FLOW OF WATER IN DREDGED DRAINAGE DITCHES.. 59 
Consequently it is impossible to formulate definite rules to govern 
one in choosing the proper value of n for any particular channel. It 
is believed that views, together with complete descriptions of chan- 
nels for which the values of n have been determined, afford the best 
means of arriving at the proper value of n to employ for any partic- 
ular channei. This applies especially to existing channels where it 
is desired to ascertain the capacity. 
In order to determine the capacity of a proposed dredged channel 
it is necessary to assume anticipated conditions of channel. As is 
readily seen from the results of these experiments, values of n for 
dredged channels vary greatly, depending principally upon irregu- 
larities of side slopes and cross section due to erosion, caving banks, or 
faulty construction; upon obstructions and growth in the channel 
due to a lack of maintenance; and, under certain conditions, upon 
the effect of a lining of silt in the channel. In most cases where 
erosion takes place in a newly-constructed and well-finished dredged 
channel, the roughness coefficient increases, but the capacity of the 
channel as a rule also increases, since the enlarged cross section more 
than offsets the effect of the increased roughness coefficient. In 
Some instances practically no difference in capacity in a newly- 
dredged channel may result due to erosion, after a certain amount of 
erosion has taken place, as was found to be the case for the experi- 
ments conducted at Trenton, Tenn. (see Table 3). 
CONCLUSIONS. 
A careful study of the results of these experiments suggests the 
following conclusions: 
(1) That a deposit of slick, slimy silt on the sides and bottom of a 
channel greatly reduces frictional resistance to flow (see results 
for Allen and Willow Creeks in Table 4). 
(2) That the clearmg of perennial growth from a channel will 
ereatly increase its capacity (see results for Old Town Creek in Table 1). 
(3) That the growth of grass and weeds in a channel during the 
summer greatly decreases its capacity (see results of experiments for 
North Carolina in Table 5). 
(4) That the accumulation of drift, trees, logs and other obstruc- 
tions in a channel greatly decreases its capacity (see results for South 
Forked Deer River Channel at Campbell’s levee in Table 3). 
(5) That after a certain amount of erosion has taken place in 
a channel, further erosion does not necessarily increase the roughness 
of the perimeter (see results for North Forked Deer River in Table 3). 
(6) That the roughness coefficient n is appreciably higher for a 
roughly dredged channel than for a smoothly dredged one (see results 
for the South Forked Deer River at Jackson and Roberts, in Table 3). 
