Electric Discharge in rarefied Hydrogen and Oxygen. 299 



As an illustration, it may be seen from the curves that 

 when the distance between the electrodes is 12 cms. the 

 quantity of water formed per coulomb is perceptibly greater 

 at the low pressure of 1 mm. than at the pressure 4 mms., 

 and even than at the pressure 9 mms. 



The points corresponding to a given pressure fall upon a 

 curve, or at least do not deviate further from it than errors 

 of observation (about 5 per cent.) should allow ; and thus 

 the remark made above is justified, that if the current 

 and pressure are kept constant, the chemical action per 

 coulomb that passes is a function of the distance between the 

 electrodes. 



It should be noticed that when D exceeds 10 cms. the curve 

 corresponding to the pressure 1*05 mm. lies above that corre- 

 sponding to 2'1 mms. which lies above that corresponding to 

 1*5 mms.; but that the curve corresponding to *65 mm. (not 

 plotted in fig. 2) lies below those corresponding to 1*05 mm., 

 2*1 mms. 



Explanation of the observed 



I shall endeavour to show that the phenomena presented 

 by the forms and relative positions of these curves can be 

 completely explained by the same hypothesis which I made in 

 my previous paper and mentioned above, namely, that the ions 

 moving swiftly under the electric force dissociate into un- 

 charged atoms those molecules which they strike under 

 favourable conditions and with sufficiently high velocity. 

 But first it is necessary to recall the theory of the electric 

 discbarge in any gas between two parallel plate electrodes 

 as established by Professor J. S. Townsend. 



When the current exceeds a certain small value there is an 

 intense field of force near the cathode, namely, the region of 

 the cathode fall of potential where ions are generated by the 

 impacts of both the positive and negative ions with the neutral 

 molecules of the gas. From this region of intense force — 

 due to the accumulation near the cathode of the more slowly 

 moving positive ions— the potential rises more gradually up 

 towards the anode ; and the electric force tends to become 

 uniform if the distance between the electrodes is great enough. 

 There seems to be no reason to suppose that a sudden rise of 

 potential occurs at the anode (Townsend, Phil. Mag. June 

 1906), as has been supposed. The region of approximately 

 uniform * force terminating at the anode and called " the 



* Although the number of positive ions which emerge from the 

 positive column during a discharge is a very small fraction of the 

 number of negative ions which traverse it, vet the numbers of ions of 



