Measurement and Division op Water. 39 
under corresponding conditions. Under the best of conditions the 
result is subject to much uncertainty. 
Thus the table gives the values of “c” for a few cases, and 
shows the great variation which comes with a different value of 
“n." Thus with a moderate sized ditch, hydraulic mean depth of 
two feet, the coefficient “c” would be 98 with “n” .017, and only 
66 for “n” .025. One result is 50% over the other. These would 
be a difference of 27% made by selection of .020 or .025. 
There is a very common inclination by courts and lawyers to 
accept any results by the Knitter formula as beyond question, but 
these show how much the result depends on the coefficient “n/’ 
and this choice can only be made with much reserve, and with 
knowledge of results of actual tests in similar conditions. 
Nor does it follow that “n” remains the same in the same canal 
at different depths, or that it remains the same with the same depth 
for any length of time. Thus I have found “n” to vary from .022 
to .027 in the same canal, the greater value of “n”, and hence the 
smaller value of “c” with the increased depth. This was attiibuted 
to the fact that with the greater depth the water partially over¬ 
flowed the bank and was retarded by underbrush. The growth of 
weeds during the summer often makes a noticeable effect on the 
carrying capacity of a canal for the same depths. 
Hence, while the Kutter formula is of great usefulness, it 
must be used with intelligence, with a knowledge of its limitations, 
and realizing that the results under the best of conditions may be 
considerably in error. The expression for “c” in English measures 
is complicated, but this arises partly liom the fact that it was 
derived in French measures, and the constants were simple in that 
system, but have become more complex in the transfer of units. 
Had the formula been derived in English units in the first place, or if 
some one should go to the trouble to obtain the expression so as to 
use English units, the apparent complexity would be much reduced. 
However, with the access to tables now easy, the engineers do not 
have occasion to use the formula itself, and this complexity does 
not matter. 
It should not be forgotten that the whole purpose is to get 
“c” in the formula on page 38. Books like Trautwines Civil En¬ 
gineer’s Pocket Book have tables extensive enough for common 
use; other more elaborate tables have been prepared. The attempt 
to compute “c” to decimals is time wasted, for the error in the 
judgment of “it’’ may introduce a much larger error. Practically 
some forms of diagrams are accurate enough and much quicker 
to use. 
