8 BULLETIN 854, U. S. DEPAKTMEITT OF AGRICULTURE. 
either side of the tile in the flume was prevented. After the tile 
were covered, the remaining space in the flume was filled with earth. 
PIEZOMETERS AND PIEZOMETER TILES. 
In order to measure the depth of flow in the tile drain, piezometer 
tubes of graduated glass were placed on the side of the flume and 
connected to the lower part of the tile line. Twelve tiles of uniform 
shape, for each size and kind, were selected, and a small hole was 
drilled through the wall of each. In each hole a J-inch iron pipe, 
2 inches long, was inserted, care being taken that the tube did not 
project inside the tile bore. This tube was set in cement (PL 11, 
fig. 2), and any unevenness on the inside wall of the tile at the en- 
trance of the tube was removed by coating the surface with a little 
cement. This method of inserting the tube was deemed the best as 
determined by Hiram F. Mills from a study of the results of some 
6,000 observations on various piezometer connections (see Trans. 
Amer. Academy of Science, 1878). Mills found that with an orifice 
whose edges are in the plane of the side of the conduit and with the 
bore of the tube normal to the plane of the waU, the piezometer 
column indicates the true height of the water surface in any open 
conduit, or the pressure in a closed conduit. 
At first these piezometer tile were so turned as to have the tube 
on the bottom of the tile in the flume. Much trouble was experienced 
from the tube openings filling up, so the piezometer tile were then 
laid with the tube leading toward the side of the flume but turned 
slightly downward. The connection was made by rubber tubing to 
a steel nipple inserted through the wall of the flume (PI. VII). On 
the side of the flume at each piezometer tile, a frame holding the glass 
tube was set. This glass tube (PI. VII) was graduated in tenths and 
hundredths of a foot. Its zero was set at a definite distance below 
the top of the flum.e. A rubber tube connected the piezometer glass 
to the nipple in the wall of the flume. 
The zero of each piezometer gage was 17 J inches below the top of 
the flume. The invert of the tile in the flume was always laid 16| 
inches below the top of the flume. The capillarity of the glass tubes 
used was found to be 0.01 foot. Thus, with water just entering the 
tile drain, the piezometer tube read 0.09 foot. In other words, in 
order to obtain the true depth of flow in the drain, 0.09 foot was 
subtracted from each piezometer reading. 
With the exception of the two piezometers near the tile entrance, 
which were only 8 feet apart, these tubes were distributed along the 
flume approximately 55 feet apart, the last piezometer being within 
a few feet of the outlet of the tile drain» 
I 
