Electric Intensity and Conductivity in Rarefied Gases. 51o 



The form of the potential-fall curve in the electric discharge 

 through a bunsen-flame is, as I have shown *, very similar to 

 that of the potential fall curve in the discharge in rarefied 

 gases, and in the flame nearly all the ionization appears to 

 occur on the metallic electrodes when these are red hot. It 

 seems to be the case in every kind of gaseous discharge that 

 the ionization occurs most easily at the surface of a metal. In 

 the ordinary discharge at low pressures it seems also very 

 probable that all the ionization due simply to the electric 

 intensity which may be called the primary ionization, occurs 

 at the surfaces of the electrodes, for it is only close to the 

 electrodes that the intensity is great enough to be supposed 

 to produce ionization directly. The curves for the intensity 

 given above do not show the sudden drops of potential and 

 corresponding enormous intensities which occur at both elec- 

 trodes, only the beginning of the rise of intensity near the 

 negative electrode is shown. The negative ions formed at 

 the surface of the cathode constitute the cathode rays, and 

 produce the ionization which the intensity-curves show occurs 

 in the negative glow. 



The effect of this secondary ionization in the negative glow 

 is to diminish the intensity there to a very small value. As 

 the negative ions get separated from the positive ions they 

 cause the intensity gradually to rise. This gradual rise in the 

 intensity goes on until it is suddenly stopped by the appear- 

 ance of the positive column. In the positive column more 

 secondary ionization occurs. In the case of a uniform positive 

 column the ionization at every point must be equal to the 

 recombination. If it is supposed that the ionization in the 

 positive column, like that in the negative glow, is due to the 

 impacts of negative ions moving with great velocity, then it 

 follows that the break in the gradual rise in the intensity 

 which occurs at the end of the positive column is due to the 

 intensity having become great enough to give the negative 

 ions a velocity sufficient to enable them to produce as much 

 ionization in the gas as the recombination which is going on. 

 In the positive column the intensity, according to Hittorf, is 

 independent of the current-strength, which shows that the 

 amount of ionization going on there is proportional to the 

 square of the current-strength. Thus if q is the amount of 

 ionization going on in unit volume, assume q = at 2 , where i is 

 the current-density and a a constant; also we have q — an x n 2 

 and i=(hn 1 + k 2 n 2 )eX, using Prof. Thomson's notation (loc. 



* Phil. Traus. A. 1899. 



