Conducting Gases on the Direction of the Electric Field. 219 



applied, the current must also tend to equality, since the 

 maximum current through the gas for a P.D. sufficient to 

 remove all the ions to the electrodes, before appreciable 

 recombination takes place, depends only on the number of 

 ions present and not on their velocity. 



We should thus expect in the present case to obtain a 

 maximum ratio of the currents for a certain P.D., and this 

 ratio to diminish when the P.D. was raised or lowered. 



The potential- gradient, near the radio-active surface, is 

 much greater with the lower plate negative than positive. 

 The reason of this is easily seen. Since the current through 

 the gas is greater when the lower plate is negative, more 

 positive ions have to reach the lower plate than negative 

 ions when the lower plate is positive. The positive ions 

 also move more slowly than the negative in the dry gas, and 

 both causes assist in requiring a steeper potential slope to 

 force the number of positive ions, necessary to carry the 

 current, towards the lower plate. 



Effect of Moisture and, Vapours on the Ratio of the Currents. 



Zeleny * has directly measured the velocity of ions pro- 

 duced by Bontgen-rays, and found that the ratio of the 

 velocity of the negative to that of the positive ion for dry air 

 is 1*375, and for moist air 1'10. Townsend f. by comparing 

 the rates of diffusion of ions, determines the ratio of the 

 velocity for dry air to be 1*54 and 1*09 for moist air. 



If the ratio of the currents in the experimental case con- 

 sidered depends on the ratio of the velocities of the ions, we 

 should thus expect to obtain wide differences between the 

 ratios of the current when the air is dry or moist. This is 

 fully confirmed by the experiments. 



The maximum ratio of the currents observed for air which 

 had been left over P 2 5 for a week was 1*45. The ratio ob- 

 served for air standing over water at a temperature of 18° C. 

 was 1*12. For intermediate stages of dryness the ratio of 

 the currents took intermediate values. 



The effect of alcohol-vapour on the ratio of the currents is 

 more marked than in the case of water-vapour. 



With fairly dry air in the apparatus, the ratio observed for 

 the currents w r as 1'37. On introducing alcohol through a 

 side tube into the closed apparatus, the ratio of the current 

 with the lower plate negative rapidly diminished, while the 

 current in the other direction was not appreciably affected. 

 After a certain stage, however, the currents in both directions 

 decreased. 



* Trans. Koy. Soc. 1900. t Ibid. 1900. 



