44 BULLETIN 376, U. S. DEPARTMENT OF AGRICULTURE. 
straight. There is one summit, as shown by the view. The maximum head is 74 
feet. Water columns were used for both gauges. Gauge No. 1 was located 88.8 feet 
from the inlet. Gauge No. 2 was located 5,965.9 feet from gauge No. 1 and 115.7 feet 
from the outlet. Velocity within the pipe was determined by fluorescein tests, the 
mean time of travel of five batches of color being accepted. Levels were determined 
by static head ; that is, on one visit to the pipe it happened that no water was running, 
so simultaneous readings at both gauges, taken 10 seconds apart for 1 minute, gave a 
true water level. The extremely slow fall of the water surface throughout these 
readings indicated that the leakage was negligible, which fact was also apparent to 
the eye. As well as an examination of the pipe outlet would disclose, the pipe was 
clean and smooth on the interior, though an examination of the long, low stretch 
across the marshy bottom of the swale might have shown deposits. The writer does 
not believe this likely, however, for the reason that the water flows for about 6 miles 
in open channels at comparatively low velocities before reaching the intake, and in 
these channels the heavier sand would have precipitated, leaving the water little 
more than clouded. The one observation at the commercial velocity indicates the 
capacity of this pipe to be about 10 per cent greater than that computed by the new 
formula. 
No. 40, Expt. S-6, 40-inch Continuous-Stave Douglas Fir Siphon Pipe, 
Burbank Co., Washington. — Irrigation water is conveyed across a depression 
between two sections of open channel by a continuous-stave siphon, built in Decem- 
ber, 1912, along the west side of section 6, township 8 north, range 31 east. This pipe 
was constructed on the surface of the ground and supported on cradles. At the time 
of these experiments, therefore, it was in its second irrigation season. As shown by 
the profile in Plate III, figure 1, there is one summit on the reach tested, but as this 
is protected by a standpipe, there was probably no air accumulation at the summit 
at the time of this test. Although the pipe is about 2,900 feet long, a reach 927.4 feet 
long was chosen near the outlet end for the reason that there is a diversion from the 
lowest point of the pipe. Gauge No. 1, a mercury manometer, was located 1,049.6 
feet above the outlet and gauge No. 2, a water column, was located 122.2 feet above 
the outlet. The water divided in the outlet structure, flowing in two directions, one 
stream continuing in an earth channel and the other in a concrete-lined channel. 
The discharge in the pipe was determined by the sum of the flows in these two chan- 
nels, as measured by current meter. The nominal area of the pipe was accepted as 
correct. At velocities exceeding 2 feet per second, it was noticeable that sections of 
the pipe immediately following the sharpest vertical curves vibrate about 1 inch, 
vertically, upon the cradles. This emphasizes the necessity for securing anchorage 
at bends. The two observations taken at commercial velocities indicate the capacity 
of this pipe to be about 3 per cent greater than that computed by the new formula. 
No. 42, Expt. S-9, 48-inch Continuous-Stave Redwood Siphon Pipe, 
Cowiche Siphon, Yakima Y alley Canal Co., Washington. — Water for irrigation 
is conveyed across Cowiche Canyon, about 4 miles from North Yakima, Wash., in a 
redwood siphon built in January, 1914. (PI. IV, fig. 1.) Gauge No. 1, a mercury 
manometer, was located 67.3 feet from the inlet (PI. Ill, fig. 2), while gauge No. 2, a 
water column, was located but 7.6 feet from the outlet. The inlet to the pipe is at 
the bottom of a concrete well about 10 feet deep. Subsequent tests to determine entry 
losses showed that much air was entrained and carried into the pipe, but no influence 
of air was apparent at gauge No. 1, which was attached to the pipe at the mid-point 
of its left side. From the intake to gauge No. 1 the pipe is straight. This is likewise 
true of the pipe for about 100 feet before gauge No. 2 is reached. For the balance of 
the distance between gauges the pipe is virtually one long vertical curve, as it is under 
a maximum head of about 100 feet and the total length is but 962.3 feet. The pipe 
has but one gentle bend in horizontal alignment. For each of the several runs made 
with different velocities in this pipe fluorescein was timed from inlet to outlet, the 
