82 BULLETIN 852, U. S. DEPARTMENT OF AGRICULTURE. 



uptake shaft. Disturbance was eliminated at this point by passing 

 the flow over stop planks below the shaft. 



In the computations it was assumed that the velocity head within 

 the uptake shaft was changed to pressure head without loss. Obser- 

 vations 1 to 4, inclusive, were computed from 31 readings at 5-minute 

 intervals for each observation. For the other observations readings 

 were taken at minute and half-minute intervals for 124 readings. 

 Observations were made about noon on each day, after which the dis- 

 charge was changed and allowed until the next noon to secure a 

 stable regimen of flow. As stop planks caused a break in the 

 hydraulic gradient immediately below the reach tested, this regimen 

 ol flow was quickly attained. 



The discharge was measured through a 210-inch Venturi meter 

 located S miles above the tunnel. The leakage was known to be 

 small and negligible from tests previously made. 



At the time of test the tunnel was new, but had been in operation 

 for sufficient time to allow the surface to become somewhat foul, 

 although the observations were not continued long enough to indicate 

 the probable ultimate deterioration, hydraulically considered, of the 

 surface which long and constant use might bring about. 



The concrete lining, 17 inches in thickness, was deposited against 

 clean, oiled steel forms, being carefully spaded near the forms to assure 

 a dense, smooth inner surface. The mixture used was, in general, 

 composed of one part cement to two of sand and four of crushed stone. 1 



In conducting the experiments, the discharge (hence the velocity) 

 was reduced from day to day until five observations had been taken. 

 It was then increased from day to day until three more observations 

 had been recorded, the velocities being about the same as those for 

 the first three observations. 



A study of Plate VI indicates that for a given loss of head the 

 velocity during the latter end of the series would be about 13 per 

 cent less than that occuring during the beginning of the series. 



In reply to a question as to " shoulders" within the tunnels, the 

 board writes: 



Regarding "shoulders," by which is understood the irregularities in the concrete 

 surface of waterways occurring at the places where form sections joined, there were 

 such irregularities as commonly occur in the pressure-tunnel linings, but this effect 

 hydraulically is unknown, because of lack of opportunity to compare, under other- 

 wise identical conditions, the measurements as made with results which might be 

 obtained with a smoother lining. The mortar linings of the pipe siphons were 

 unusually free from form irregularities, such as exist being clearly of negligible effect, 

 hydraulically. 



No. 40, Experiment M-3. — Wallkill pressure tunnel, Catskill 

 Aqueduct, N. Y. — At the Wallkill River crossing the Catskill 

 Aqueduct takes the form of a circular pressure tunnel, 4.4 miles 

 long, excavated in solid rock with a minimum cover of 150 feet, the 

 tunnel being then lined with concrete. Thus, like the Rondout 

 tunnel, the conduit is a circular pipe, when considered from the 

 standpoint of capacity. (See PL IX, fig. 1.) 



The concrete was placed as a moderately dry mixture of 1 part 

 cement to 6 parts of aggregate. 2 Careful spading against the steel 



1 Engin. Rcc., Jan. 1, vol. 61, 1910, p. 20; id., Jan. 29, vol. 61, 1910, p. 138; id., Feb. 28, vol. 69, 1914, p. 240; 

 id., Sept. 17, 1910, vol. 62 p. 312. 

 * Eng. Rec., Apr. 2, vol. 61, 1910, p. 450. 



