426 



REPORT 1887. 



inches in depth, and -was supported upon two piers 20 feet apart in the 

 clear. The top flange consisted of a plate 4* inches X i inch and two 

 angle irons 2 inches X 2 inches X j\r inch, giving a sectional area of 

 4"30 square inches. The bottom flange consisted of a plate 4 inches 

 X ^ inch, and two angle irons 2 inches X 2 inches X y^^r inch, 

 giving a sectional area of 2'40 square inches, or 1-775 square inches, 

 when the necessary deduction is made for the rivet holes. The statical 

 breaking strength of the girder does not appear to have been accurately 

 known, but it was estimated at 12 tons in the centre, that of the iron being 

 taken at 22-6 tons per square inch. By the revolution of a crank driven 

 by belting a given load was alternately allowed to rest on the centre of 

 the girder, and was lifted oS" again a great number of times in succession. 

 The following table shows the number of applications of the different 

 loads, the calculated stresses produced in the bottom flange, and the general 

 results. 



These experiments show that the girder was apparently able to bear 

 any number of applications of from 3 tons to 3| tons at the centre, pro- 

 ducing a stress from 5 '67 tons to 6'62 tons in the bottom flange, without 

 any signs of failure or of decreasing strength, but that the greater loads 

 caused fracture after a certain number of applications. 



In a paper by Sir William Thomson, entitled ' On the Elasticity 

 and Viscosity of Metals,' published in the Proceedings of the Royal 

 Society for May 18, 1865, an account is given of experimental researches 

 instituted by him and conducted in his laboratory in the University of 

 Glasgow, through which some new and previously unsuspected properties 

 in the elasticity of metals were discovered. These cannot be fully de- 

 scribed here in detail, but it may be mentioned that the new results, of 

 greatest interest and probably of greatest practical importance, related to 

 tempoi'ary and gradually subsiding effects left in wires by previous elastic 

 oscillations. Energy was expended (dissipated) much more in any one 

 torsional oscillation of a wire which had for some time previously been 

 kept actively oscillating, than in a like oscillation either of the same or 



' The beam broke by tension, the bottom flange failing near the centre ; the 

 fracture having been repaired, the experiments were continued. 



2 The bottom flange broke under tension close to the plate riveted over the 

 previous fracture. 



