TESTS OF WOODEN BABRELS. 6 
in the preceding test, while the second was discontinued whenever 
one-half the contents had escaped or would have escaped had the 
barrel been in the reverse position. This test is illustrated in 
Plate II. 
SIDE-DROP TEST. 
In this test the barrels were dropped on a wooden platform about 
3 J inches thick resting on the concrete floor of the laboratory. On 
top of this platform was a steel plate one -eighth inch in thickness. 
The barrel was suspended with its axis horizontal. The first drop 
was 3 inches, the next 6 inches, etc., increasing each time by 3 inches. 
Each drop was upon the same point of the barrel. After the first 
apparent leak the drops were made at 3-minute intervals. The 
weight of the barrel and contents was taken immediately before 
each drop. The test was continued until half the contents of the 
barrel had escaped. Complete notes were made to show the character 
and sequence of the failures. A picture of this test is shown in 
Plate III. 
DIAGONAL-DROP TEST. 
This test was conducted in the manner described for the side-drop 
test, except that the barrel was suspended so that the lowest point of 
the chime was directly below the center of the barrel, which was 
dropped on the chime. Each drop was upon the same point. A 
picture of this test is shown in Plate IV. 
INTERNAL-PRESSURE TEST. 
In this test the barrel and connecting pipes were filled with water in 
such a way as to exclude as nearly as possible all air. The pressure 
was then raised to 2 pounds per square inch and held for 2 minutes. 
It was then raised to 4 pounds and there held for 2 minutes. This 
was repeated, increasing the pressure 2 pounds each time and holding 
it constant for 2 minutes after each increase, until 1 pound of water 
ran from the barrel in 1 minute or less. The test was then discon- 
tinued. Complete notes were made as to the character and sequence 
of the failures. 
In these tests connection to the barrel was made by screwing a 
special tapered bush into the bunghole. The apparatus is shown in 
Plate V. 
MINOR TESTS. 
STAVE TESTS. 
In order to find out something of the variability of the barrel 
material tests were made on 36 staves, two from each of 6 barrels of 
each thickness. The best and poorest appearing stave of each barrel 
was chosen. Pieces 2 inches in width, cut from these staves, were 
tested in static bending under center loading. The span was 28 
inches. The staves were placed with the outer side up. 
