THE PLOW OF WATER I]ST CONCRETE PIPE. 43 



intake chamber. It was 1,091.3 seconds before the color appeared 

 at the outlet and 84 seconds more before the last trace disappeared. 



In computing the velocity the elapsed time is taken as from the 

 moment of injection of the solution to the mean between first and last 

 appearances of the color at the outlet. Being injected under a pneu- 

 matic pressure of 60 or 70 pounds against a static head of but 2 or 3 

 pounds, the "shot" of color leaves the color gun in but a fraction of a 

 second, the lever handle of the valve being turned across the orifice 

 slowly but continuously. When tried out in an open channel, where 

 the injection can be watched, a great cloud of dense color, 3 or 4 feet 

 in diameter, can be shot into the water, with an instantaneous opening 

 and closing of the valve, merely by turning the handle through 180°. 



As could be seen in the photographs from which the plate was made, 

 each crack between the boards of the forms was clearly defined in the 

 concrete structure. No opportunity has offered for examination of 

 the interior of the siphon, but the velocities are such that little or no 

 silt should collect in the pipe. The retardation factor of 0.313 con- 

 firms the recommendation of 0.310 for pipes of similar construction. 



No. 35, Experiment S-40. — 63^-inch monolithic reinforced con- 

 crete pipe, Simms Creek siphon, Sun River project, United States 

 Reclamation Service, Montana. 1 Water for irrigation is carried over 

 Simms Creek in a siphon pipe inserted between open channels. 



This siphon was constructed in the winter of 1907-8. It was built 

 in place, steel in 6-foot sections being used for the interior and wood 

 for the outside forms. The inside was then washed with cement 

 grout. From intake to outlet the pipe is 1,556 feet long and 5.3 

 feet in diameter. 



Preparatory to the experiments conducted by the writer, piezome- 

 ter connections of type C (fig. 4) were set in the zenith of the line 

 at distances of 40 and 400 feet from the intake and 20 feet above 

 the outlet. These will be referred to as taps 1, 2, and 3, respect- 

 ively. At the time these taps were set the project manager inspected 

 the pipe and found the interior in a very smooth, clean condition. 

 So far as he is aware, the pipe had never been cleaned, although 

 in its eighth year of operation. 



The loss of head for the various velocities was measured between 

 taps 2 and 3, a mercury column being used at tap 2 and a stilling-box 

 type of water column at tap 3. In order to fill the pipe completely 

 at tap 3 it was necessary to insert flashboards in the slots of the 

 outlet structure. The velocity was determined by accepting the 

 mean time of two or three batches of fluorescein. For observation 1 

 the color was injected at tap 2, but for the other runs the color was 

 injected at tap 1. The first and last appearance was noted at the 

 outlet of the siphon. Thus for all observations other than No. 1 

 the color traversed a reach of 379.5 feet longer than the reach between 

 manometers, while for run 1 the color reach was but 20 feet longer. 

 The line was designed for a capacity of 175 second-feet of water, 

 based on a value of n in the Kutter formula of 0.012. Mr. C. P. 

 Williams, project manager, stated to the writer that "The maximum 

 amount that had been carried by the siphon was about 225 second- 

 feet. At this time there was 255 second-feet diverted into the head 

 of the canal and the loss between the head and the siphon was about 

 30 second-feet." Our experiments also bear out the fact that the 



lEng. Rec, vcfl. 59, 1909, p. 716. 



