Chemical Affinity in terms of Electromotive Force. 289 



about to be dipped before immersion therein. Operating in 

 this way, the difference between the values obtained before 

 and after exchanging represents double the sum of the poten- 

 tial difference that would be set up between the plates when 

 opposed to one another in the same solutions respectively. It 

 is noteworthy that the figures thus got by exchanging plates 

 are of the same order of magnitude as those resulting from 

 the opposition of two plates of the same kind. Thus in the 

 above-described experiments the maximum variation observed 

 with amalgamated zinc plates is 1*55 millivolt, and that with 

 electro-copper plates '66 millivolt, giving a sum of 2*21; 

 whilst the corresponding average values give a sum of less 

 than '5 millivolt. With two opposed Daniells set up with 

 amalgamated zinc and electro-copper plates, the halves of the 

 differences observed on exchanging plates averaged less than 

 •5 millivolt, sometimes, however, amounting to 1*0, and occa- 

 sionally, but only rarely, to 2*0 millivolts; and similarly in 

 other cases. 



150. In order to eliminate errors arising from depreciation 

 ofE.M.F. through non-adjuvancy, arising from the current- 

 density being greater than the limiting-density below which 

 such effects are inappreciable (" polarization "), several series 

 of observations were made in order to find out what the 

 limiting-densities are for different kinds of plates and fluids, 

 such that the depreciation in E.M.F. shall not exceed a 

 definite small quantity, say *1 to *2 millivolt. The experi- 

 ments already published made with single Daniell and other 

 analogous cells (§§ 103-105, Part V.) show that with current- 

 densities of from 10 to 20 microamperes per square centi- 

 metre of plate-surface (both plates being equal in area, and 

 one only being reckoned), diminutions in E.M.F. of from 

 1 to 6 millivolts may be brought about, quantities in many 

 cases far exceeding the potential-differences due to varying 

 solution-strength to be measured. Manifestly, therefore, only 

 very minute currents can be employed when two Daniell 

 or other analogous cells are opposed, and consequently the 

 effect of " polarization ?' about doubled. Similarly the ex- 

 periments described in Part IV. (§ 93) indicate that when 

 a current passes between two zinc plates in zinc-sulphate 

 solution, or between two copper plates m copper-sulphate 

 solution, a " counter electromotive force " of a considerable 

 number of millivolts, and even some centivolts, may be set up 

 with current-densities of but a few microamperes per square 

 centimetre; so that in a " diffusion-cell " serious diminutions 

 in the potential-difference actually set up may be brought 

 about by very small current-densities. 



