6 BULLETIN 832, U. S. DEPARTMENT OF AGRICULTURE. 



ence in velocity heads was found to be unnecessary, so far as the 

 practical value of the results was concerned. 



In calculating values of n, the values of V, R, and S were first deter- 

 mined as explained in the foregoing, C was then computed by Chezy's 

 formula, and with C, R, and S known, the value of n was obtained 

 from a large-scale diagram similar to, but larger than, the one bound 

 in the treatise, "The Flow of Water in Rivers and Other Channels," 

 by Ganguillet and Kutter, as translated by Hering and Trautwine. 



TABULATED RESULTS. 



In Tables 1 to 6 are given the hydraulic elements and values of C 

 and n for the six different sets of experiments. In the third column 

 is shown the average of the maximum depths as measured at each of 

 the cross sections. At the foot of this column is shown the average 

 maximum depth for a bankful stage of each channel, which enables 

 the reader to estimate about what proportion of the channel is filled 

 for each measurement. In the fourth column is given the average 

 surface width along the slope course at the time of each measurement. 

 With these two dimensions, the average cross-sectional area given 

 in column 6, and the accompanying cross-sectional figures for 

 each channel, a good idea can be obtained as to the shape and size of 

 the channel. The headings at the tops of the other columns in the 

 tables are self-explanatory. On most of the channels values of n 

 were determined for several stages ranging from low to high. 



Attention is called to the specially interesting features of some of the 

 experiments. For instance, values of n for Old Town Creek, Table 

 1, were obtained before and after clearing, in order to show the effect 

 upon n as produced by the presence of growth in the channel. 



On the South Forked Deer River at Campbell's levee, near Jackson* 

 Tenm, experiments were conducted to determine the relative values 

 of n for three consecutive courses of channel; the first, an old straight 

 course of the river channel ; the second, an old crooked course of the 

 river channel; and the third, a course of an irregularly dredged channel. 

 The results which are given in Table 3, show a remarkable difference 

 in the values of n obtained for the three different conditions of channel. 



In the results of the Iowa experiments (Table 4) attention is called 

 to the unusually low values of n obtained for earth channels. These 

 low values, which were obtained for Boyer River and Allen and Wil- 

 low Creeks, were due, no doubt, to the smoothing up of irregularities 

 by silting and to the lubricating effect of silly mud deposited on the 

 bottom and side slopes of the channels. The increase in the value of 

 n, due to a caving bank covered with a heavy growth, is shown by the 

 experiments made on the Little Sioux River cut-off channel before 

 and after the caving took place. 



