﻿288 Mr. E. D. Kleeman on the Recombination 



readings in Table VI. are comparable with one another. 

 Each uranium leak is the mean o£ two readings, which seldom 

 differed by more than two per cent. 



In the case of a mixture of air and a vapour, the air was 

 drawn by means of a water-pump through a wash-bottle 

 containing the volatile liquid. 



It will be seen that the uranium saturation leak was not 

 appreciably affected when the air was bubbled through methyl 

 alcohol, but the lack of saturation for low voltages was 

 approximately doubled. 



In the case of ethyl chloride, an amount' of ethyl chloride 

 gas, about double the cubical contents of the chamber, was 

 passed into the chamber so that the ethyl chloride entered at 

 the bottom while the air was expelled at the top ; the chamber 

 ought to have contained a large percentage of the gas at 

 the end of the process. It will be seen that the uranium 

 saturation leak is almost double the air leak, and the lack of 

 saturation for low voltages is increased enormously, while the 

 X-ray saturation leak is also about doubled, but the lack of 

 saturation for low voltages only slightly increased. 



On the whole, it appears from these observations that the 

 effect of initial recombination is also small in other gases 

 besides air. 



§11. 



The fact that initial recombination is practically absent in 

 a gas ionized by X rays is due, probably, to the electron 

 being ejected with a greater speed from its parent molecule 

 than is the case when a molecule is ionized by an a. particle. 

 We know from Dorn's experiments that the electrons ejected 

 from a metal plate on which X rays fall, have a speed varying 

 from 1*8 x 10 9 to 8%5 x 10 9 cm./sec, and it is very probable 

 that an electron is ejected from its parent molecule with a 

 velocity of the same order. 



Curie and Sagnac have shown that the penetrating power 

 of these ejected electrons is of the same order as that of the 

 Lenard rays, a piece of aluminium *46 X 10~ 4 cm. thick re- 

 ducing the stream of electrons about 40 per cent. If we 

 take 3'42 for their coefficient of absorption in air, as found 

 by Lenard for the cathode rays, and apply it to the electrons 

 ejected from air molecules, we find that less than one per 

 cent, is absorbed in a distance of '003 cm., and about 50 per 

 cent, in a distance of '2 cm. 



The electric field at a distance of '003 cm. from a charged 



molecule is ' , =10~ 2 of a volt per cm. approximately. 



