36 The Colorado Experiment Station. 
the edge should be sharp on the upstream side; it should be at 
some distance above the bottom of the channel; the two sides of 
the weir should be equally inclined to the vertical and there should 
be free flow of the water away from it. The point of the notch 
would not need to be as much above the bottom of the channel 
as the ordinary weir. 
RATING FLUMES. 
Weirs can not be conveniently used to measure water into 
canals, or even into large laterals. The fall required makes them 
inapplicable in most cases. Hence the method used is some form 
of rating flumes. These are recognized by law and the state 
officers may require them to be constructed. 
A rating flume is commonly a flume built in the ditch some 
distance below the head gate with a rod graduated to shew the 
depth. Measurements are made with a current meter to determine 
the quantities at several depths, and tables may then be prepared 
which will show the quantity at intermediate depths. There should 
be at least four such guagings made. If the conditions are good 
the results when plotted to scale, fall on a regular curve, and from 
that curve the quantities for intermediate depths may be taken with 
considerable confidence. A table is then prepared giving the quan¬ 
tities for different depths, and especially indicating the depth re¬ 
quired to give the various appropriations of the canal. 1 have 
agree within one per cent. 
The results depend on the skill of the gauger, on the accuracy 
of his meter, and the skill exercised in its “rating,” and then on 
the conditions of the rating flume so that conditions are essentially 
the same at different times for the same depth, and there are no 
irregularities in the flow. 
The flume should be far enough from the head gate to avoid 
the rush of the current which sometimes may be felt for consid¬ 
erable distances when the stream is high and the gates nearly shut; 
it should be in a straight reach of the canal, so that the current is 
central in the canal and free from eddies or boils due to roughness; 
it should not fill up with sand in some stages, to be scoured out 
at others. 
The purpose of the flume is primarily to secure a constant 
bottom and cross section. Often this may be done by a simple 
construction, or even in rocky channels by one or more timbers 
across the bottom of the canal to fix the points of reference. It is 
common to make the sides of the flume vertical. In large canals 
or where the velocity is great, this may cause a considerable per¬ 
turbation or contraction in the current. Hence I have sometimes 
caused such flumes to be built with the sides flaring so as to corres¬ 
pond to the slope of the canal bank, and thus the water passes 
