is, 5 Valencia: Commercial Philippine Timbers 501 
of the same tree. This is largely due to the defects they con- 
tained, and to the fact that they are not so homogeneous as are 
manufactured materials such as steel and other metals. The 
mechanical properties of a certain timber of a given variety can, 
therefore, be predicated upon the results obtained in tests of 
timber of that variety in a most general way only, until detailed 
information is obtained concerning the many factors governing 
the mechanical properties of both test timber and commercial 
timber. 
Per cent of average modulus of rupture. 
65 70 75 80 Q5 90 95 100 105 HO H5 /ZO 125 150 135 
Fig. 11. Variability in modulus of rupture; tangile, structural sizes, 16 tests; average 
modulus of rupture, 7,640 pounds per square inch. 
Per cent of average modulus of rupture. 
Fig. 12. Variability in modulus of rupture; apitong, structural sizes; 18 tests; 
average modulus of rupture, 6,820 pounds per square inch. 
Figs. 11, 12, 13, and 14, inclusive, show how the individual 
results of the modulus of rupture vary from the average. The 
cross-hatched areas between the adjacent vertical lines repre- 
sent graphically the percentage of the total number of tests that 
fell within the limits indicated on the scale at the bottom, which 
gives the strength of the pieces in percentages of the average 
strength. The percentage of the number of pieces is expressed 
by the figures at the left. For example; by referring to fig. 11, 
it will be seen that 18 per cent of the tests fell between 110 and 
177655 2 
