282 EXPEKIMENT STATION EECORD. [Vol.36 



that the application of this formula " meets (within 1 per cent) the mean of all 

 observations and the mean capacity of all wood pipes upon which experiments 

 have been made." 



General conclusions regarding the capacity of wood-stave pipes are drawn 

 as follows : "A very conservative factor of safety should be used whei*e a guar- 

 antied capacity is to be attained. . . . The data now existing do not show 

 that the capacity of wood-stave pipe either increases or decreases with age. 

 This statement does not consider sedimentation, a purely mechanical process. 

 If silted waters are to be conveyed, the pipe should be designed for a worliing 

 velocity of from 5 to 10 ft. per second. If sand is present in the water, the 

 design should be for a velocity of about 5 ft. per second, which will be high 

 enough to carry out a large part of the sand and at the same time not so high 

 as to seriously erode the pipe. The better method is to remove the sand by 

 sumps or other means. Air should be removed from the intalie end of every 

 pipe line, especially when the capacity load is approached. Wood pipe will 

 convey about 15 per cent more water than a 10-year-old cast-iron pipe or a 

 new riveted pipe, and about 25 per cent more than a cast-iron pipe 20 years 

 old or a riveted pipe 10 years old." 



An appendix devoted to abstracts of descriptions of experiments made by 

 agencies other than the Division of Irrigation Investigations of the U. S. 

 Department of Agriculture is included, together with a discussion of the flow 

 of water in wood-stave pipe, by G. S. Williams, T. A. Noble, D. C. Henny, 

 E. A. Moritz, E. W. Schoder, and L. M. Hoskins. 



New method of deriving- weir formulas, T. R. Running {Engin. News, 16 

 (1916), No. 15, pp. 695, 696, figs. 4)- — -A- new method of determining the rate of 

 flow of water over weirs, especially adapted for experiments on the flow of 

 water through pipes, is described as developed at the University of Michigan. 



The main advantage of the method is that it requires only a single experiment 

 for determining the rate of flow of water over a weir for different heads and is 

 .applicable to weirs of high as Avell as of low heads. The apparatus for record- 

 ing the time and head is described, and the derivation of the v/eir formula on 

 the basis of head and time is mathematically presented. 



Design of small lined canals {Jour. Electricity, 37 {1916), No. 15, pp. 279, 

 280, figs. 3). — Curves showing the carrying capacities of small wooden flumes 

 in second feet, miner's inches, and gallons per minute are given. The flumes 

 vary in area of water cross section from 0.5 to 3 sq. ft. and have a fall in 

 feet per 1,000 ft. of from 0.5 to 20. The curves are based on a value of " n " in 

 Kutter's formula of 0.015. " The carrying capacity of small concrete flumes or 

 lined ditches can also be obtained from these curves, as there is usually little 

 difference in the friction factor." 



Surface water supply of Pacific basins in Washington and upper Columbia 

 River Basin, 1913 {U. S. Geol. Survey, Water-Supply Paper 362-A {1916), pp. 

 VI-\-26ft, pis. 2). — This report, prepared in cooiieration with the Slates of Wash- 

 ington. Montana, and Idaho, presents the results of measurements of flow made 

 on the Quinault River, Puget Sound, and upper Colimibia River drainage basins 

 during 1913. 



Rogue River Valley Project and Willamette Valley Investigations, J. T. 

 Whistler and J. H. Lewis {Oreg. Cooper. irorA;, Dept. Int. U. S. Rcclam. Scrv., 

 1916, Feb., pp. Ill, pis. 15). — This report was prepared in cooperation with the 

 State of Oregon. The part covering the Rogue River Valley Project deals with 

 the irrigation and water supply of that portion of the Rogue River Valley above 

 Tolo, and more especialy the area in the immediate vicinity of Ashland and 

 Medford. The average texture of soils of the area is that of a clay loam. 

 •'A duty of 1.5 acre-feet delivered on the land has been adopted. It is based 



