﻿814 Mr. A. J. Saxton on Impact Ionization by 



potentials for electrons in the gas. They concluded that 

 there was no sharp ionization potential for positive ions in 

 the gas, and also found that the minimum potential at which 

 ionization could be detected was low r er the greater the 

 intensity of the source of positive ions. Horton and Davies * 

 made a thorough investigation of the ionizing properties of 

 the positive ions emitted from an incandescent tantalum 

 filament in helium. Their results indicated the production 

 of fresh ions by collision of positive ions accelerated through 

 20 volts. Further investigation led to the view that the 

 ionization produced was not due to the ionization of the gas 

 molecules by direct positive ion impact, but to the bombard- 

 ment of the walls of the ionization chamber by positive ions 

 releasing S-rays. They conclude that the positive ions do 

 not produce ionization when accelerated through potentials 

 of 200 volts. Thus the only investigations satisfying the 

 necessary conditions are to some extent contradictory. 



4. Description of Apparatus. 



The following experiments were undertaken with the hope 

 of obtaining more definite information about the conditions 

 under which ionization is produced by low-speed positive 

 ions. Other workers have used the positive ions from glow- 

 ing filaments so that their nature was not definitely known. 

 In the present work the nature of the ionizing positive ions 

 was known with greater certainty. The}' were produced by 

 electron impact in hydrogen, and the speed of the colliding 

 electrons was great enough to dissociate the molecule on 

 ionization. Thus the positive ions formed would be H- 

 nuclei (protons). They were accelerated through hydrogen 

 at very low pressure and their ionizing properties studied. 



In figs. 1 and 2, F is a tungsten filament heated by a 

 battery of 8 volts supplying a current of 2 to 3 amps, and 

 insulated on paraffin-wax blocks. The filament leads are 

 sealed into a glass tube, the end of which fits as a ground- 

 glass stopper into the side tube B. With this arrangement, 

 when the filament burnt out, it could be replaced more 

 easily. The electrons from the glowing filament F are 

 accelerated towards the nickel electrode A by the P.D. of 

 V volts between the negative end of the filament and A. 

 If Y is greater than the ionization potential for electrons in 

 hydrogen, ionization by electron impact occurs near to A, 

 and the resulting positive ions are accelerated towards the 

 filament by V. Some strike the filament, but some pass on 



*' Hovton and Davies, Proc. Roy. Soc, March 1919, p. 333. 



