THE FLOW OF WATER IN WOOD-STAVE PIPE. 17 



Bureau of Standards the year previous, but had been used only a few 

 times, and then as a standard. Its rating curve was checked by the 

 writer and his assistant in the channel of the hydrauhc laboratory at 

 Cornell University, just prior to these tests. No change was found 

 necessary. 



Fluorescein} — ^Direct measurements of velocity of water in a pipe 

 were made by injecting solutions of fluorescein and timing the passage 

 of the resultant green-colored water through the reach tested. 



Weirs. — Where weirs had been installed to measure the quantity 

 of water from pipes the writer made use of them. Each is described 

 in the report of the test in which it was used. 



Hook gauge. — ^A small hook gauge of the Boyden type, with vernier 

 reading to thousandths of a foot, was used to determine surface 

 fluctuations for head on a weir. (See PI. 1, fig. 1.) 



Piezometers. — ^Two types were used: 



Water column: This was employed where the pressure in the pipe 

 was low. A simple glass manometer-tube, engine-divided to tenths 

 and hundredths of a foot, was connected to the tap in the wood pipe 

 by a piece of rubber pressure tubing. (See PI. II, fig. 1.) 



Mercury manometer: Where otherwise the pressure would have 

 compelled the use of a long water column, a mercury manometer of 

 the U-tube pattern was selected. Two of these U-tube mercury 

 gauges, as shown in figure 3, were provided. They consisted of 

 wrought-iron U tubes with unequal legs. The short leg was a glass 

 manometer tube surmounted by a tee connection provided with the 

 necessary cocks for manipulation. The long leg was formed of 2-foot 

 imits of one-eighth inch wrought-iron pipe until a length had been 

 attained which permitted the top of the high mercury column to 

 show in another glass tube. 



These glass tubes were engine-divided into tenths and htmdredths 

 of a foot, the tenths and half-tenths fines extending completely 

 around the tube and the other fines but half way around. By sight- 

 ing through the tube across the front and back of any one line, all 

 tendency toward parallax was removed and the mercury column 

 could be correctly read to thousandths of a foot. The relative vertical 

 positions of the two sets of graduations were of immaterial conse- 

 quence as the graduations on each gauge glass were brought into the 

 general scheme of levels above an assumed datum, as shown in 

 figure 2. 



AU abutting pipes screwed against fiber gaskets. The ends of the 

 gauge glasses were permanently set into sleeved coupfings with seafing 



1 The writer is indebted to R. B. Dole, of the U. S. Geological Survey, for suggesting the use of this won- 

 derful coloring matter. See"Useof Fluorescein in the study of underground waters," R. B. Dole, Water 

 Supply Paper No. 160, U. S. Geological Survey. 



42463°—Bull. 376—16 2 



