lumber meanwhile. Mr. Harding suggested in designing a flume where such 
waves are to be expected, that a liberal freeboard be allowed for or else that the crests 
of the waves be considered the water surface. He has solved n for both assumptions 
and finds n runs about 0.004 higher where crests are considered the water surface. 
This difference was quite close throughout the four tests. 
No. 70, Expt. 11-39, Bitter Root Valley Irrigation Co.'s flume No. 2, Montana. This 
flume is built of 2^-inch finished tongued and grooved siding. The side boards are 
12 inches wide, with the cracks battened with £ by 2-inch strips. On each side of 
the cracks in the bottom is laid a \ by 1-inch strip, leaving a groove about one-half 
inch wide over the crack. This groove is then poured with an asphaltic filler. The 
reach chosen follows a contour with the following alignment: Station about 100 
feet below upper end of flume; thence 100 feet of about 10° curve right; thence 200 
feet tangent; thence 100 feet of about 5° curve right; thence 100 feet tangent; thence 
50 feet of about 10° curve left; thence 100 feet tangent to a small bend; thence 50 
feet tangent, to end of reach which is about 150 feet above outlet to flume. This 
reach is very similar to that of No. 71. Coefficient ri=0.0125. 
No. 71, Expt. H-37, Bitter Root Valley Irrigation Co.'s flume No. 1, Montana. The 
reach tested is at the lower end of a flume about 1 mile long. The sides were formed 
of 2£ by 12-inch finished, tongued and grooved lumber with cracks battened with \ by 
1^-inch strips. There are no battens on the bottom. There was a thin coating of 
sand and gravel in places. The alignment is about as follows: Station to 1 is about 
20° curve right; thence 200 feet tangent; thence 300 feet of about 10° curve left; 
thence 200 feet tangent; thence 50 feet of about 20° curve left; thence 150 feet tan- 
gent. The value of n should be compared with those found in Nos. 58, 70, 73. It is 
higher than for No. 58, because the latter is on a clean tangent. It is lower than for 
No. 73, which is on a very crooked reach. Coefficient n =0.0127. 
No. 72, Expt. H-28, Hedge Canal, Bitter Root Stock Farm, Montana. This test is 
on a new reach of flume built of 3 by 12-inch finished tongued and grooved lumber. 
The canal is rough both above and below the flume, which is too short for the best 
grade of results. Coefficient n=0.0129. 
No. 73, Expt. H-32, flume No. 7 of the Bitter Root Valley Irrigation Co.'s main 
canal, Montana. This is a veiy crooked flume of the same construction as in No. 58. 
There was no moss or gravel. The flume follows mountain contours, having sharp 
curves right and left, with short tangents between. The higher value of n found in 
this test is to be expected and gives comparative results between straight and very 
crooked flumes, when compared with the three other tests on flumes of this canal. 
Coefficient n=0.0135. 
No. 74, Expt. H-ll, lateral No. 4, Billings Land and Irrigation Co., Montana. This 
is a small flume 2 feet wide and 2 feet deep. There are no battens and the butt joints 
are calked with oakum. The inside of the flume is weathered, but appears to have 
been originally of surfaced material. The flume is generally straight, there being 
wavy places which slightly distort the true alignment. The ends of the reach are 
far enough from entrance and exit to the flume to be uninfluenced. Coefficient 
n=0.0138. 
No. 75, Expt. S-16, Telluride flume, Utah Power & Light Co., near Logan, Utah. 
This test is on a straight reach of flume with an angle of about 20° left, located 35 
feet below the end of the reach. See Plate VI, figure 3. The sides are of 1-inch 
Oregon pine with all cracks battened with f by 3-inch strips, the vertical battens, 
about 16 feet apart, being beveled on the upstream edge to reduce friction. The 
bottom battens appear to be about 6 inches wide, but examination was difficult. All 
surfaces of wood in contact with water appear to have been surfaced, but are so weath- 
ered and waterworn that this point is indeterminate. Coefficient n=0.0141. 
79256°— Bull. 194—15 3 
