2 BULLETIN" 86, U. S. DEPARTMENT OF AGRICULTURE. 
composed of four pieces, though two heads were each composed of 
three pieces. The pieces of the head were joined together with 
-^6-inch hickory dowels. There were two dowels per joint, each 
about one-third or one-fourth the length of the joint from its end. 
The head and bilge hoops were If inches by 17 gauge, while the 
quarter hoops were lh inches by 18 gauge. The average thick- 
nesses of hoops used for tension tests (see p. 4) were 0.051 inch 
and 0.061 inch for the 18 and 17 gauge, respectively, while the I T . S. 
standard gauges of these numbers are 0.05 inch and 0.05625 inch. 
The average hoop spacing, dimensions, weights, and capacities of 
the barrels are shown on figure 1. The hoop splices were always 
placed over the bung stave, and the heads were placed with their 
end grain toward this stave as shown in I and II, figure 1. 
The barrels were received at the laboratory on November 24, and 
were stored in a closed and unheated shed until the tests were begun 
on December 10. 
BARREL TESTS. 
The barrels were brought in from the storage shed shortly before 
the time for test. Each barrel was then carefully inspected and the 
hoops driven tight by a representative of the St. Louis Cooperage 
Co. Just before test each barrel was completely filled with water, and 
with the exception of those barrels to which a pressure gauge was 
attached, was closed with a wooden bung. These bungs, after soaking 
for a few seconds in warm water, were driven to a tight fit. They 
were placed with their grain parallel to that of the staves. The bungs 
bore the brand "U. S. Bung Mfg. Co., Cincinnati, O." No bung 
straps were used. 
Two barrels of each group were tested in side compression, two in 
diagonal compression, one each in side and diagonal drop, and two 
by internal pressure. 
SIDE-COMPRESSION TESTS. 
In this test the barrel was placed between two flat surfaces and 
compressed in the direction of its diameter. The rate of compression 
was 0.25 inch per minute. Simultaneous readings of load, com- 
pression, and loss of water from the barrel were taken. The test 
was discontinued when one-half the water had escaped. Notes were 
made of the character and sequence of failures. In about one-half 
of these tests a pressure gauge was attached to the barrel, and read- 
ings of internal pressure were taken. The method of test is shown 
hi Plate I. 
DIAGONAL-COMPRESSION TEST. 
In this test the barrel was compressed between two flat surfaces, 
being supported upon one point of the chime and loaded at a point 
on the other and diagonally opposite. The rate of compression was 
0.25 inch per minute. Notes were taken as in the side compression 
test, The test on the first barrel of each group was discontinued as 
