684 EXPERIMENT STATION RECOED. 



higher heads." It is stated that the principal source of error in using this box 

 in practice will be the difficulty in measuring the depth over the weir closely. 



The Iliverslde box consists of a shallow box set over the end of the delivery 

 pipe line. The water enters through the bottom of the box and is measured out 

 through an adjustable cast-iron mea.suriug plate in the end. In tests of this 

 device the average difference between the number of inches actually received 

 and the area in square inches of the opening was about 2 per cent. For all 

 tests the area in square inches of the opening averaged 1 per cent greater than 

 the inches actually received. It is thought that where care is used to adjust 

 the width of the opening to the amount running this box will measure water 

 very closely. 



In tests of the Foote inch box the amount of water supposed to have been 

 passed averaged 4 per cent greater than was actually run and the error did not 

 vary with the amount of discharge. " From these tests it appears that the 

 slide can be set within an average 4 per cent of correct if care i^ used. This 

 box will measure water up to 150 in. satisfactorily under conditions to which 

 it is adapted, although it is not in general an economical box to use." 



A general discussion of different types of weirs in use is given, including 

 weir tables which are in general use. 



Submerged orifices are also described and tables developed by the U. S. 

 Reclamation Service for determining their flow under different conditions are 

 given. Tests of a submerged orifice 2 ft. wide and 1 ft. high gave a coefficient 

 of 0.61 which agrees with that determined by the Reclamation Service. In 

 tests of a submerged orifice gate under two conditions it was found that the 

 mean of all measurements using the level board gave a mean coefficient of 0.8 

 while the measurements on the gate gave a mean of 0.72. " From these results 

 it is seen that the coefficient for such measuring gates varies with the type of 

 gate. . . . Where the lack of sufficient fall for the use of a better measuring 

 device makes the use of this type of submei'ged orifice necessary, a standard 

 size and structure should be adopted, and special discharge tables prepared. 

 This should then be rated under the condition in which it will be used." 



Three mechanical devices for measuring and registering the total flow were 

 tested. The Dethridge meter consists of a wheel or drum to which projecting 

 pieces of sheet metal are fastened. The drum is placed with its axle horizontal 

 and is so set that the projecting blades are in the current' of the ditch to be 

 measured. A special box is built around the wheel so that all water in pass- 

 ing has to strike against the blades. The tests of this device showed the meter 

 to be quite accurate under constant ditch conditions between rates of flow of 

 from 1 to 3.5 second-feet. For both larger and smaller discharges the meter 

 passed more water than it did between these limits. " The amount of water 

 going through the meter varies with the depth of drov\-ning. A meter set high 

 in the ditch will discharge less water per revolution than one set low. Check- 

 ing up the ditch below a meter so that the depth is increased at the meter may 

 increase the discharge by as much as 10 per cent in some cases. . . . The 

 Dethridge meter of this size is adapted for accurate measurement of streams 

 varying from 1 to 3 or 4 cu. ft. per second." 



The Graut-Michell meter consists of a wheel turning in a horizontal circular 

 opening through which the water is made to pass. The meter consists of four 

 flat blades set so that the water in flowing through the circular opening strikes 

 against them at an angle. On the upper end of the shaft carrying the wheel 

 is a counter which records the number of revolutions of the wheel. Tests 

 made of the 21-in. meter showed that for discharges of over 2 second-feet and 

 up to 6.5 second-feet the meter makes one revolution for every 6.1 cubic feet of 



