TESTS OF WOODEN" BARRELS. 7 
DROP TESTS. 
The broken heads of the made-up barrels were replaced by other 
J-inch heads, and the barrels subjected to drop tests. The barrel 
dropped upon the side showed much better resistance than the 8-hoop 
barrel five-eighths inch in thickness throughout, but was not quite 
the equal of the f-inch barrels. 
In dropping on the chime the made-up barrel was the equal of any 
barrel tested. 
These tests of made-up barrels seem to justify the previously sug- 
gested changes in thickness of head and spacing of hoops. 
The detailed results of these tests are given in Table 5. 
SUGGESTIONS REGARDING TESTS OF SHIPPING CONTAINERS. 
There are two classes of tests to which containers such as barrels 
may be subjected: 
First. Tests, such as the ones described in this bulletin, where the 
object is to determine the most economical and efficient designs. 
Tests of this class are usually carried to the destruction of the con- 
tainer and entail damage or complete loss of contents. It is neces- 
sary to fill the containers with material which is relatively inexpensive, 
safe to the investigators, and which will produce stresses similar in 
character to those which would be produced by the commodity which 
the container is intended to carry. 
Second. Tests to determine the suitability of the container for 
specified commodities under practical conditions. Such tests should 
be made upon containers filled with the material to be shipped in 
them or with some other very similar in its action on the container. 
In the case of the first class of tests seepage through the pores and 
the first leak depend largely upon the nature of the lining and of the 
contained liquid. A material difference might be expected in the 
behavior of barrels fined with paraffin and filled with water as com- 
pared with barrels lined with glue and filled with gasoline. In the 
drop test the height of drop also depends upon the specific gravity of 
the contained liquid. The height of drop required to produce given 
stresses is in approximately inverse proportion to the combined weight 
of barrel and contents. 
Having made tests of the first class, and so determined the best 
construction, it then remains to manufacture containers in accordance 
with specifications based upon the results of these tests. Tests of 
the second class made upon such containers lined according to com- 
mercial practice and filled with the commodity they are to carry 
would show their limitations under practical conditions. 
In the case of barrels internal-pressure and side-drop tests are 
recommended for this purpose. 
