496 Wellisch and Bronson — Distribution of the Active 



difficulty in saturating the cathode activity ? Even under the 

 most favorable conditions which have been employed in our 

 work there is still about 10 per cent of the activity deposited 

 on the walls of the testing vessel. In an attempt to answer 

 this question we are at once led to the corresponding problem 

 in connection with the ionization current which passes through 

 the gas during the activation of the electrodes. Mention has 

 already been made of the researches of Bragg, Kleeman, Mou- 

 lin, and Wheelock in connection with this aspect of the prob- 

 lem. Bragg explained the difficulty of saturating the ioniza- 

 tion due to a-particles by introducing the conception of initial 

 recombination ; viz., that the electron which is expelled in the 

 process of ionization returns in a number of cases to its parent 

 atom, and exceptionally strong electric fields are needed to 

 exercise any appreciable preventive effect upon this tendency 

 to recombine. Kleeman extended the experimental work, 

 using Bragg's theory as a basis. Moulin ascribed the difficulty 

 of obtaining saturation to the fact that the a-particles ionized 

 in columns, so that the density of ionization is not uniform 

 throughout the gas when sufficiently small volumes of gas are 

 considered. This localization of the ions would naturally 

 result in a recombination more intense than that which would 

 correspond to a uniform distribution in the usual acceptation 

 of the term. Moulin showed that saturation appeared to be 

 most difficult when the columns were parallel to the lines of 

 force of the electric field ; when, however, the a-particles 

 moved across the lines the ionization tended more readily to 

 saturation. Wheelock continued the work, adopting the idea 

 of columnar ionization ; in particular, he showed that when 

 the pressure was reduced to about one-third of an atmosphere 

 saturation set in fairly readily. There can be little doubt as 

 to the reality of this columnar effect and the intense recom- 

 bination resulting from the local distribution of the ions ; this 

 is clearly brought out by the slowness with which the ioniza- 

 tion current increases with the potential in the early stage of 

 any curve for a-particle ionization. The experimental results 

 obtained in the present research appear to lead to a radically 

 different explanation of the shape of the ionization curves at 

 the higher potentials. It has been shown that for potentials 

 which are not too low the ratio of the percentage cathode 

 activities for two different potentials is equal to the ratio of the 

 corresponding ionization currents due to the a-particles, and 

 over a wide range is independent of the amount of emanation 

 employed, that is, of the intensity of the intrinsic ionization. 

 This experimental result has already led to the suggestion that 

 the fraction of the total activity w T hich is deposited on the walls 

 of the testing vessel is a measure of the lack of saturation of 



