Compounds hy homogeneous X-radiation. 181 



X-radiation from the ends and sides must have been very 

 small, producing a correspondingly minute ionization, while, 

 being very soft, it would be completely absorbed even in the 

 case of hydrogen alone. Its effect was thus entirely elimi- 

 nated in the process of calculation, or at any rate, any error 

 due to it was reduced to something far below the errors of 

 experiment. 



Corpuscular radiation from the ends of the chamber was 

 small in amount, and Avas liberated in all the experiments in 

 equal amounts. The only error it could introduce would be 

 negligibly small in most cases, depending on whether equal 

 numbers of corpuscles produced equal ionizations in every case. 



The corpuscular radiation from the gases and vapours in 

 the ionizing- chamber was absorbed differently in the various 

 experiments, however. In the mixtures containing the sub- 

 stances of higher vapour-pressure, it would be mainly absorbed 

 by the vapour, but when compounds of low vapour-pressure 

 were used, the corpuscles would be chiefly absorbed by the 

 hydrogen. The ionization produced in a vapour is not neces- 

 sarily the same as that produced in hydrogen, when equal 

 numbers of corpuscles are absorbed in each — the ionization 

 being usually larger in the vapour than in the hydrogen *. 

 The effect of this is that, in the mixtures containing a larger 

 proportion of vapour, the ionization due to the corpuscular 

 radiation from the mixture is greater than that which would 

 be produced by the liberation (and subsequent absorption) of 

 an equal number of corpuscles in the mixtures containing 

 the low pressure vapours. The observed ionizations deviated 

 from an " atomic " law in this kind of way, being relatively 

 much higher in the higher pressure vapours than in the low- 

 pressure vapours, and accordingly it was in this direction 

 that a further investigation was made. 



The paper by Kleeman, to which reference has already 

 been made, gave the ratio of the ionization produced in 

 certain vapours alone, to that produced in air by the total 

 absorption of equal quantities of corpuscular radiation. 

 From these ratios it is possible to calculate the corre- 

 sponding ratio for a mixture of gases or vapours, provided 

 the relative amounts of corpuscular radiation absorbed by 

 each of the constituents are known. The relative amounts 

 absorbed by the vapour and by the hydrogen in the various 

 mixtures were calculated assuming the absorptions to bo 

 proportional to the densities of the gases actually present 

 (except for hydrogen which behaves anomalously). In this 

 w r ay calculations were made to determine how much corpus- 

 cular radiation must have been liberated in the various 

 * Kleeman, Roy. Soc. Proc, March 1910. 



