472 BELL SYSTEM TECHNICAL JOURNAL 



streams of bombarding electrons are used : that is to say, it occurs be- 

 cause light of less than the threshold frequency puts some of the atoms 

 into abnormal states, in which less energy is required to ionize them 

 than in the normal state. 



The study of ionization by positive ions is also very troublesome. 

 This is partly because there are no such convenient sources for con- 

 trollable positive ions as there are for electrons. The ions emerging 

 from hot filaments are generally not all of one kind. If ions of a par- 

 ticular sort, hydrogen ions for instance (these would give the most 

 \aluable information of any) are produced by bombarding the proper 

 kind of gas by electrons having a suitable ionizing-energy, they cannot 

 be used for ionizing except in the same tube and therefore upon the 

 same gas; further, it is necessary to keep the bombarding electrons out 

 of the region where the positive ions are meant to ionize, by an elabo- 

 rate system of gauzes and opposing potentials. If the collecting elec- 

 trode is maintained at a positive potential so as to receive electrons 

 produced by the ionization, it receives also the electrons which are 

 knocked out of the walls of the tube by positive ions which strike them. 

 It is scarcely surprising, then, that the published data are scanty and 

 not always concordant.^ 



The considerations about conservation of momentum during im- 

 pacts, mentioned in dealing with ionization by electrons, show that we 

 should hardly expect a positive ion to be able to ionize unless it has 

 much more energy than must be transferred to the atom to detach the 

 loosest electron from it; twice as much, if the ion is of the same mass 

 as the atom. 



Identification of Ions Produced by Electron-Impacts 



The methods hitherto described for detecting the onset of ionization 

 in a gas show when free positive charges appear in a gas, but give no 

 further information about them. The methods employed by J. J. 

 Thomson and F. W. Aston reveal the charge-to-mass ratios of ions 

 occurring in a gas carrying a self-maintaining discharge, but give very 

 little information about the precise conditions necessary to produce 

 them. A combination of methods of these two kinds was first effected 

 by H. D. Smyth.io 



One of the tubes employed by Smyth is sketched in Fig. 3. Electrons 

 from the filament Fare accelerated through the potential-rise Fi to the 



^ For work published up to 1922 sec the review and bibliography by A. J. Saxton, 

 Phil. Mag. 44, pp. 809-823 (1922). See also J. T. Tate, Phxs. Rev. (2) 23, pp. 293- 

 294 (1924). 



^"Proc. Roy. Soc. A102, pp. 283-293 (1922-23); A104, pp. 121-134 (1923); Phys. 

 Rev. (2) 25, pp. 452^68 (1925) and references there given. 



