48 BULLETIN 194, U. S. DEPARTMENT OF AGRICULTURE. 
II. 71=0.013 for construction as in type I, but with curves as in the 
usual mountain canyon. Same construction and alignment as in 
type I, but with small amount of sand or debris in water. Construc- 
tion as in type III, but in very favorable alignment or for water that 
carries a small amount of fine silt that vail eventually form a slick 
coat. 
III. 7i=0.014 for Hnings made by good construction under favor- 
able conditions. The surface to be as left by smooth-jointed forms 
or to be roughly troweled. Joints to be good, but causing some re- 
tardation. Alignment about equal in curves and tangents, with no 
spirals between. The bed to be clean and sides free from rough 
deposits. In the opinion of the writer this is the value to use for 
most hnings on moderate-sized channels. 
IV. 7i=0.015 for construction as in type III, but with sharp curves 
and clean bottom or moderate curves and much debris on the bottom 
but clean-cut sides. 
V. 7i=0.016 for concrete as constructed by the average gang of 
laborers, using forms that leave prominent lines at the cracks, no 
finish coat being applied. Bed to have the usual small amount of 
rock fragments and patches of sand and gravel. Average amount of 
curvature. In climates where a rough deposit accumulates, as in 
southern California, a hning that originally had a value of n about 
0.013 quickly assumes about this type. For this reason it appears 
to the writer that labor and money spent in securing a very smooth 
surface is lost where the deposit accumulates on smooth or rough alike. 
VI. 7i=0.017 for roughly coated Hnings with uneven joints. This 
value also is applicable where rough deposits accumulate on the sides 
and conditions of alignment are poor. 
VII. 71 = 0.018 for very rough concrete with sharp curves and 
deposits of gravel and moss. A broken gradient, irregular cross 
section, and the like, contribute to such a high value of n. 
Where experiments show higher values of n than are given above 
for concrete linings, the conditions are such, as a rule, that the con- 
taining material has lost its identity as concrete or cement, and thick 
coatings of sand, accumulations of moss, or deposits of sand change 
the general classification of the channel. 
VALUE OF n FOR WOODEN FLUMES. 
While experimental flumes have been constructed that showed 
values of n in the neighborhood of 0.009, yet in making tests on 
flumes in commercial service values of n below 0.012 are so rare that 
it is not thought advisable to recommend any value less than this. 
Wood is subject to so many changes, due to the influences- of climate, 
wind, settling of earth, moss accumulations, warping, and so on, that 
even when well constructed the value of n becomes greater after a 
