THE FLOW OF WATER IN CONCRETE PIPE. 35 
a round pipe flow line. The siphon pipe is made of precast units, 
each 3 feet in length. The pipe is practically straight in both hori- 
zontal and vertical planes, the maximum head being about 8 feet. 
This line was constructed in 1911, being thus in its fifth year of 
operation at the time of the experiment. It carries water through- 
out the year. A reach was chosen for test between a point 79 feet 
from the intake chamber and a point 69 feet upstream from the 
outlet manhole. Water columns attached to piezometers of type 
A were used at both ends of the reach. Velocities were determined 
by the use of solutions of potassium permanganate, injected at gauge 
No. 1 and observed at the manhole be} r ond gauge No. 2. Measure- 
ments of the initial pipe section at the intake chamber showed the 
diameter to be 2.49 feet, or 0.01 foot less than the nominal diameter. 
No. 23, Experiment $-34. — 30-inch Jointed Reinforced Concrete 
Pipe, "D x Line 1 ', Umatilla project, United States Reclamation Service, 
Oregon. — Water for irrigation is conveyed across the depression near 
Hermiston, Oregon, in a 30-inch pipe laid in sections, each 4 feet long. 1 
This line was constructed in the winter of 1909-10 and tested for 
friction losses in 1911 by Mr. H. D. Newell. 2 The writer conducted 
tests on the same reach of pipe during the season of 1915. See No. 
12a (page 77) and No. 24a, for pipe description. Thus while com- 
paratively new when first tested it had been in use nearly six seasons 
when these experiments were made. The pipe units were each 4 feet 
in length, cast in wood forms that had been coated with No. 26 sheet 
steel. The mixture and wash coating is described under pipe No. 32, 
page 41 . Gage No. 1 was a piezometer tube of type B inserted into the 
water section of the pipe through a stuffing box, located above the 
valve A in figure 3, Plate 2. Newell appears to have used the pressure 
directly from a f-inch valve set in the pipe line. The writer found so 
much air in the water column taken directly from this tap that he 
judged a piezometer tube, even of type B, would give more nearly the 
true pressure head (see discussion on page 15). This tube was con- 
nected with a mercury manometer. At the outlet a piezometer tube 
of type A was thrust 3.6 feet into the pipe, against the current. This 
tube was connected with a stilling box and the water surface in the 
latter was read by a hook gauge. The relation of the points plotted 
on Plate VI for the Newell tests (No. 24) and for those made by the 
writer (No. 23) indicate that the capacity of this pipe has diminished 
slightly. This was to be expected, for Newell speaks of deposits in 
the pipe after but one season while the writer's tests were made in 
the sixth season. Until July of each year the water for this pipe 
comes directly from the Umatilla River, through about 4 miles of 
open canal, the last reach of which discharges in a direct line into 
the pipe intake. This river water, taken during the high-water 
period, would undoubtedly contribute greatly toward silt and debris 
within the pipe as the maximum demand so far has not required a 
velocity in the siphon in excess of about 2.5 feet per second. Veloci- 
ties were determined directly with solutions of fluorescein, the mean 
time of several batches being accepted in computing the mean velocity. 
This pipe is quite straight in horizontal alignment, while the vertical 
curves are long and gentle. There are five 6-inch valves and three 
manholes on the reach tested. Examination of the outlet indicates 
that the pipe is smooth and slimy inside. The color was injected into 
- Eng. News, Feb. 16, 1911, Vol. 65, p. 208. * id. May 1, 1913, Vol. 59, p. 904. 
