the Electrical Conduction produced by it. 143 



gradual loss of charge which continued until the electrode 

 was discharged. 



When the electrode A was charged to 30 volts there was 

 no rate of leak of B. The rate of leak of B or A is thus 

 proportional to total number of ions in the gas. The ions 

 recombine in the interval taken for the air to pass between 

 A and B. The rate of leak of B for a saturating voltage, 

 when A is to earth, is thus less than that of A. 



For a particular experiment the rate of leak of the electrode 

 A was 146 divisions per minute. When A was connected to 

 earth, the saturation rate of leak of B was 100 divisions per 

 minute. The distance between A and B was 44 cm., and the 

 mean velocity of the current of air along the tube 70 cm. per 

 second. In the time, therefore, of *63 sec. the conductivity 

 of the gas has fallen to '68 of its value. 



If we assume, as in the case of Bontgenized air *, that the 

 loss of conductivity is due to the recombination of the ions, 

 the variation of the number with the time is given by 



dn „ 



dt=-« n > 



where n is the number of ions per c.c. and a a constant. If 

 N is the number of ions at the electrode A, the number of 

 ions n at B after an interval t is given by 



n N 



JSTow the saturating rates of leak at A and B are propor- 

 tional to N and n, and it can readily be deduced that the 

 time taken for the number of ions to recombine to half their 

 number is equal to 1*3 sec. This is a much slower rate of 

 recombination than with Bontgenized air near an ordinary 

 Crookes tube. 



The amount of ionization by the uranium radiation is in 

 general much smaller than that due to Bontgen rays, so that 

 the time taken for the ions to fall to half their number is 

 longer. 



The phenomenon of recombination of the ions is very similar 

 in both uranium and Bontgen conduction. In order to test 

 whether the rate of recombination of the ions is proportional 

 to the square of the number present in the gas, the following 

 experiment was performed : — 



* Phil. Mag. Nov, 1897, 



