393 



on, after which they became more rapid ; from which Mr. 

 Barlow seems to draw the conclusion, that a piece of timber 

 will safely bear about half the weight which is necessary 

 for producing its ultimate rupture. I should, however, men- 

 tion, that in the building of dwelling-houses, the strength 

 of the beams, and the consequent dimensions, must not be 

 calculated with relation to the breaking weight, but to the 

 degree of flexure of the beams which may be permitted 

 without endangering the ceilings under the beam. 



Pillars with flat ends should have their breaking weights 

 at least four times greater than the weight they are in- 

 tended to bear, as it is not always possible to ascertain 

 that the ends of the beam are properly bedded, or that 

 they will remain so, if the foundation on which they are 

 placed should settle. And as appears from Mr. Hodgkin- 

 son's experiments, a pillar with rounded ends is only capable 

 of bearing one-third the breaking weight which a pillar with 

 flat ends would; and a pillar improperly fixed has its strength 

 reduced at least to that of a pillar with rounded ends, 

 that is, one-third. Therefore, if a pillar with flat ends (and 

 the inferiority of pillars with rounded ends being known, 

 they will seldom be used) have not its breaking weight 

 three times greater than the weight to be borne, any 

 negligence in fixing, or settling of the foundation, might 

 cause the pillar to be broken by less than the weight it 

 was intended to support. 



Mr. Ward concluded by recommending that the members 

 of the Society, especially those interested in mines and 

 maufactures, should preserve the records of such struc- 

 tures, beams, or pillars as might be broken in their works, 

 with the data requisite for calculating the strength, the weight 

 causing fracture, and the situation and the time which the 

 structure had stood before breaking, in order to furnish more 

 precise knowledge on the points to which he alluded. 



