Frictional Resistance in Artificial Waterways 23 
sistance offered by various kinds of material (gravel, rock, soil, or 
vegetation) covering the sides and bottom of the channel. 
Sections on tangents were selected for experiment, and these 
were made as long as possible, that an accurate determination of 
the true slope of the water surface might be made. At the same 
time due care was exercised in choosing sections in which values 
of the hydraulic elements were to all appearances fairly constant 
throughout the channel tested. In short sections of 600 feet or 
less, cross-sectional measurements of water surface elevations were 
determined at the ends and middle of the section, while for longer 
sections these measurements were taken regularly from 300 to 800 
feet apart, depending upon the uniformity of the section. In se¬ 
curing the elevations of the water surface two stakes were driven 
into the channel, so that their heads were below the water surface. 
Into the top of one stake was driven a nail whose head was flush 
with the water surface at the lowest pulsation, and similarly a nail 
was driven into the other stake to mark the highest level of pulsa¬ 
tion. Opposite each point at which measurements were to be made, 
permanent bench marks were established and the elevations of the 
cross-sectional profile from bank to bank were tied into these. To 
eliminate the error due to a slight fluctuation in the discharge dur¬ 
ing the test, the position of the water surface at the various sec¬ 
tions was marked simultaneously with the taking of the current 
meter measurements, or immediately thereafter. In the latter 
instance elevations of the water surface were marked beginning at 
the upper end, rather than by proceeding from the lower to the 
upper end. Cross-sectional elevations were taken of the bed of 
the canal at intervals from 0.5 feet to 4 feet, depending upon the 
regularity of the profile. The profiles were subsequently plotted 
and from these the wetted perimeters were measurd with a pair of 
dividers. Wetted areas were computed from the depths rather than 
by measurement with the planimeter. The mathematical mean of 
these values was used in the computations. 
The accompanying diagrams, Figs. 6 to 17 inclusive, illustrate 
typical cross-sections assumed by the canals after years of use. 
Along with some of these profiles are shown dimensions and form 
of the canal as originally constructed, but unfortunately with most 
of them there is no record concerning the shape and dimensions of 
the original cross-section. The cross-sections shown were obtained 
by plotting on a large scale the several cross-sections taken in the 
field, and from these a new perimeter was adopted to represent an 
average. 
