20 L. G. Raub 



for in no case was there any change in the pressure as read on the 

 McLeod gauge which could detect a change of about five-tenths 

 of one per cent. 



If this is true, the vakie of the mobihty must be very sensitive 

 to impurities in the gas, since they would be given up very slowly, 

 and the values found for the gas after a fresh filling may be 

 higher than the true value of U for argon, since it is not strictly 

 pure. Franck^' found that the negative ion was very sensitive to 

 impurities, but he did not find that impurities seemed to show any 

 influence on the mobility of the positive ions. However, measure- 

 ments made by this method might well be more sensitive to the 

 effect of impurities — at least to the presence of hydrogen if Skin- 

 ner's^^ value for the mobility of hydrogen is correct — than meas- 

 urements made by the ordinary methods. 



SUMMARY 



It has been shown that in the negative glow there are equal 

 positive and negative currents. Assuming, then, that the negative 

 glow is the source of positive ions, and that the negative ions 

 originate from the extremely small electron current caused by the 

 bombardment of the cathode by the positive ions, it has been shown 

 that at a distance of one ionizing interval toward the cathode from 

 the negative glow the positive ions carry three-fourths of the cur- 

 rent, at the distance of two intervals seven-eighths, and so on, so 

 that at a comparatively short distance the positive ions carry prac- 

 tically all the current. 



From this assumption a theory has been developed for the 

 potential at any point in the gas between the point of minimum 

 gradient and the polarization region for wire cathodes. 



This theory is tested for an aluminium cathode in the monatomic 

 gases, helium and argon, and is found to correspond very satis- 

 factorily with the experimental results. 



The values for the mobility in helium, which was not entirely 

 pure, computed from the theoretical equation (8), were not con- 

 stant for different current densities, and the values of the mobility 



^''' J. Franck, Ber. d. D. Phys. Gcs., 8, p. 291, 1910. 

 ^8 C. A. Skinner, Phys. Rev., XII, p. 143, 1918. 



