312 
TIMBEE 
spruce, oak, and birch timber, 4 inches square. Between 
1893 and 1895 the moduH of breaking and elasticity had 
increased while deflection had diminished — the wood had 
become stiffer — it had during this period been stored in 
dry lofts of the building. In 1901 further tests made on 
similar timber which during those eight years had been 
seasoning, and which was further dried for ten or twelve 
days at 100° Fahr., showed that the modulus of breaking 
under compression, bending, and shearing stresses had all 
increased, as well as that of elasticity, and that that of 
deflection, which had diminished from 1893 to 1895, was 
on the increase, but not in any particular relation to the 
lapse of time.^ 
Doubtless one reason why dry timber shows higher 
tests than wet timber is that the timber shrinks in drying 
and its volume is diminished, in the case of pines and 
spruces by about 10 per cent, and some timbers considerably 
more, but the numbers of fibres in the wood resisting 
strain remain the same ; its cross section, too, is smaller, 
although the result is generally calculated upon the original 
section ; this, however, would only account for a very small 
portion of the increase. The remarks as to the gradually 
increasing strength of wood as it dries point to the con- 
clusion that beams and joists in buildings are capable of 
bearing safely a heavier load some years after erection than 
when originally put up. 
Timber columns are fairly uniform in tests up to, say, 
15 diameters long, and up to this point give way by direct 
crushing ; in longer columns the larger proportion fail by 
lateral flexure or " buckling " sideways, and generally, as 
was the case in Lanza's tests, fail at knots. In the West 
Austrahan tests, with columns of a ratio of 18 to 1, 60 
per cent, failed by side flexure. 
' Mill, of Proc. Inst. C.E., Vol. CLVII., p. 452. 
