26 BULLETIN 376, TJ. S. DEPARTMENT OF AGRICULTURE. 



OFFICE EQUIPMENT AND METHODS. 



Original nmltiplication, division, and addition were performed on 

 mechanical devices. Checking was done by 20-inch shde rules and 

 graphic methods. AH percentage comparisons were made on 20-inch 

 shde rules. Estimate diagrams were checked by proving random 

 examples. 



Office procedure. — Where water columns were used at both ends of 

 the reach of pipe tested the loss of head in the pipe for the given 

 velocity was the difference in elevation between the top of the mean 

 water column at gauge No. 1 and the top of the mean column at 

 gauge No. 2. Where a mercury manometer was used at one or both 

 of the gauges the equivalent water column for each reading of the 

 merciu-y colimm was computed. The mean of the elevations of the 

 tops of the equivalent water columns was accepted as the elevation 

 for that gauge. The loss of head was then computed as before. 

 Standard methods were employed in computing current meter data 

 or weir discharge. Where color was used in timing the velocity of 

 the water the time was computed as from the instant of injection to 

 the mean between first sight and last appearance of the color at the 

 outlet. 



ELEMENTS OF FIELD TESTS TO DETERMmE FRICTION LOSSES AND 

 COMPARISON OF OBSERVED VELOCITIES WITH VELOCITIES COM- 

 PUTED FROM VARIOUS FORMULAS. 



In the following pages two tables are arranged (Tables 2 and 3). 

 Table 2 gives the elements of nearly all known observations on wood 

 pipes, either round or square. The various series are arranged in 

 ascending sizes of pipe and within one series the observations are 

 arranged in ascending order of velocities. 



The tests of one experimenter are omitted from these tables as 

 extraordinary friction values were found. The writer made an inde- 

 pendent set of tests upon some of the same pipes and found them so 

 choked with ravelings from the rock cuts above the siphons that erro- 

 neous values were obtained. In the omitted tests the error lay m 

 making current meter measurements for Q and then accepting 



V = ^ where A was taken as the nominal area of the pipe when as a 



matter of fact the true value of A was about 90 per cent of the nominal 

 A; therefore the true velocity was much higher than that found by 

 the erroneous assumption of A. 



EXPLANATORY NOTES ON TABLE 2. 



Column 1 gives the consecutive numbers of the pipes as followed in column 1, Table 

 3, also in the discussions in the following pages and in the appendix. The small letter 

 a after the numbers refers to discussion in the appendix. Experiments conducted by 

 this department are discussed in the text while the essential data secured from other 

 soiirces are abstracted in the appendix. 



