﻿of Ions made hy «, ft, y, and X Rays. 283 



on Langevin's curve, represents the leak in the lead chamber, 

 the corresponding ordinate representing the uranium chamber 



leak is — -, that is 6, where -=- is the ratio o£ the 



y «/ 



current in the uranium chamber to that in the secondary 



X-ray chamber. 



If we put 1 = 6, x = '5, and r = 3 in the foregoing equation, 

 we get K?r7rr 2 =r07. We are now able to plot the relation 

 between x and I. This was done, and is shown by the carve 

 B. Curve was obtained by taking the current in the 

 uranium chamber equal to that in the secondary X-ray 

 chamber. 



It will be seen that the ionization was far more dense near 

 the lead plate of the X-ray chamber than near the uranium 

 oxide, and lack of saturation in the latter case was therefore 

 not due to diffusion, or the ionization not being uniform. 



The effect of diffusion must always be small when 

 the field applied gives to an ion a velocity which is great 

 in comparison with the velocity of diffusion. Consider 

 a slab of ionized gas of thickness one or two mean free paths 

 of a molecule and lying, close to a metal plate: it will rapidly 

 lose its ions on account of their rapid to and fro motion 

 bringing them in contact with the plate, whereby they lose 

 their charges. But the ions adjacent to this slab will cross, 

 as a whole, the boundary with a velocity which is the sum 

 of the velocity of diffusion and the velocity given to them 

 by the attraction of their induced charges, for the ions that 

 cross the boundarv through their rapid to and fro motion do 

 not at once give up their charges, and therefore we have also 

 ions which cross the boundary in the opposite direction. 



The force on an ion due to the induced charge in a metal 

 plate at a distance of 5 x 10~ 4 cm. from the ion is about ^ 

 of a volt. 



Thus the drift of ions towards a metal plate is equivalent 

 to the application of a fraction of a volt to the ions of both 

 signs, and the loss of ions therefore confined to a slab of 

 ionized gas of thickness several mean free paths of a molecule, 

 when the external field is strong. The loss is therefore in 

 general very small. 



It can be shown that the lack of saturation observed with 

 weak ionization by a rays is independent of general recombi- 

 nation, whatever the value of the recombination coefficient, 

 and independent of the velocity of an ion under unit electro- 

 motive force. 



Consider an ionized gas between two metal electrodes at 

 a difference of potential. 



