466 BELL SYSTEM TECHNICAL JOURNAL 



very large one whenever ionization commences. This scheme has been 

 adopted by K. T. Compton. 



Another, and the most common, device for distinguishing ionization 

 from radiation consists in surrounding the collector with a sheath of 

 metal gauze, maintained at a potential slightly (say 3 volts) more nega- 

 tive than the electrode which it screens. Positive ions pass through its 

 meshes to the collector, somewhat slowed down but not driven back. 

 Radiation also passes through the meshes to the collector, but the elec- 

 trons which it drives out are turned back by the adverse field and re- 

 enter the metal whence they came, so that the net result is the same as 

 though they had never come out. Ionization thus produces a current 

 of positive charge into the collector, and radiation none; or radiation 

 may even produce a current of negative charge into the collector, thus 

 accentuating the contrast, for electrons which are ejected from the 

 metal gauze are drawn to the screened electrode. This is the scheme 

 devised by F. S. Goucher. 



Another, and possibly the best, method for measuring ionizing- 

 potentials is quite insensitive to radiation. A hot filament is immersed 

 in the gas, which may be supposed to be surrounded by connected metal 

 walls so that its boundaries are all at the same potential. If the efflux 

 of electrons from the filament is so plentiful that it is limited by space- 

 charge,2 and this condition persists as the potential-difference between 

 walls and filament is raised to the value just sufficing to give to the 

 electrons energy enough to ionize the gas, then at the moment of in- 

 cipient ionization the space-charge limitation is partially or totally 

 cancelled, and the current increases sharply. This is I. Langmuir's 

 method. It is better to keep the potential-difference between the walls 

 and the filament small and constant, and admit into the gas electrons 

 with controllable energy from another source; when the energy of 

 these auxiliary electrons is raised to attain the ionizing-potential of the 

 gas, the current from the filament suddenly increases. This is the 

 method of G. Hertz^ and K. H. Kingdon.^ 



Most of the accurate measurements of ionizing-potentials have been 

 made with a collecting-electrode sheathed by a gauze, according to the 

 precept of Goucher. The apparatus is a complicated affair, for the 

 parts already mentioned are by no means all that are required; in 

 some cases the whole interior of the tube appears to be webbed with 

 gauzes. A hot filament (occasionally an illuminated metal plate) is 

 provided as source for electrons, and its potential — or the potential of 



^ See the sixth article of this series (December, 1924). 

 ^ZS.f. Phys. 18, pp. 307-316 (1923). 

 *Phys. Rev. (2) 21, pp. 404-418 (1923). 



