THE FLOW OF WATER IN WOOD-STAVE PIPE. 9 



test is the more remarkable in view of the fact that there are "in 

 addition to a succession of sweeping horizontal and vertical curves, 

 27 cast-iron bends with a radius of curvature of 5 feet, and with an 

 average central angle of about 31°." (See PI. XIV, fig. 1.) 



The next test (No. 32), spoken of by F. B. Gutelius, ^ was conducted 

 by D. C. Henny ^ on the Butte, Mont., 24-inch pipe in 1892. In this 

 test the value of n was found to be 0.0103. Here again was a value 

 so close to 0.010 that it tended to estabhsh the fact that 0.010 was, 

 closely, the proper value. After these three tests by three different 

 experimenters an extensive series of tests would naturally be required 

 to convince the profession that a higher value of n should be used. 

 Attention may here be directed to the fact, however, that in the 

 three last-mentioned tests but one or two runs of water, with Httle 

 or no variation in velocity, had been observed. 



In 1897 Profs. Marx, Wing, and Hoskins ^ of Leland Stanford 

 Junior University, made a careful and vastly more extensive series 

 of tests than any previously carried through (No. 47). The values 

 of n varied inconsistently between 0.010 and 0.0204. (See Table 2 

 column 10.) Of the 22 runs in this series, all but one showed a value 

 of n above 0.0123, and most of them showed it above 0.013. 



The experimenters did not pubhsh their results as values of n, 

 merely stating : * 



In regard to the applicability of Kutter's formula it is to be said that the experiments 

 on the wooden pipe herein described give values of n ranging from 0.012 to 0.015, an 

 average value being perhaps 0.013. The difference between this value and those given 

 for the Denver and Butte city conduits can hardly be attributed to the greater rough- 

 ness of the Ogden pipe. It is rather to be supposed that the Kutter formula is defec- 

 tive. (See p. 56.) 



In correspondence relating to these tests T. A. Noble offers the 

 values of n for the various observations.^ These have been checked 

 by ther writer and are found in Table 2, column 10. In the same 

 correspondence (p. 544) A. L. Adams offers his tests on the West 

 Los Angeles pipe (No. 20) where the values of n range from 0.0105 to 

 0.0111. Mr. Adams voices the following warning: 



These values * * * do not indicate 0.01 as being a safe assumed value for n 

 as have all previous experiments. 



Various arguments were brought forward to furnish a reason for 

 the imprecedentedly high values of n in the Ogden pipe. These 

 included the following: That Kutter's formula did not apply to pipes 

 as large as 6 feet in diameter; that sediment had deposited in the pipe; 

 that the nominal area was not the true area; that a constant reduc- 

 tion factor should not be used in computing the equivalent water 

 column from the mercury column. 



1 Journal Assoc. Engin. Socs., 12 (1893), p. 219. < Id., p. 516. 



s Journal Assoc. Engin. Socs., 21 (1898), p. 250. » Id., p. 547. 



» Trans. Amer. Boo. Civ. Engin., 40 (1898), p. 471. 



