accompany the changes in volume of Solid Bodies. 331 



a 1 a 2 



For this purpose it was at first placed above the axis b, 

 and then made to slide towards the end p. This motion was 

 easily effected by the aid „. 



of a wire, as the piece c 

 rested on the lever by a 

 roller e, and the upper sur- 

 face of the lever was care- 

 fully polished. The an- 

 gular displacement of the 

 lever could be estimated 

 by a mirror g, in which 

 the image of a divided ver- 

 tical scale was observed by 

 reflexion. The scale was 

 divided into millimetres; 

 each division corresponded 

 to an angular displacement 

 of the lever of 47"*6, and 

 an elongation of the wire 

 of 0-0114 millim. As the 

 total length of the wire was 

 about 0*590 metre, the ac- 

 curacy of the experiments 

 was sufficiently great. 

 Lastly, to exclude as much as possible accidental changes of 

 temperature, and to retain its temperature uniform throughout 

 its entire length, M. Edlund arranged a glass case, B C (repre- 

 sented open in the figure), lined inside with tinfoil. 



In making an experiment, the weight d was rapidly passed from 

 the axis b to the end of the lever ; the angular displacement was 

 noted, as well as the first range of the galvanometer needle. 

 M. Edlund took this range as a measure of the diminution of 

 the temperature of the wire, and established this proportionality 

 by the following reasoning. 



Let x be that division of the scale which at the moment t is 

 brought, by reflexion on the mirror of the galvanometer, to coin- 

 cide with the vertical wire of the telescope ; x the value of oc 

 corresponding to the time / = 0at which no current passes in the 

 wire of the galvanometer; m the quotient of the magnetic mo- 

 ment of the needle by its moment of inertia; 2n a constant 

 which depends on the induced currents developed by the motion 

 of the needle in the conducting parts of the apparatus, and which 

 is inversely as the moment of inertia of the needle ; k another 

 constant directly proportional to the electromotive force of the 

 couple employed, and inversely proportional to the resistance of 

 the circuit, as well as to the moment of inertia of the needle ; 



Z 2 



