64 



Royal Society : — 



Time after immersion. 



In air. 



InC0 2 . 



1 minute. 

 5 minutes. 

 10 „ 



165 

 135 

 135 



76 

 62 



58 



As the cell in an atmosphere of carbonic-acid gas showed consi- 

 derable action, in fact nearly half as much as that in the air, each 

 cell was short-circuited for 23 hours, with the expectation that 

 any oxygen in the closed vessel would be used up ; and, indeed, the 

 most prominent crystals of silver in the cell in carbonic-acid gas 

 became reddened, while a cuprous deposit extended over the whole 

 of the crystals in the other cell. When, however, the short wires 

 were removed and the galvanometer interposed, the cell in the air 

 gave a deflection of 136,. practically the same as before, but that 

 in carbonic-acid gas, instead of showing a great decrease, rose to 80. 

 It was then found that the vessel containing the latter slowly 

 admitted air ; so the contents were swept out by a fresh stream 

 of carbonic-acid gas, and it was made properly air-tight. After con- 

 nexion by a short wire for 3 days the galvanometer indicated a de- 

 flection of 20, that of the cell in the air being 110, temperature 

 10° C. As this showed a very great reduction of the chemical 

 action, carbonic-acid gas was again passed through the vessel for 

 an hour or two ; and after a connexion of two more days the in- 

 dication of the galvanometer was only 3, while the other cell gave 

 115, the temperature being now 10 o, 5 C. The action, therefore, 

 was at last reduced almost to nothing ; and the original fault in 

 the experiment brought out, perhaps more clearly than would other- 

 wise be seen, how eagerly the solution will absorb even minute 

 quantities of oxygen from the surrounding gas. 



An important point to determine was the best strength of the 

 copper nitrate solution. Six per cent, was generally preferred, for 

 two reasons : — first, it gives about the maximum of effect — a solu- 

 tion four times as strong gives less than half the deflection, and 

 a solution only a quarter as strong gives only two thirds ; secondly, 

 a stronger solution than this 6 per cent, is apt to produce a deposit, 

 not of pure cuprous oxide, but of a subnitrate, which was supposed 

 to clog up the silver crystals to a greater extent. 



Another point investigated was the best proportion between the 

 areas of the metallic surfaces. Experiments were made with ver- 

 tical plates, in which the silver was kept at a uniform size and 

 the copper was diminished by covering it more and more with 

 varnish ; and another set was made in which the copper remained 

 the same, while the silver plate was reduced. 



The results may be thus exhibited : — 



