Page:Timber and Timber Trees, Native and Foreign.djvu/72

52 weight varied from 500 to 1,002 lbs., and from 753 to 1,005 lbs., the average results of one parcel being 193.96 lbs., and of the other 209.21 lbs. on the square inch, the difference between the two being only 15-25 lbs.

Taking the mean of the figures in the Tables V. and VI., we have a deflection of 2-489 inches with 390 lbs.; but only .151 inch after the weight had been removed, the ultimate deflection at the time of breaking being 7.166 inches, the breaking weight, 806.83 lbs., and the specific gravity 810. By applying the formulæ used by Professor Barlow, viz., $$\frac{l^3w}{16{ad}^3\delta}$$ and $$\frac{lw}{4ad^2}$$ respectively for elasticity and strength, we obtain the following result: E = 386400, S = 2116.

It should be borne in mind that in determining the scantlings to be employed, there are to be taken into account the possible chance of a short or twisted grain, a spiral turn of the fibre, knots, faulty or otherwise, and the risk which the practical builder must always run of having some defects hidden beyond the possibility of detection in, perhaps, his best-looking pieces. It would, therefore be obviously unsafe to subject them to anything like the strain which the ascertained average strength of the specimens tested would seem to warrant charging them with.

Considering the importance of this, it was determined to extend the experiments by testing a series of pieces taken from a longitudinal section cut through the centre breadth of a very fine-looking Oak tree. In setting out the specimens, the centre piece containing the pith and a very small heart-shake, was allowed to drop out as being of little or no value for the trial.

The six pieces cut from one side of the centre or pith