THE FLOW OF WATER IN CONCRETE PIPE. 83 
forms, which were well oiled or greased at each set-up, assured a 
dense, smooth surface. 
The equipment and method of taking the observations were the 
same as those employed in the tests of the Rondout tunnel (p. 81). 
It is to be noted from Table 3 that the same discharge, hence the 
same velocity, held for the observation taken on this tunnel and the 
Rondout tunnel, for any given day. The tunnels are 5 miles apart, 
and this one is 18 miles beyond the Venturi meter, where the dis- 
charge was determined. (See p. 82.) 
When the position and sequence of points is studied on Plate VI 
the position does not appear to follow a rather definite relationship 
to the sequence, as was the case for the Rondout tunnel, but the 
points appear to be indiscriminately placed on either side of the line 
through the centers of gravity, as is the case with nearly all hydraulic 
experimentation. 
No. 41, Experiment RDJ. — 18-foot monolithic concrete-lined 
tunnel, No. 2 conduit of Ontario Power Co., Niagara Falls, Ontario, 
Canada. — For the following data concerning some unpublished experi- 
ments the writer is indebted to the Hydro-Electric Power Commission 
of Ontario. These experiments, conducted under the supervision of 
R. D. Johnston, are of especial importance for two reasons. So far 
as the writer is aware, the conduit is the largest on this continent, if 
not in the world. The velocities encountered in commercial opera- 
tion are three or four times as great as those ordinarily considered 
as high velocities in long penstocks. The following description is 
submitted by the commission : 
This conduit consists of approximately 6,726 feet of concrete pipe, of which 6,646 
feet, located between the gatehouse and surge tank, was tested. Of this 6,646 feet 
approximately 5,170 feet is straight pipe and 1,475 feet in length is bends to the right 
of 800 feet radius. These bends consist of short curves separated by short tangents. 
The pipe is of oblate shape, approximately 18 feet in diameter, the greatest hori- 
zontal dimension being 19.26 feet and the greatest vertical 16.55 feet. The surface 
of pipe is unusually smooth and even. During construction great care was taken to 
D ee that the concrete was carefully spaded next to the oiled steel forms, and after the 
whole pipe had been erected all defects due to imperfect alignment of forms were 
removed by chipping and then the whole inside was rubbed down by hand with 
carborundum bricks. This pipe delivers water by means of a distributor and pen- 
stock to 7 turbines developing a total of approximately 91,500 horsepower. The 
pipe was put in operation in 1910 and the tests were made in 1913 and 1914. An 
examination of the interior made in April, 1918, showed no signs of cavitation or 
wear and a total absence of vegetable growth. This latter condition is probably due 
to the extremely high velocities, which at times reach 28 feet per second. The 
inside surface is shown on the accompanying photographs taken during the inspection 
mentioned above. (PI. XL) 
For the tests the velocity was measured by the "color method." This consists of 
liberating some coloring substance in the water at a known point and at a given time 
and recording the time taken for this color to be carried to another known point. 
The computations were based on an elapsed time, as from the moment of injection 
to the mean between first and last appearance of the color. The volume of the pipe 
or other container between the two points is measured or calculated, and this gives 
the total volume of water which has passed the second point of observation in the 
recorded time. The velocity of water at any section can then be calculated. In 
this case the point of introduction of the color was in the mouth of the pipe in the 
gatehouse and the point of observation was at the tailrace weir opposite No. 7 unit, 
which is the first one served by this conduit. Adjustments necessary for the change 
in velocity in the penstock, etc., were made after further tests had been made on the 
latter. (For a more detailed description of this method see Engineering News of 
Sept. 23, 1915.) 
The loss of head was obtained by observing the water levels at the gatehouse and 
at the surge tank. The surge-tank riser served as a huge manometer, in which all 
