692 BELL SYSTEM TECHNICAL JOURNAL 



and assumes the potential of the plate. I say "almost all" of the gas; 

 a narrow cylindrical sheath about the wire remains comparatively dark, 

 and between the wire and the outer frontier of this sheath the entire 

 potential-rise is spread. Having traversed the frontier, the electrons 

 shoot into the luminous zone of the gas with the energy corresponding 

 to the full potential-rise from the filament to the end-plate. It is as 

 though the boundary of the sheath were a grid connected to the plate; 

 the discharge itself creates its own impalpable grid. The method, then, 

 consists in measuring the current into the cylinder over a range of 

 values of the potential thereof, beginning when the cylinder is at the 

 same potential as the filament and the only electrons which can attain 

 it are those which come clear through the luminous zone without 

 deflection or loss of energy, and ending when it is at the same potential 

 as the end-plate and the luminous gas (or at any rate when it is well 

 above the filament) and electrons can reach it despite their collisions 

 en route. By analyzing the shape of the curve, Langmuir and Jones 

 are able to deduce the values of the cross-section for interception 

 for the various gases they tested, and values of several other things 

 as well; but the analysis is intricate. I shall therefore say only that 

 for the gases neon, hydrogen, argon, helium, nitrogen and mercury, 

 and applying to the electrons voltages ranging from 30 to 100, they 

 found for the cross-section values departing by less than ten per cent 

 from the gas-kinetic cross-section <tq. 



Interception of Positive Ions (Atom-nuclei and Charged Atoms) 



We consider next the results of experiments like these of the fore- 

 going pages, in which the beam traversing the gas is a stream of posi- 

 tive ions — protons, or atoms of heavier elements lacking each an 

 electron — and the quantity measured is the number of ions disappear- 

 ing from the beam, or something more or less nearly equivalent. The 

 experiments as yet are few, and mostly not so accurate as those on 

 electron-beams. This is partly because of the difiiculty of obtaining 

 steady reliable sources of positive ions — a difiiculty which amounts 

 almost to an impossibility, except for alkali-metal ions, and particles 

 issuing from high-voltage discharge-tubes with kinetic energy amount- 

 ing to thousands of equivalent volts. 



The cross-section for interception, a, is defined in the same way as 

 for electrons — as the ratio {Q — R)/NQdx, to return to the notation of 

 equation (1) of this article. But unless the positive ion is a proton, 

 we should certainly not visualize a- as the cross-sectional area of a mole- 

 cule of the gas; the size of the flying ion itself is involved. The cus- 

 tomary procedure is to compute the "radius" R corresponding to the 



i 



