1884.] of Conductors through which a Current is passing. 187 



In the somewliat crude state of my preliminary experiments I have 

 not considered it necessary in calculating to make any allowance for 

 heat lost at the ends of my wires by conduction to the masses of metal 

 to which they were soldered. The following appended note, however, 

 contains an investigation of the correction which, in future experi- 

 ments, it will be necessary to apply. 



Note by Sir W. Thomson. 



To estimate effect of conduction of heat from the tested wire 

 through its ends: suppose the ends to be kept at the atmospheric 

 temperature. This supposition corresponds to the greatest possible 

 degree of the effect in question. (Adopting Fourier's notation, in 

 the first four symbols), let v be the excess of temperature in the wire, 

 at distance x from one end : 



a, the length of the wire ; 

 k, its thermal conductivity ; 

 h, the thermal emissivity of its surface ; 

 g, the girth of its cross section ; 

 A, the area of its cross section ; 



s, the specific electric resistance of its substance (C.G.S.) ; 

 ^o Ac, the strength of the current (C.G.S.) ; 



(or c, the current in amperes per square centimetre of cross- 



section) ; 



and J, Joule's equivalent multiplied by the force of gravity at 



Manchester (4'2344 X 981'34=41'55 centimetre-dynes), the 

 dynamical equivalent of the thermal unit (C.G.S.), 



we have by Fourier and Joule : 



* Temperature probably much too low. The wire, sagging down, touched the 

 glass tube in several points. 



