428 MR ANGUS R. FULTON ON EXPERIMENTS TO SHOW 
TENSION TESTs. 
From the results of the tests for elasticity and in compression it was evident that 
no great difference of stress would arise in the tension and shearing tests. 
The specimens in tension were of rectangular section, 1” to 1%” wide by }" to 3%," 
thick, with enlarged ends. 
The average results were as follows :— 
broad side radial 7-05 tons per sq. in. Maximum 7:45 tons. 
Oak > 
» 9 tangential 6°93 ¥ Bs M 1503) %, 
{ >»  Fadial CE es i bs 958 a, 
AsH : 
»  », tangential 8:28 = Neils a, 
ree ie mo untolelll 9°47 5 . A Werle 
» », tangential 8°8 vs 3 % 10:2 ~—C, 
Here a slight balance is shown in favour of the broad side being radial, though in 
the Boxwood tests the maximum value was reached by a specimen cut with its broad 
side tangential. Owing to the careful selection of the timber, the results were very 
uniform. 
SHEARING. 
The test pieces here were cylindrical in shape and of 14 inch diameter. In the 
case of Oak the shear was measured radially, tangentially, transversely, and obliquely. 
Radial. Tangential. Transverse. Oblique. 
Oak, average . ‘48 6 1:04 “8 tons per sq. in. 
Ash . : = are 6 1:03 
Box . : a Sho 8 1°24 
According to these results, there seems to be a greater resistance to shearing along 
the rings than to shearing along the lines of the medullary rays, a result which, how- 
ever, we would naturally expect from the previous experiments. 
CROSS-BREAKING. 
It has been the practice to measure the stress or strength factor in cross-breaking by 
the well-known formula, 
where W =central breaking load, 
L= length of span, 
6 = breadth, and 
d = depth of rectangular section. 
This formula is based on the theory of the elastic bending of beams, and gives a 
value of “/f” for the rupture stress which in timber is too great for the ultimate 
compressive stress, and too small for the ultimate tensile stress of the material. If, 
however, the value of W inserted in the above formula be the central load at the 
limit of proportionality of deflection to load, then the stress so calculated agrees 
