MECHANICAL PROPERTIES OF WOODS GROWN IN UNITED STATES. 3 
SCOPE AND METHOD OF EXPERIMENTS. 
ORIGIN OF DATA. 
The data in this bulletin are based upon about 130,000 tests, 
probably the greatest number ever made in one series upon any ma- 
terial. For this reason, and for others explained later, the data are 
the most thorough and accurate that are available on the mechanical 
properties of American woods. The tests were begun about six years 
ago at the Forest Products Laboratory, which is maintained by the 
United States Forest Service with the cooperation of the University 
of Wisconsin. One hundred and twenty-six species of wood have 
been tested, and it is planned to continue the series until all species 
which are important, or which give promise of becoming so, have 
been included. 
SMALL CLEAR SPECIMENS USED. 
Small clear specimens are used in the tests in order that considera- 
tion of the influence of defects may be eliminated from calculations 
to determine the relation between strength and density, moisture, 
locality of growth, soil conditions, etc. These various relations are 
referred to in the present bulletin, however, only when it is necessary 
in order to render the data thoroughly understandable. The speci- 
mens are 2 by 2 inches in cross section. Bending specimens are 30 
inches long; others shorter, depending on the kind of test. 
SELECTION OF MATERIAL. 
The material for any given species and locality is cut from typical 
trees, usually five in number. These are selected by representatives 
of the Forest Service, careful descriptions being made of each tree and 
of the conditions under which it has grown. As a rule the test 
specimens are taken from the top 4 feet of the 16-foot butt log. The 
number of test specimens from each tree varies from 40 to 120, 
depending on the size of the tree. Eventually each important species 
will be represented by tests from at least five typical trees from each 
of several localities distributed throughout its range of growth. 
OTHER DATA INCLUDED. 
Data derived from tests previously made by the Forest Service and 
under practically the same conditions as the present series are in- 
cluded in Tables 1 and 2. The tests were made at Purdue Univer- 
sity and at the Universities of Colorado, California, and Washington 
in cooperation with those institutions. 
TESTS ON LARGE TIMBERS. 
A large number of tests have also been made by the Forest Service 
on full-sized timbers, such as bridge stringers, factory-building tim- 
bers, and car sills. These tests have demonstrated the influence of 
defects such as knots, shakes, and checks on strength, and they serve 
