34 BULLETIN 852, U. S. DEPARTMENT OF AGRICULTURE. 



typical of present-day careful practice in the citrus groves of southern 

 California where long, well-made units are being used. The value of 

 C s is 0.345. . 



No. 20, Experiment $-57. — 24-inch monolithic pipe, Clavey siphon, 

 Oakdale irrigation district, California. — As described under No. 28, 

 the first portion of the Clavey siphon is 36 inches in diameter from the 

 inlet to a "booster pump" which is used to force part of the water 

 through 3,000 feet of 24-inch pipe to a point 8 feet higher than the inlet 

 to the siphon. Tests were conducted on a reach 1,046.2 feet long 

 between an air valve and a point 6.6 feet upstream from the outlet 

 of the 24-inch pipe. Water columns attached to piezometer tubes 

 of type A were used for both gauges. Velocities were determined 

 with potassium permanganate solutions injected at gauge 1 and 

 observed at the outlet. The color dragged out from 6.2 to 9.6 per cent 

 of the time from moment of injection to the mean time of first and last 

 appearance at the outlet, depending on the intensity of the injection. 

 Although nominally a 24-inch pipe, the mean of 6 measurements made 

 in the last 30 feet of the pipe indicates the actual diameter was 23.7 

 inches, which was used in computations on this pipe. This pipe was 

 constructed with wood forms, both inside and out. The longitudinal 

 ridges made by the cracks in the boards, the circular ridge at the end 

 of one setting of the forms and cemented concrete scraps are shown 

 clearly in Plate II, figure 2, which was photographed from a point 

 about 20 feet in the pipe at the outlet end. The values of C s are 

 erratic, varying from 0.328 to 0.373. 



No. 21, Experiment $-47. — 24-inch jointed concrete pipe, lateral A, 

 British Columbia Fruitlands Co., Kamloops, British Columbia, Can- 

 ada. — A reach of lateral A pipe line 2,306.5 feet long was chosen for 

 test. Water entered the pipe from a box similar to the one described 

 under No. 16. Likewise the installation of the piezometers was identi- 

 cal withNos. 16andl7. The first standpipe was located 64.7 feet from 

 the inlet of the pipe. Between the two gauges were located two turn- 

 out boxes (not in use at time of tests), one 12-inch standpipe, and one 

 waste valve. Water in the standpipe rose to the hydraulic gradient. 

 The writer had been notified that the pipe was partially filled with 

 gravel from the hillside above the main canal. A heavy rainstorm 

 had eroded the hill, filled the main canal near the inlet and regulating 

 box with gravel, and washed some of the latter into the pipe line. 

 The ratio between the measured size of several samples of this pipe 

 and the true area of the water section was found in the following way: 

 For one particular observation, the velocity of the water through the 

 reach was determined with two injections of color — one permanganate 

 and one fluorescein. These agreed within 2 seconds in an elapsed 

 time of 1,800 seconds. The quantity of water was determined by 

 hook-gauge readings on the 3-foot contracted weir at the pipe inlet. 



The area was then found by solving the equation J.=-^« The area 



thus found was 2.754 square feet or 87.5 per cent of the nominal area 

 of a 24-inch pipe and about the equivalent of the area of a pipe 23.5 

 inches in diameter. The mean value of C s is 0.252. 



No. 22, Experiment S-61. — -30-inch jointed concrete pipe, Temescal 

 Water Co., Corona, Calif. — Water for irrigation and domestic use is 

 conveyed over a shallow depression by a siphon pipe 1,028 feet long, 

 inserted between an open concrete-lined channel in sandy soil and 



