of Recombination of the Ions of Gases. 433 



depending as it does on whether the gas is freshly made or 

 has stood undisturbed for some time. 



For a determination of the velocity of the ions the value of 

 T requires to be known with accuracy ; but it is a difficult 

 matter to determine T accurately with a pendulum inter- 

 rupter, and moreover the intensity of the radiation from the 

 Crookes' tube, on which the value of T largely depends, is 

 very liable to change over a long range of experiments. 

 Recourse was therefore had to a simpler method of deter- 

 mining T, which was found to give very consistent and 

 reliable results. It has been shown that the final number of 

 ions N is given by 



and also 



N =aT ' 



therefore 



T=- 



Now if a very large B.M.F. is applied to the testing-vessel 

 the instant after the rays have ceased, the deflexion of the 

 electrometer is proportional to N, and q is proportional to the 

 rate of leak per second for a saturating E.M.F. The ratio of 

 these two quantities is thus readily determined and the value 

 of T known. 



Arrangement of the Experiment. 



The testing-vessel consisted of two parallel plates, the lower 

 being of aluminium and the upper of sheet lead. These 

 plates were separated by blocks of paraffin, which were 

 melted together and formed the sides of the testing- vessel, 

 and at the same time insulated the top from the bottom plate. 

 In order to measure the after-conductivity with accuracy fairly 

 large plates were required. In the apparatus used the plates 

 were 22 cm. square and 4'7 cm. apart. The rays passed 

 through the lower aluminium plate and made the gas inside 

 the vessel a conductor, but were completely stopped by the 

 top lead plate. Care was taken that the radiation fell only 

 on the central portion of the plate where the electrostatic 

 field was sensibly uniform. 



The arrangement of the apparatus was the same as that 

 given in fig. 2, with the exception that the bell-jar was re- 

 placed by the vessel with parallel plates. One pole of a 

 battery of small accumulators was connected to the lower 

 plate, and the electrometer to the upper. 



Phil. Mag. S. 5. Vol. 44. No. 270. Nov. 1897. 2 I 



