26 BULLETIN 831, U. S. DEPARTMENT OF AGRICULTURE. 



necessary to effect a complete priming were less under such conditions 

 or when the outlet lip of the discharge leg was so deeply submerged 

 as to set up reverse action in the expulsion of the entrained air. 



iSince the tests referred to were made a system of relief valves has 

 been installed in connection with the battery of seven siphons placed 

 in the Huntington Lake dam in California, the discharge from which 

 amounts to 5,000 second-feet. The description of these siphons fol- 

 lows in another part of this paper. 



SIPHON AT GIBSWIL. SWITZERLAND. 



A siphon at Gibswil in Switzerland, built with a sloping instead of 

 a vertical outlet leg, is described as consisting of a |-inch riveted 

 steel tube tapering from 31.5 to 23.6 inches in diameter, the assumption 

 being that the taper would tend to keep the water column from part- 

 ing under the 52.48-foot head which was utilized. The inlet pipe 

 was cut on a horizontal plane at the normal high-water surface and 

 was incased in a reinforced concrete hood projecting 3.28 feet below 

 normal water surface, so as to prevent the entrance of ice or float- 

 ing debris. The air control, to break the action of the siphon and to 

 prevent the water from being drawn down into the reservoir below 

 normal surface, consisted of long narrow slots cut through three of 

 the sides. When the water rises these slots are closed and siphonic 

 action begins. A series of tests to determine over-all efficiency for 

 this siphon gave a discharge of 98.9 cubic feet per second, but this 

 was considered inaccurate and lower than the real efficiency of the 

 siphon, because it was found that some of the air slots were not fully 

 sealed. A maximum efficiency jdelding 123.G cubic fed per second 

 was determined as a more accurate assumption of the real capacity. 

 Computation of the end area at 3.03 square feet would give a corre- 

 sponding velocity of 40.8 feet per second. The velocity due to 52.48 

 feet head is V= V2gH = 58.06 feet per second and thus the efficiency k 



is ' nr = 0.70 as a coefficient of discharge. The computed friction loss 



in the pipe alone equaled 10.2 feet, which certainly is high and con- 

 firms the statement made in another part of this paper that the in- 

 crease in the length of the tube beyond 34 feet would reduce the 

 value k. In addition, this tube had the added friction produced by 

 building the draft tube on a slope and thus making it longer. 



In another siphon installation, also in Switzerland, the conditions 

 to be overcome as the result of conflicting requirements of several 

 plants and their water rights, were peculiar. There was a spinning 

 and weaving mill operated by hydraulic machinery, the tail water 

 *rom which, up to a maximum of 56.5 second-feet, was appropriated 

 by a twine plant farther down the stream. The discharge from the 

 first plant in excess of the 56.5 second-feet had to be led over a weir 



