Ionization Power of Negative Ions by Collisions. 293 



pressures. To account for this he assumes that there must 

 be a denser layer of absorbing vapours surrounding the arc 

 under the heavier pressures. He thus introduces a density 

 change in part of the metallic vapour, and it does not require 

 much extension of the assumption to say that this change 

 permeates the whole of the vapour in the arc. 



There seems, therefore, to be a fair amount of evidence to 

 justify the assumption that an increase in the density (or 

 pressure at constant temperature) of the metallic vapour in 

 the arc corresponds to an increase in the pressure of the 

 surrounding atmosphere. The evidence is, however, only 

 qualitative. If the theory here developed accounts for the 

 Humphreys' pressure shift, as determined in his experiments, 

 we must conclude that the pressure of the metallic vapour is 

 proportional, if not equal, to the pressure of the surrounding 

 atmosphere. This fact is not quite in agreement with the 

 results obtained by Haschek and Mache, but their experi- 

 ments were conducted at very low pressures, and are hardly 

 suitable to use quantitatively in any discussion referring to 

 Humphreys' experiments where the pressure of the surround- 

 ing air varied between 1 and 100 atmospheres. 



If, however, there is this change in density in Humphreys' 

 experiments then the phenomena investigated by him would 

 seem to admit of the simple explanation of the shift in the 

 lines here developed. 



XXII. Note on Ionization Power of Negative Ions by 

 Collisions at High Pressures. 



To the Editors of the Philosophical Magazine. 



Gentlemen, — 



IN a recent paper * Mr. E. S. Bishop has published results 

 giving the number of fresh ions a produced per cm. by 

 negative ions in air, hydrogen, and carbon dioxide under 

 various pressures and electric forces. Taking the electric 

 force as X volts per cm. and measuring p in mm. of mercury 



the range of values of — covered in this paper is from 45 to 



70 for air, 20 to 70 for hydrogen, and 40 to 90 for carbon 

 dioxide. 



The values previously obtained by Professor Townsend f 



* Phys. Kev. xxxiii. p. 325, Nov. 1911 ; Phys. Zeitsckr. xii. p. 1148, 

 Dec. 8, 1911. 



t ' Theory of Ionization of Gases by Collision.' 



