508 The Philippine Journal of Science 1921 
safety; namely, 5 for extreme fiber stress in bending, 4 for 
shear along the grain, 8 for compression along the grain, and 
2 for compression across the grain. For steady loads these 
factors become 34, 2%, 2, and 14, respectively. The variability 
of the factors given by different authorities indicates that the 
factor of safety for timber is largely a matter of guesswork on 
account of lack of confidence in the reliability of values of 
strength upon which designing is based. An exhaustive series 
of tests on the mechanical properties of Philippine timbers will 
some day give very reliable data upon which to base the 
design of timber structures. The United States Forest Service ** 
seems to have solved this problem for timbers grown in the 
United States and recommends the following factors of safety: 
5 for modulus of rupture, 3 for compression parallel to grain, 
1.5 for compression perpendicular to grain, 8 for shearing, and 
5 for tension perpendicular to grain. 
In applying factors of safety to strength values for timbers 
it is important to know the moisture of the wood tested, and 
whether the tests were made only on small, thoroughly seasoned 
specimens free from defects or on timber of structural sizes. 
This information is essential, because the results of the tests 
show that the stresses developed in large timbers were less 
than the stresses developed in small specimens. The ratio for 
modulus of rupture varies from 60 to 90 per cent. This differ- 
ence in values was due to the facts that the small specimens 
were free from defects and that they contained less moisture. 
In conclusion I will say that the character and location of 
defects in timber have much to do with its strength. Checks 
in beams weaken their resistance to horizontal shear, especially 
if the defects are found near the neutral plane. The tests have 
shown that knots occurring in the central lower part of a beam 
weaken the timber much more than similar defects found in 
another part. 
APPENDIX 
FORMUL& * USED IN COMPUTATIONS 
LEGEND 
A = Area of cross section, square inches. 
B = Area under plate, square inches. 
CS = Crushing strength, pounds per square inch. 
* Bull. U. S. Forest Service 556 (1912). : 
Instructions to Engineers of Timber Tests, Circular (revised) U. 5S. 
Forest Service 38 (1909). 
