﻿Theory of Ionization hy Collision. 857 



importance, and I do not know of any previous experiments 

 which show that the negative ions generated by Rontgen rays 

 from molecules are electrons and not atoms. 



The view that negative electrons combine with molecules 

 or form groups of molecules when the electric force is so 

 great that ionization by collision takes place, is a contribution 

 to the theory which was subsequently made by Professor Sir 

 J. J. Thomson in a paper that appeared in the Philosophical 

 Magazine, April 1901. I do not know what experimental 

 evidence there is in favour of it, but I did not neglect to 

 consider it, as Mr. Campbell implies. In my book on the 

 subject of Ionization by Collision, I dealt with trie matter 

 and mentioned two reasons against that hypothesis. One, 

 which is explained in the latter portion of the book, was 

 based on some accurate determinations of currents between 

 parallel plates when negative ions were produced by col- 

 lisions ; the other was based on experiments on the motion 

 of negative ions when forces much smaller than those re- 

 quired to generate ions by collisions were acting. On the 

 latter subject the reader was referred to some papers pub- 

 lished in the ' Proceedings ' of the Royal Society. 



Perhaps I may mention some values I have recently obtained 

 for the rates of diffusion of negative ions into air at various 

 pressures, as they illustrate the substance of the second 

 argument. 



The coefficient of diffusion is not independent of the electric 

 force. When a small force 1*5 volts per centimetre is acting 

 the coefficients of diffusion of the negative ions into air are 

 •045, 54, and 33000, at pressures 760 millimetres, 16*5 milli- 

 metres, and 4*5 millimetres respectively. 



The number corresponding to the pressure 4*5 was obtained 

 with ions that had made on an average 8000 collisions with 

 molecules of air before their motion was investigated. If 

 there was any tendency for the ions to become attached to 

 molecules they had ample opportunity of doing so, and their 

 rate of diffusion would then have been smaller than the rate 

 of diffusion of hydrogen molecules. 



The rate of diffusion of hydrogen into air at 4*5 millimetres 

 pressure is about 105. This number differs considerably 

 from 33000, which shows that, even after making 8000 

 collisions with molecules, the large majority of the ions must 

 be in the electronic state at any time. If this state persists 

 to such an extent when a low force 1*5 volts per centimetre 

 is acting, it is more than likely that it should persist to a 

 greater extent under such forces as are sufficiently large to 

 produce ionization by collision. 



The second conclusion as to the erroneous nature oi' the 

 mathematical work is. as Bar as I can gather, to be found in 



