Measurement and Division oe Water. 37 
through smoothly and without ripple. In such cases it is desirable 
to measure the depth in a well of still water connectly freely with 
the water in the flume. Without some such method the results 
are often discordant. 
The results with fairly moderate quantities of water ought to 
be correct within 5%, and with proper care in construction of 
weir and its rating, to still less error. It is not so reliable for 
smaller channels and smaller quantities, but still may often be 
used with advantage where the weir cannot be used. One of the 
principal reasons why it is not adopted for measuring water to 
consumers is that the rating of these flumes calls for considerable 
skill as well as an instrumental equipment not frequently available. 
The errors in small ditches like laterals are also much larger 
than in large ones. Hence it is not to be used for this purpose, 
though for measurement into canals or into the larger laterals it 
may be entirely satisfactory. There may be reasons which force one 
to choose rating flumes as the last available method for the condi¬ 
tions, as, for instance, where sand or silt fills up the weir box, or 
where no fall is available. If there were enough need, I have no 
doubt that the rating flumes for small ditches could be developed 
so as to be nearly as accurate as the weir under the ordinary con¬ 
ditions. It would require special current meters adapted to the 
conditions. 
COMPUTATION OE QUANTITY FROM THE SECTION AND EAUU. 
It is often necessary to determine the amount which a canal 
will carry, or to decide on the size required. For this purpose 
the formula commonly known as Kutter’s is generally used. The 
results are to be taken with considerable latitude, and as will be 
seen are not exact enough to serve as a basis for division of water. 
These variations may be easily as much as 20%, and even then 
require accurate determinations of the slope of the water surface, 
(not the bottom of the canal) not commonly available, while the 
ordinary demands of the ditch call for accuracy whhin 5% or 10%. 
The calculation by computation attempts to determine the 
velocity of the water, while the quantity, which is usually wanted, 
is the product of the velocity and the area of the cross-section. 
These should be expressed in feet-per-second, and in square feet, 
in order that the result be in cubic-feet-per-second which is usually 
wanted. 
It is a matter of common observation that the velocity of 
water depends upon the section, upon the grade, and also upon the 
roughness. That is, given two canals of the same grade, the 
larger one will have the greater velocity. It is common observation 
that when a canal is enlarged it scours and checks must be built. 
The sides have a retarding influence somewhat in proportion to the 
