THE FLOW OF WATER IN IRRIGATION CHANNELS. 45 
irregular; from stations 4 to 6, more uniform in width but rough on rock side; from 
stations 6 to 7, rock wall is rough and width variable; from stations 7 to 8, rock wall 
quite smooth, bottom clean or little gravel, the width uniform. Coefficient n=0.0228. 
No. 264, Expt. S-17a, Logan, Hyde Park and Smithfield Canal, Utah. This test 
made on a reach 337 feet long between an earth section and the reach in No. 266 
below (PL XX, fig. 1). The excavation is on a steep hillside. The upper bank is 
mostly of willow roots, while the lower bank is a well-made concrete wall. The 
bottom is covered with coarse gravel. This reach is nearly straight with bends at both 
ends. This is classed as B rating by reason of these bends. Coefficient n=0.0256. 
No. 265, Expt. H-26, Hedge Canal, Montana. This reach is excavated in rock cut 
with concrete floor and a rubble masonry lower wall, faced with 3 inches of concrete, 
deposited against wood forms. The bottom is mostly covered with sand and rav- 
ellings of small rock. The upper bank is rough rock excavated quite true to cross 
section. The alignment is practically straight except for one sharp curve between 
stations 0+80 and 1+50. The value of n is higher than in No. 262, which has a 
smoother section. Coefficient n=0.0269. 
No. 266, Expt. S-17b, Logan, Hyde Park and Smithfield Canal, Utah. Just below 
the reach described in No. 264 the canal enters the section covered in this test. The 
same concrete wall formed the lower bank, and the bottom was about the same, but 
the upper bank is a rough vertical rock cut. The difference in the value of n is about 
what is to be expected. Coefficient n=0.0278. 
MISCELLANEOUS SECTIONS. 
Nos. 267 and 268, Expts. S-74 and S-73, lower canal, Riverside Water Co., Cali- 
fornia. These tests made on a straight reach of canal in a sandy soil with a shifting 
sand bottom and a wood lining on the lower side (PL XX, fig. 2). The canal in test 
No. 267 is in the shade of a dense row of trees and is free from moss accumulations. 
Coefficient n=0.0249. 
The canal in test No. 268 is in the sun and moss has accumulated on the wood 
lining. Coefficient n=0.0291. 
In both tests the water was retarded by a rank growth of grass for about 1 foot from 
the bank opposite the wood lining. The difference in the values of n is directly due 
to the moss which grows in sunlight and does not in shade. 
No. 269, Expt. S-17c, Logan, Hyde Park and Smithfield Canal, Utah. This reach 
is fairly straight, excavated in rough rock. The bottom is strewn with coarse gravel 
Coefficient n=0.0298. 
THE USE OF VALUES OF n. 
The engineer is required to exercise his judgment as to the value 
of n in two general ways : 
1. The value to be substituted in the formula in the design of a 
channel or in the changing of a channel from one containing material 
to another. 
2. The value that is to be used when the engineer is called upon 
to determine the maximum carrying capacity of a canal already 
constructed but at the time of inspection carrying no water or but a 
small portion of its capacity. 
In the design of channels it has been customary to allow little or 
no factor of safety. In the opinion of the writer it is as necessary to 
allow for a slight overload in a canal as in any other conservative 
construction. This should be accomplished by choosing as nearly 
