Chemical Affinity in terms of Electromotive Force. 201 



average value of a normal Daniell cell with fresh plates by an 

 amount sensibly equal to the sum of the numerical alterations 

 that had occurred in the zinc-silver and copper-silver cells 

 jointly. It is specially noticeable that, whilst in zinc-copper 

 cells the alteration in the surface of the copper (probably 

 through oxidation) on standing diminishes the E.M.F., in 

 copper-silver cells the alteration is in the opposite direction : 

 with zinc and cadmium the direction of this alteration when 

 opposed to silver is the same as when opposed to copper. 



It would seem from all these results that the effect of a given 

 alteration of the surface of one of the plates of a voltaic pair 

 upon the E.M.F. of the pair is independent of the nature of the 

 other plate as regards its numerical value, although the nature 

 of this second plate regulates the direction of the variation in 

 the E.M.F. produced (increase or decrease), and also exerts 

 an influence upon the rate at which the alteration of the plate- 

 surface takes place. Thus it was repeatedly observed that, 

 whereas an amalgamated zinc plate (or an electro-copper one), 

 when forming part of a normal Daniell cell, did not become 

 sensibly oxidized, so as to diminish the E.M.F. of the cell, 

 until after several hours at least had elapsed, a precisely similar 

 plate immersed in the same liquid, but forming part of a zinc- 

 silver cell (or of a copper-silver cell), did become perceptibly 

 oxidized in much less time. In other words, although no 

 measurable current was generated in either case, yet the 

 different amounts of strain (so to speak) set up in the chain 

 of liquid particles between the two plates, according as one 

 was silver or not, did affect the rate of change in the surface 

 of the more oxidizable metal (presumably by varying the 

 rate at which it combined with the oxygen dissolved in the 

 fluid). 



127. The following Table exhibits the average results ob- 

 tained in about 150 sets of observations and comparisons, only 

 those values made during the first hour (or sometimes less) 

 after setting up being taken into account, and all subsequent 

 values being rejected where any diminution through oxidation 

 &c. began to be perceptible *. The values cited as the average 

 effects of substituting for bright silver electro- and amalga- 

 mated (liquid) silver are the means of the three sets above 

 quoted (§ 125) obtained respectively with zinc, cadmium, and 

 copper : — 



* Notwithstanding all care, it is probable that the average results with 

 the zinc-silver cells are too low by a few thousandths of a volt, and those 

 with the other cells slightly too high (vide § 128). 



