80 BULLETIN 376, U. S. DEPARTMENT OF AGEICULTUEE. 



the end of about 3,000 feet a head of nearly GO feet is attained and the pipe ckanged 

 to a 55f-inch continuous-stave pipe of Douglas fir. A reach of this pipe 2,848.2 feet 

 long was tested in 1909 and again in 1910. At the time of the first series of tests the 

 pipe had been in operation about five months. It was new, well rounded, and con- 

 tained but minor distortions. There was no deposit or growth inside. Cross-sec- 

 tional areas of pipe interior were determined by taking four diameters every 200 feet 

 throughout the reach tested. Velocity in the pipe was determined by dividing the 

 discharge, as found by the 18-foot weir, by the mean inside cross-sectional area of the 

 pipe. Mercury manometers were used at both ends of the reach. A comparison of 

 the capacities of this pipe and nearly all other large pipes shows that this siphon is 

 remarkably smooth. This fact is also borne out by the tests on the 48|-inch pipe dis- 

 cussed as No. 43, which is part of this same siphon. This fact is also clearly shown by 

 the relative positions of the points for this pipe in Plate VI. The two series of tests 

 on this pipe were the only ones on any pipe of greater diameter than 18 inches not 

 rejected by Moritz in deriving his formula. This accounts for the difference between 

 the Moritz formula and those of Williams-Hazen, Tutton, and the writer. Giving all 

 weight for large pipes to these two series develops a formula indicating a far greater 

 capacity for large wood-stave pipes than a study of all available tests on such pipes 

 will warrant. If the new formula represents the flow in an average pipe, shown in 

 Tables 2 and 3 to be true, then this pipe will carry more than 18 per cent more water 

 than the average pipe. While conducting tests for the Department of Agriculture the 

 writer visited this pipe after a lapse of four years with a view to securing additional 

 information, but the pipe leaked so badly that tests were not feasible. The pipe was 

 rebuilt in the winter of 1914-15. 



No. 49. Moon Island Conduit, Boston, Mass.^ — In October, 1884, E. C. Clarke 

 made one test on a rectangular conduit, flowing full; that is, as a pipe. This condmt 

 is a tight wooden flume 6 by 6 feet, made of planed plank, laid lengthwise. The ex- 

 perimental section was straight, 2,486.5 feet in length. During this test the flow con- 

 sisted of about one-fourth sewage and about three-fourths salt water. The sides of 

 the conduit were covered with from one-eighth to one-fourth inch of slime below the 

 ordinary flow line. Above this line, on the sides and top, there was some slime but 

 not so much as below the line. Discharge was measured with approximate accuracy 

 by the strokes of the pump pistons. This test was used by Tutton in deriving his 

 formula but rejected by other authorities as the conditions did not parallel those for 

 which the usual pipe is designed. 



Nos. 47-48. 723^-inch Continuous-Stave Douglas Fir Power Trunk Line, 

 Pioneer Electric Power Co., Ogden, Utah.^ — Soon after the construction of the 

 Ogden Canyon pipe line supplying the Pioneer Electric Power Co. plant, near Ogden, 

 Utah, tests were made by Profs. Marx, Wing, and Hoskins, of Leland Stanford 

 Junior University. These tests covered loss of head in the 6-foot wood-stave pipe 

 and the riveted-steel pipe leading from the stave pipe to the power house. Experi- 

 ments were first made in 1897 ^ but were supplemented by a second series of tests in 

 1899.* In both series the discharge was measured through the Venturi meter installed 

 at the plant. The loss of head was measured by the mercury manometers afterwards 

 used by Moritz in the Sunnyside experiments. The relative elevations of the gauges 

 were determined by the static head in the piezometers with the valves closed so that 

 there was no velocity in the pipe. A constant reduction factor was used in converting 

 the mercury column to the equivalent water column. These experiments have been 

 criticized for this reason, but the writer is of the opinion that no error of moment was 

 thus introduced since, in the tests conducted by him, hydrometer readings were taken 



1 E. C. Clarke. Main Drainage Works of the City of Boston, Mass., 2d ed., 1886. 



2 Trans. Amer. Soc. Civ. Engin., 38 (1897), p. 246. 

 3Id., 40(1898), p. 471. 



* Id.. 44 (1900), p. 34. 



