Chemical Affinity in terms of Electromotive Force. 287 



found to exceed the latter, sometimes by a very considerable 

 amount. 



149. In carrying out observations by means of either the 

 method of opposition or that of diffusion-cells, it is obvious 

 that the current observed produced through a given total re- 

 sistance only represents the B.M.F. due to difference of solu- 

 tion-strength, provided that the two zinc plates and the two 

 copper plates are identical in surface-characters, and provided 

 also that the current-density is small enough to reduce polari- 

 zation to negligible amounts. In practice it is impossible to 

 ensure equality of surface with any pair of plates, even though 

 prepared with the greatest care in precisely the same way 

 from the same mass ; and the result is that when two plates, 

 apparently identical, are immersed in the same solution, one 

 is sufficiently electropositive with reference to the other to 

 cause a distinctly measurable current to flow through even a 

 large external resistance. In order to form some estimate as 

 to the probable magnitude of this source of error, a large 

 number of observations were made with pairs of zinc plates 

 immersed in zinc-sulphate solutions of varying strengths, and 

 pairs of copper plates immersed in copper-sulphate solutions. 

 The following results in millivolts were obtained with upwards 

 of 120 pairs of plates, readings being taken every few minutes 

 for about half an hour after the first immersion, and being then 

 averaged. As a rule the currents observed increased somewhat 

 as time elapsed, so that the first readings were usually the 

 minima. The galvanometer-readings were translated into volts 

 by means of the formula 



m Ro D r> rn 



e= n'Rri000 =m:R2 - C ' 



where m = the galvanometer-reading to be translated (no 

 shunt being used). 

 n = the galvanometer-reading when a Daniell cell of 

 E.M.F. D is used to send a current through 

 a total resistance B^. ohms, employing yoqo 

 shunt. 

 R 2 = total resistance in circuit when the reading m is 



taken. -pv 



C = current representing 1 scale-division = .,, ^ _ 

 _ amperes. * _ _ 1000mR i 



If the potential differences are expressed in millivolts, then 



m R 2 -r* 

 n xii 



