TIMBER: ITS STRENGTH, AND HOW TO TEST IT. 5 
point of view of forestry, is almost entirely lost. It is only by 
such elaborate records that those responsible for the upkeep 
and the maintenance of State forests can obtain data upon which 
they can work in the management and treatment of the trees 
in the forest, and of course it is equally important to the 
engineer and to the architect to know, when he has the data 
of the tests before him, what was the previous history of the 
timber from which the specimens were prepared. 
As in Bauschinger’s tests, so also in those of the American 
Board of Agriculture, careful determinations were made of the 
moisture conditions of the tested bars. The plan adopted by 
the American observers was to cut a thin disc across the whole 
section of the tested bar, as close up to the point of fracture as 
possible; these thin discs were then at once weighed, and 
immediately placed for a number of hours in a warm current of 
air, maintained at a temperature of 220° F.; they were then 
re-weighed when perfectly dry, and the difference between the 
two weighings gave the total moisture present, and the ratio of 
this difference to the dry weight was the percentage of moisture 
present. 
A reference again to Fig. 1, which shows this law of 
variation of strength with variation of moisture, brings out 
clearly the interesting fact that the maximum strength is not 
reached when the timber is perfectly dry, but when there is 
about 3 per cent. or 4 per cent. of moisture present in it. It is 
rather a difficult piece of experimental work, however, to 
determine this point very accurately, because, as I have already 
mentioned, when timber is dried below ro per cent. of moisture, 
it tends to reabsorb moisture from the air extraordinarily rapidly. 
In connection with this question of the influence of moisture 
upon the strength, the American experimenters investigated 
whether or not moisture reabsorbed into a specimen which had 
been dried had the same weakening effect. as the original sap of 
the tree, and this was found to be the case. 
The two curves in Fig. 2 represent (1) a series of tests with 
timber which was dried by gradually driving off the natural sap, 
and (2) a series of tests with very dry timber which was 
gradually allowed to reabsorb moisture, and you will notice 
that the two curves practically coincide. This is a matter of 
considerable practical importance in the case of timber used in 
underground situations, or in water itself, or in a water-logged 
