the Ions produced in Gases hy Rontgcn Bays. 



239 



other. That one of these parts which was not traversed by the rays 

 was connected to a pair of quadrants of an electrometer, so that it 

 was possible to tell when any ions reached it. A series of readings 

 was taken for the charge reaching the electrometer in a given time 

 for different values of the potential of the outer cylinder. From this 

 was determined the value of this potential for which the ions starting 

 from the outer edge of the ionised layer were just able to reach the 

 juncture in the inner cylinder. 



The ionic velocity in a unit electric field is given by the equation — 



where U is the mean velocity of the gas stream between the cylinders, 

 1 is the inner radius of the outer cylinder, 

 a is the outer radius of the inner cylinder, 

 A is the potential of the outer cylinder, corresponding to 

 X the distance defined above. 



To avoid the presence of vortices in the gas at the place where it 

 was exposed to the rays, a sufficiently small velocity was used, and 

 the gas was previously passed through a long portion of the cylinder 

 to allow the motion to assume a steady state. 



The disturbing influence upon the electric field between the cylinders 

 of the free charges formed in the gas during the conduction was 

 diminished by using weak rays. The fall of potential at the electrodes'^ 

 was also reduced by this means. For diminishing the amount of 

 ionisation due to the secondary radiation produced at the metal sur- 

 face,! the cylinders were made of aluminium, for which metal the effect 

 is the least. 



The spreading of the ions due to diffusion produces an error, the 

 amount of which increases with the time required for the ions to travel 

 between the two cylinders. 



The value of this time is found from the equation T = X/U, where 

 X and U have the same significance as above. 



The experimental values obtained for the velocity decreased as T 

 increased, and from a series of results with different values of T the 

 velocity could be obtained corresponding to T = 0. Since in that 

 case the effects of diffusion and similar causes disappear, this result 

 was taken as the desired value of the ionic velocity. 



For testing the accuracy of the method, in addition to using different 

 values of U and X, changes were also made in the intensity of the 

 rays, in the diameter of the internal cylinder, and in the metal which 

 formed the inner surface of the outer cylinder. 



Determinations were made with the gases when dry and when satu 



* J. Zeleny, ' Camb. Phil. Soc. Proc.,' vol. 10, Part I, p. 17. 

 t J. Perrin. * Comptes Rendus,' vol. 124, p. 455. 



