474 On the Controversy concerning Voltas Contact Force. 



charged to the potentials Y ± and V 2 respectively by the 

 differential electric affinities at work 



Now transfer unit charge reversibly from one condenser to 

 the other, l. e. from potential Y 1 to V 2 , by shifting the 

 neutral plates ; the work done, equivalent to loss of energy, 

 has usually been written as 



V 2 — Y 1 = ni2.H-w 1 — w 2? 

 but it ought to be written 



v 2 — Vi =n 01 + Mi + n 12 -u 2 + n 20 . 



The total entropy change is zero, 

 or, writing II for S(II), 



11 = 6(8,-8,), 



w heref ore 



V +u — 6s = constant. 



w — u = j Ods = 6 s — \scl6, 



Y—\sd6 = constant, 

 d6 ""*' 



the result to be proved. 



But there is nothing here more than has been admitted in 

 this paper over and over again. 



If surface boundary forces are supposed zero, if electricity 

 can enter or leave the free surface of a metal without aid or 

 obstruction, then "V^ — Y 2 must represent the only remaining 

 force, viz. that at the metallic junction ; and the only II will 

 be situated there ; but if there are other boundary forces 

 then Vj— Y 2 represents th.e whole E.M.F. of the circuit, and 

 not the localised junction force at all, and II 12 is not necessarily 

 the sole Peltier effect. The metallic junction force may still 

 be measured by II12 ; the combined entropy of the boundary 



or 



But 



so 



or 



or 



