1886.] 



A Theory of Voltaic Action. 



299 



The consideration as to which oxide is formed leads to the question 

 whether any difference of potential would be found near the contact 

 of two different oxides of the same metal — I experimented on litharge 

 and red lead, and on stannous and stannic oxides, but did not get 

 results sufficiently definite to be reliable — but I think the experiments 

 worth repeating with freshly-formed oxides, which mine were not. 



12. It has been suggested in criticising my former experiments on 

 the effect of various gases on the difference of potential near certain 

 metals in contact, that it would have beeu better to have had quanti- 

 tative results. Although for the reasons given above these are 

 perhaps not of very much value, I have, however, made measurements 

 in several cases. The following are the results with a copper-iron 

 pair in hydrogen sulphide gas. 



Quadrants of these metals after being freshly cleaned with glass- 

 paper were put in position, and the difference of potential in air 

 measured as quickly as possible was found to beO'059D, 0'062D, mean 

 0"06D, iron side positive. 



Hydrogen sulphide was now sent into the electrometer, reversing 

 immediately the relative potentials near the metals. The stopcocks 

 were then closed and the following measurements taken : — t repre- 

 senting the time elapsed after admission of the gas, and P/D the ratio 

 of the difference of potential near the metals to that of %he terminals 

 of a Daniell cell. 



t. 8 m. 15 m. 25 m. lh. 18 m. 2h. 10 m. 2h..8o.m. 12 h. 10 ra. 



P/D. 0-4 0-48 0-53 055 047 043" 03 



Some fresh hydrogen sulphide now sent in caused ; a slight rise, 

 P/D = 0*32, after which it fell slowly, and in about 11 hours more 

 (£=23 h. 25 m.) it was 26. The instrument was now cleared of gis 

 and left open for about \\ hours, when P/D fell to 0*1. Re-admiss.on 

 of gas raised it to 0'32. This was repeated with the same result, 

 after which the instrument was left open for 24 hours, when the iron 

 side was found positive, P/D = 32. This somewhat high value may 

 be due to oxidation of the iron sulphide, but more probably perhaps 

 to oxidation of the iron itself, coupled with the protective action of 

 the very thick coating of sulphide on the copper. After again ad- 

 mitting the gas the copper side became again positive, P/D = 26D. 

 Several more changes of the atmosphere were made, the corresponding 

 change in difference of potential taking place each time. The electro- 

 meter was then charged with the gas and closed up, the water vessel 

 having been removed (so that its vapour might not interfere with 

 insulation inside the instrument), and its place supplied by paper 

 vanes cemented to the weight. The difference of potential then fell 

 gradually till at the end of eight days after the gas had first acted on 

 the metals it was sensibly zero. On opening the electrometer plenty 



