of the Heat of Combustion. 105 



transitory currents did not entirely disappear as was antici- 

 pated. The gas, though prepared carefully and in different 

 ways, could not be obtained perfectly pure, and when exposed 

 to an alkahne solution, ^^^^'^ '^^ ^"^ would remain uncondensed. 

 In order therefore to remove any free oxygen which the gas 

 might contain, I exposed it during two days and two nights to 

 the action of a stick of phosphorus. After this, immersion 

 caused no, or at most, very trivial transitory effects ; but on ad- 

 mitting only one per cent, of oxygen they became very consi- 

 derable, — a striking example of the power possessed by metals 

 of collecting and condensing oxygen upon their surfaces. I 

 do not bring forward this experiment as a proof of the entire 

 non-action of carbonic acid, because the phosphorus was found 

 to have decomposed it partially. 



30. All these phaenomena are easily understood, if, with the 

 great body of philosophers, we keep in view the intimate rela- 

 tion which subsists between chemical affinity and the electric 

 current. For let^J, fig. 2, represent a plate of platinum ; z^ a 

 plate of zinc, or other electro-positive metal ; and c, one of a 

 series of atoms of water extending from p to z. The intensity 

 of the current along the wire w, is proportional to the affinity 

 of oxygen for the positive metal, minus the affinity of oxygen 

 for hydrogen. But if/? be covered with a film of oxygen, the 

 current will be entirely proportional to the affinity of the po- 

 sitive metal for oxygen. In the former case, c = rr — ]i\ in 

 the latter, c' = z. 



31. Considering these equations, it is obvious why, as I have 

 observed ( 1 5.), the transitory currents are better exhibited with 

 iron than with zinc as a positive element ; for in proportion to 

 the smallness of c, provided it remain greater than /z, will the 

 difference between c and c' be more manifest. If d — c be 

 the same for both iron and zinc, we shall have a proof of the 

 accuracy of these principles. 



32. Thus from (21.), turning the deflections of the needle 

 into quantities of electricity, we have 63^° = 0°'034' Q, and 

 29^° = 0°-0072 Q, of which the difference" is 0°-0268 Q, when 

 iron is the positive element. We have also 76° = 0°'056 Q, 

 and 63° = O°-027 Q, of which the difference is 0°-029 Q, when 

 zinc is the positive element. I consider these differences as 

 nearly equal as could have been expected from the nature of 

 the experiments. 



33. I might now proceed to consider in detail several phae- 

 nomena (such as the very rapid corrosion of metals when 

 they are exposed to the joint action of air and moisture, &c.) 

 which are occasioned by the great intensity of galvanic action 

 in constuiuencc of the mixture of oxygen witJz the liquid 

 But I hasten to fulfil my principal design. 



