378 Bath Ulerary and Philosophicai Society. 



The first experiment was made on a square piece of Dantzic 

 oak cut into twenty-five pieces, each piece five feet long and 

 two inches square. The mean weight to produce fracture was 

 1674 lbs. ; the variation being from 114 lbs. to 214 lbs. The 

 section which included the heart of the tree bore 154 lbs., and 

 the strength of the other sections diminished in proportion ta 

 their distance from the heart. The angles of curvature did not 

 correspond to the weights employed ; in one piece which bore 

 196 lbs. just before fracture, the angle of curvature was 6" 12'; 

 another which supported nearly the same weight, viz. 193 lbs., 

 only bent to an angle of 3° 25' ; while a third piece which broke 

 with considerably less weight formed an angle of 70' 12'. — 

 These variations in elasticity appeared to be determined by 

 the central relationship of the respective sections of wood ; the 

 heart section being the least, and the outer section the most 

 elastic. 



It appeared further from Colonel Beaufoy's experiments, that 

 a piece of wood divided into any number of pieces loses strength 

 nearly in tlie proj)ortion of the square root of the number of 

 pieces : thus the original piece of Daiitzic oak here employed, 

 would have supported five times more weight than the sum of 

 the weights of the twenty-five pieces into which it had been di- 

 vided. 



Experiments were made on timber spliced three different ways, 

 and on an average the splicing diminishes the strength nearly 

 one-half. 



On a comparison of four different hinds of wood : viz. pitch- 

 pine, Engli-.h oak, Riga fir, and Dantzic oak. Colonel Beaufoy 

 found the pitch-pine tlic strongest ; and taking tiie mean of its 

 power of sujiport at lOOO, the rehitive sneagth of the other 

 woods mav be stated as follows : 



Pitch pine 1()(!0 I Riga fir .. 7S2 

 English oak 923 | Dantzic oak 663 



The weight of Riga fir being to that of English oak as 659 to 

 1000, whilst it stjrugth is as 846 to 1000 ; it is thciefcre pre- 

 ferable in dry buildings to oak for beams, both on account of 

 weight and expense. 



Additional experiments, it was stated, had been made on this 

 s\ibiect by Mr. G. Smart, carpenter, Westminster Bridge, from 

 which it appeared that there was a remarkable difference in 

 point of su|)port between timbers lying loose at their ends, and 

 timbers carefully coaked down on plates of buildings instead of 

 peing dovetailed, as is often done in carpentry. A lath lying 

 loose at both ends, bent like a hoop and broke with eleven pounds 

 weight, while a similar lath, which was rendered tight at each 

 end by bhouldfere, ^ipported no less than 240 lbs, Mr. Sej)ping, 



in 



