﻿of Ions made by «, (3, 7, and X Rays. 281 



that of the aluminium chamber together with the electro- 

 meter was found to be about 150 E.S. Us. In the lead 

 chamber the rays fell on an area of 140 cm. 2 , and in the 

 aluminium cham'ber on an area of 50 cm. 2 If we take 100 

 scale-divisicns per second as the average leak in each 

 chamber, we get a current of 9 x 10 _5 E.S. Us. per cm. 2 for 

 the lead chamber, and a current of 10 x 10~ 5 E.S. Us. per 

 cm. 2 for the aluminium chamber. The number of ions lost 

 by recombination is therefore, in both cases, somewhat 

 greater than occurred in the experiment with uranium oxide 

 (about three times greater in the case of the aluminium 

 chamber), but it is still a small fraction, in each case, of the 

 total number received. 



It appears from these observations that there is very little 

 lack of saturation (when the number of ions lost by general 

 recombination is small) with primary and with secondary 

 X rays, and initial recombination is therefore practically 

 absent in air ionized by X rays. 



The greater lack of saturation in the case of the lead plate 

 chamber is probably due to non-uniformity of its ionization, 

 which is certainly very much more dense near the lead plate 

 than elsewhere. 



It may, perhaps, be argued that when the air was ionized 

 by the a rays from a layer of uranium oxide, the ionization 

 was very dense near the layer, and lack of saturation was 

 due to loss by diffusion to the plate. It becomes of im- 

 portance, therefore, to compare the density of ionization at 

 various distances from the uranium oxide with that in the 

 lead chamber due to X rays, for the experiments show that 

 in the latter case the loss of ions is comparatively small. 



Let us take an element of surface 8s on the surface of the 

 uranium oxide, and draw a thin cylinder through it in any 

 direction, and produce it both ways. The number of uranium 

 molecules per c.c. which eject their a particles along the 

 cylinder towards the surface of the uranium, is everywhere 

 the same. Therefore, since the range of an a particle is 

 diminished in proportion to the matter it has to traverse, the 

 ranges of the a. particles along the cylinder from the surface 

 of the uranium consist of a set of ranges diminishing in 

 arithmetical progression, beginning with the maximum range 

 of an a. particle in air and ending with zero. In the last case 

 the a particle from some depth in the uranium oxide just 

 manages to reach the surface. The total ionization in the 

 cylinder may therefore be approximately represented by the 



r 2 K 

 area of a triangle, and written — - , where K denotes a 



