234 Oti the strength of Pine and Spruce Timber. 



the lower ones. The fractures were not sufficiently regular to per- 

 mit any comparisons to be drawn as to the relative number of fibres 

 crushed and drawn asunder in the different kinds of wood. 



The results in columns [g) of tables (B) and (C), indicate the 

 weights which were required to bend the specimens through an ad- 

 ditional division on the scale of deflexions. If the deflexions were 

 always as the weights producing them, it is evident that all the differ- 

 ences in columns [g), would, for the same kind of wood, be equal. 

 It is easily seen, however, as might have been anticipated, that after 

 having passed with considerable regularity through the two or three 

 first divisions, the deflexions begin to increase in a much greater ra- 

 tio than the weights. From this cause it is very important that in 

 experiments on the stiffness of timber, small deflexions only should 

 be tried, and that those deflexions should be accurately measured. 



It may be seen that the weights producing deflexions of /^ths of 

 an inch to one foot are less in table (C) than in table (B), the rea- 

 son for which is that the stiffness of the wood was diminished by the 

 first trials. 



A circumstance of a practical nature connected with these exper- 

 iments, deserves to be noticed, as it may possibly have modified the 

 results in a slight degree. It was necessary in suspending the load 

 from the beam to adopt some precaution to prevent the iron band 

 connected with the 23latform, from cutting into the wood, and there- 

 by crippling the timber. The annexed sketch will indicate more 

 clearly than any description, the means that were employed. 



Fig. 2. 



(A), (1) and (2) is a strong iron collar, three inches wide and 

 half an inch thick, embracing the specimen under trial, from which, 

 by means of the hook below, the platform containing the load was 

 suspended : (6) represents a bar of iron 2.75 inches wide, six inches 

 long, and half an inch thick, resting upon the wood underneath the 

 collar. 



