THE FLOW OF WATER IN" CONCRETE PIPE. 39 
The large pool just above the 10-foot Cipolletti weir was cleared of 
sand just prior to the experiments. The end and bottom contrac- 
tions were equal to about twice the depth of water over the crest. 
The velocity of approach was about 1 foot per second. More exact 
figures were not feasible, due to shape of pool. The current-meter 
measurements were made by the hydrographer of the project simul- 
taneously with the reading of gauges and the passage of the color in 
the pipe. 
During the summer of 1917 the writer again visited this pipe and 
made additional tests on approximately the same reach of pipe. 
There being an abundance of water it was feasible to vary the dis- 
charge through the pipe and thus secure tests at various velocities. 
Gauge No. 1 was attached to a piezometer of type A while the 
piezometer at gauge No. 2 was very similar to that in the 1915 tests, 
but a glass gauge tube was used instead of a stilling box. 
The velocities were determined by the travel of fluorescein solu- 
tions, which were timed from the intake of the siphon to the outlet. 
The average value of C s for all the tests was 0.395. The high veloci- 
ties probably assure a pipe free from sediment and the conditions at 
the intake are such that no gravel or heavy detritus would enter the 
siphon. As may be seen from the photograph, this pipe is excep- 
tionally straight and under most favorable flow conditions. The 
writer would consider the coefficients found as verifying a working 
coefficient of 0.370 for glazed pipe under average conditions of 
curvature. 
No. 30, Experiment S-49. — 42-inch jointed reinforced concrete 
pipe, Victoria Aqueduct, siphon No. 1, Vancouver Island, British 
Columbia, Canada. — A pipe line quite typical of present-day high- 
grade construction has recently oeen "built to convey water for 
municipal purposes from Sooke Lake, a mountain reservoir, to 
Humpback Reservoir, about 10 miles from Victoria, British Columbia. 1 
This line, 27.3 miles in length, follows the hydraulic gradient as a 
flow line with the exception of 6 inverted siphons. The slope 
throughout is 0.001 foot per foot or 1 foot fall per thousand feet of 
length. Slightly more than one-half the line is curved. The pipe 
units are of the lock-joint type, each 4 feet long and exactly 42 inches 
in diameter. The shell of the flow line is 3 inches thick and that of 
the siphons 4^ inches thick. The maximum head on any of the 
siphons is 90 feet. The flow-line pipe is reinforced with triangular 
mesh wire and square steel bars are used in the pressure units. The 
concrete, a "wet mix/' composed of 1 part cement, 2 parts clean 
sharp sand, and 4 parts of beach gravel (maximum diameter three-* 
fourths inch), was tamped into steel forms with a suitable slice bar, 
care being taken to force the coarse aggregate away from the forms, 
thus making a smoother surface. 
The forms were well wiped with oil just prior to each pouring. 
After being filled they were placed in a steam room for three hours; 
the forms were then stripped and the pipe, resting on a base ring, 
again subjected to. steam for an additional four hours. The interior 
surface of pipes made in this way is unusually smooth. At times 
there was small pitting, the cause of which remained obscure. If 
1 Engirt. Rec, vol. 69, Feb. 21, 1914, p. 225; id., vol. 72, Oct. 2, 1915, p. 406; id., vol. 72, Oct. 23, 1915, 
p. 507. Can. Engr., July 23, 1914, vol.27, p. 197; id., June 10, 1915, vol. 28, p. 652. Western Engineering, 
vol. 5, Sept. 1914, p. 105. 
