410 



burst itself in the effort to expand. The character of the 

 failure is shown in the second figure (fig. 2). Another similar 

 block of insignisj 2| in. x 3| in. in area, treated in the 

 same way, gave a maximum load of 1,460 lbs. This also failed 

 in compression. The average maximum pressure exerted by 

 the two blocks was 139 lbs. per sq. in. A block of Pinus 

 canariensis was dealt with in the same way, and for three days 

 it was left in the testing machine, and gave a pressure gradu- 

 ally increasing up to 128 lbs. to the sq. in., when it had to be 

 renioved to make way for other tests. This block showed no 

 sign of failure. 



When the timber was first received small cylinders about 

 3 in. in diameter and 1 in. long were accurately turned out 

 of the green wood from blocks whose moisture contents had 

 just been determined. These were then weighed and accurately 



'////////A 



w///MW/7ft///?m//. 



Fie. 2. 



measured along marked diameters in directions parallel and 

 perpendicular to the rings. The average measurements in 

 August, 1917, at the end of March, 1918, and in October, 

 1919, are shown in Table I. It will be seen that by the end of 

 March, 1918, these small pieces had lost all the moisture 

 they were free to lose and at that time of the year showed a 

 percentage of only 7 to 9 per cent. The contraction in the 

 direction parallel to the rings was in all cases greater than that 

 in the perpendicular direction, and was most for Pinus 

 maritima and least for insignis. For Pinus maritima it 

 amounted to 4*8 per cent., which is less than half the contrac- 

 tion that might be expected from a Eucalypt with the same 

 initial quality of moisture. It will be noticed that when meas- 

 ured again in October, this year, the blocks all showed an 



