MECHANICAL PEOPERTIES OP WOODS GROWN IN UNITED STATES, 5 
Also, in making comparisons, it is important that the data should 
really be representative of the classes of material which it is proposed 
to compare. For example, it is not just to take the figures derived 
from Rocky Mountain Douglas fir, which is known to be inferior to 
the Pacific coast type, 1 as representative of the coast fir. Nor in 
general can a comparison of species properly be made from results of 
tests on large timbers alone; for in practically all cases the large 
timbers tested have not been selected as representative of the species, 
but have been chosen to determine the effect of defects, the effect of 
preservative treatment, or for the solution of other and similar 
problems. 
Comparisons should not be made with greater refinement than the 
data justify. The change which additional tests would probably 
make in the average values and the probable variation of a given 
stick or lot of material from these average values should be considered. 
Numerical measures of these probable variations are given in Table 3. 
CAUSES OF VARIATIONS IN STRENGTH. 
Variations in strength of timber can be accounted for more accu- 
rately than is usually supposed. In some species there is a difference 
in strength in wood from different positions in the tree, different 
localities of growth, etc. But such variations have been overesti- 
mated, and a knowledge of them is not essential in order to estimate 
with a fair degree of accuracy the properties of a piece of timber. 
Differences in strength are usually due to differences in defects, 
moisture content, or density, or to combinations of these. 
Defects are not considered in this publication. Their effects on 
structural timbers are discussed in Forest Service Bulletin 108; and 
limitations on their size, character, and location are given in the 
grading rules for structural timber which have been recommended 
by the Forest Service. 2 
Differences of moisture content cause considerable variation in the 
strength values of air-dry or partially air-dry material, but have no 
effect as long as all material is thoroughly green. 
One of the principal factors causing differences in strength is 
variable density. As might be expected, the greater the density of a 
given stick or the more wood it has 3 per unit volume, the stronger is 
the stick. 
1 See also "Localities Where Grown," p. 8. 
2 See "Discussion of the Proposed Forest Service Rules for Grading the Strength of Southern Pine 
Structural Timbers," by H. S. Betts, Proceedings of Am. Soc. for Test. Materials, Vol. XV, 1915, p. 368. 
3 Accurate determinations made at the Forest Products Laboratory on seven species of wood, including 
both hardwood and coniferous species, showed a range of only about 4§ per cent in the density of the wood 
substance, or material of which the cell walls are composed. Since the density of wood substance is so 
nearly constant, it may be said that the density or specific gravity of a given piece of wood is a measure 
of the amount of wood substance contained in it. 
