36 BULLETIN. 832, U.S. DEPARTMENT OF AGRICULTURE. 
[Cross-sectional area=315.6 square feet. Hydraulic radius=5.45 feet. Slope=0.0003.] 
} 
| 
Fe Relative 
Channel, i rete | Vv lect eimrae = 
y: S®- | charges. 
| > Pt. 
| per sec. | Sec. feet. | Per cent. 
UIE er ee BR nn a | 0.0367 56.2 2-97 715.1 100.0 
OL HSA STH eR oe Seb ote co cee ade st boeSoe ene - 0619 34.8 1.40 443.5 62.1 
Old crookeds = "5- 2445222 eee Saas fa acta - 1460 15.1 0. 61 192.7 27.0 
Results of the above computations show that for similar cross 
sections and siope the dredged, old straight, and old crooked channels 
would carry 715.1, 443.5, and 192.7 second-feet, respectively, the 
capacity of the old straight channel being only 62.1 per cent of that 
of the dredged, and of the old crooked channel only 27 per cent of the 
dredged. The difference in the capacities of the old straight channel 
and old crocked channel is not due entirely to the curves in the latter, 
since the accumulation of drift, trees, and logs was greater in the 
erooked channel. However, the difference in the condition of the 
two channels may be directly attributed to the presence of the 
curves, since there is a greater tendency for drift and other obstruc- 
tions to accumulate in a crooked than in a straight course of channel. 
DISCUSSION OF WESTERN TENNESSEE EXPERIMENTS. 
The experiments in western Tennessee cover a greater variety of 
conditions in channels than do any of the other six sets, the values 
of n obtained ranging from 0.024 for the dredged channels near 
Roberts and Trenton to 0.162 for the old crooked river channel near 
Jackson. The results, however, do not justify the use of a coefficient 
as low as the first named above in the design of dredged channels, 
since it is not the rule that a dredged channel maintains its original 
efficiency. It is not believed that a value of 7 less than 0:030 should 
be used for this section of the country, and it will be noted that this 
statement is in agreement with the conclusions deduced for the two 
sets of experiments made in Mississippi. Where it is desired to com- 
pute the capacity of an existing channel, the proper value of n can 
be selected by a comparison of the conditions in the channel with 
the conditions described for the various channels for which n has 
been determined in these experiments. 
EXPERIMENTS IN WESTERN IOWA. 
Experiments in western Jowa were conducted on the following 
streams: Allen Creek, Willow Creek, Pigeon Creek, Boyer River, 
the Monona-Harrison Ditch, and Little Sioux River. 
