Disappearance of Gas- in the Electric Discharge. 235 



hitherto, is almost confined to the case when the thermionic 

 current is far from saturation ; it would be equally interest- 

 ing when saturation is attained. It must be remembered 

 that, even when it is saturated, some positive ions must be in 

 the neighbourhood of the cathode, and since the field in its 

 immediate neighbourhood must be such as to drive them 

 towards the cathode, some must fall into it. A state in 

 which all positive ions arrive at the grid is unattainable, 

 whatever the form of the grid. The bearing of these remarks 

 will appear in par. 15. 



Faraday's Law for Gaseous Reactions. 



6. The considerations just presented show that the absorp- 

 tion of gas may be associated with its ionization. The 

 experiments about to be described show that, in some cases, 

 it is associated in a very simple manner. 



The gas mainly studied was CO. In par. 11 of the 

 .first paper, it was shown that the disappearance of this 

 gas involves its conversion into C0 2 , one molecule of G0 2 

 being produced for two molecules of CO disappearing. It 

 was suggested also that CO could disappear by adhesion to 

 the walls ; this action is left out of account here ; the obser- 

 vations were never begun till the discharge had been passing 

 for some little time, long enough to make the walls take up 

 all the CO that could adhere to them. In any case this 

 adhesion is much less important at the relatively high 

 pressures of the experiments about to be described than at 

 the relatively low pressures of the previous work. 



Accordingly, the chemical reaction must apparently be 



2nCO = nCO a + nC. 



No direct evidence of: the liberation of carbon has been 

 obtained, but the difficult and elaborate experiments neces- 

 sary to detect it certainly have not been undertaken. If this 

 is indeed the reaction, if we suppose at the outset that n = l, 

 and if we attempt to interpret the reaction in terms of 

 ionization, we have the alternatives that a positive CO ion 

 unites either (1) with another positive ion, (2) with a 

 neutral molecule, or, possibly, {?>) with a negative ion — though 

 at the low pressures of these experiments the existence of 

 negative ions is very doubtful. 



If (1) is correct, the rate of the reaction should be propor- 

 tional to the square of the number of ions present and 

 independent of the pressure, which measures the concentra- 

 tion of neutral molecules ; if (2) is correct, it should be 



112 



