1886.] 



A Theory of Voltaic Action, 



307 



equalised (sensibly) the potentials of the films, by altering those 

 of the metals which were previously at one potential.* 



29. Evidently then, as I have before suggested ('Phil. Mag.,' 

 Feb., 1879), in experiments purporting to give the difference of 

 potential between a metal and a liquid by the condenser method, such 

 as those of Hankel, Grerland, Clifton, Ayrton and Perry, &c, we have 

 really a two-fluid cell (one fluid being that under examination and the 

 other that condensed on the metal plate), with a dielectric division of 

 air between the two fluids. Any observed differences of potential can 

 readily arise from differences between the nature and constitution of 

 the film and those of the liquid under examination, even if the latter 

 be water. The film can, for instance, more easily replace by absorp- 

 tion from the air any oxygen in solution which may have entered into 

 combination with nascent hydrogen liberated by the oxidation of the 

 metal. 



These differences naturally cause different actions on the metal, 

 entailing different states of its surface, which again react on the 

 electrolytes differently. 



If in experiments of this kind the metal dipping in the liquid be 

 different to that of which the plate is formed, we have of course then 

 a two-metal two-fluid cell. 



30. This aspect of the Volta condenser as a copper-fluid-zinc cell 

 divided in its electrolyte, suggested the possibility of joining the 

 films only on the two metals, without bringing the metals them- 

 selves in contact, and so producing a real galvanic current-producing 

 cell from the apparently dry metals. This I succeeded in doing 

 after a very great expenditure of time and patience ; an expenditure 

 partly owing to the unsuitability of the apparatus — the condenser, 

 fig. 3, described § 20, having been designed for another purpose, 

 was not capable easily of sufficiently fine adjustment. After the 

 first few preliminary trials had given promise of a decided result, 

 a micrometer screw was added, instead of the support at M, in order 

 to be able to form an approximate idea of the distance between the 

 plates during an experiment. 



The copper and zinc plates of the condenser were first carefully 

 faced up to a plane surface by the use of a surface plate, and then 

 lightly ground together with washed emery-powder, after which the 

 zinc received a light rubbing with fine emery-paper. Wire connexions 

 were fixed by screws to each plate, and the pair could thus be joined 

 by means of a mercury cup arrangement either to a reflecting galva- 

 nometer, giving a deflection of about seventeen divisions for a millionth 



* [When the quadrants of copper and iron were disconnected and then joined by 

 a drop of potassium sulphide solution, the difference of potential near them was 

 (as would be expected) increased instead of annulled. It became about double its 

 former value. — November 1, 1886.] 



