of Recombination of the Ions of Gases. 435 



It will be seen from the above table that the velocities of 

 the ions follow the inverse order of the densities with the 

 exception of chlorine gas. The velocity of the hydrogen ion 

 through hydrogen is nearly four times as fast as the velocity of 

 the oxygen ion in oxygen. The ions of sulphur dioxide gave 

 the slowest velocity, being only about ^ of that of hydrogen. 



The velocity of the ions through a gas was found to be 

 independent of the amount of ionization of the gas. The 

 velocities deduced from the two different sets of experiments, 

 when the intensity of radiation in one case was six times that 

 of the other, were found to be the same. In the one case, 

 therefore, six times as many ions per c.c. were present as in 

 the other, but the velocity remained unaltered. There is 

 thus no correction to be applied for the velocity of the ions 

 in a gas like chlorine, whose ionization is large compared 

 with that of air. The fact that the velocity under a given 

 small electromotive force is independent of the number of 

 ions per c.c. in the gas, shows that the movement of the 

 positive and negative ions does not produce any resultant 

 electrostatic field between the plates. 



In the method which has been used for determining the 

 velocity, it has been assumed that the conductivity of the 

 gas is purely due to volume ionization of the gas. Perrin 

 (Comptes Hendus, March 1st, 1897) has, however, recently 

 shown that the rate of leak between two plates is made up 

 of two parts, one due to the volume ionization of the gas, 

 and the other due to surface ionization at the surface of 

 separation of the metal plate, on which the radiation im- 

 pinges, and the surrounding gas. The rate of leak due to the 

 surface action is quite comparable with that due to the 

 volume ionization, when the plates are 1 cm. apart, especially 

 in the case when the electrodes are of silver, gold, or zinc. 



It seems probable that the gas close to the surface of the 

 plate on which the radiation falls has a much greater density 

 of ionization than the gas between the plates, and since this 

 increase is confined to a very thin layer close to the elec- 

 trode, that the rate of recombination for the surface-ions is 

 far more rapid than for the Rontgenized air some distance 

 from the surface. The existence of this effect would tend 

 to increase the rate of leak q ; while the number of ions, N, 

 as determined by the application of an E.M.F. a short 

 interval after the rays had ceased, would not be appreciably 

 affected, since the ions near the surface probably recombine 

 with great rapidity. The value of T which is obtained from 



N 

 the equation T = — would thus be too small. The correction 



l l 



