﻿of Ions made by a, ft, y, and X Rays. 285 



effect on the ions after they are separated, for, as has been 

 mentioned before, an electric field has no influence on the 

 rate of recombination of ions after they have once been 

 uniformly distributed in the field. The number of ions 

 undergoing recombination in the stream of gas at any instant 

 depends, of course, on the density of the ions. Thus we may 

 have a lack of saturation of any magnitude, and in this way 

 both Rutherford's and Madsen's results can be explained. 



On the other hand, experiments of this kind are liable to 

 a source of error, if no special precautions are taken. The 

 stream of ionized air passed between two plates at a difference 

 of potential, must necessarily pass first through a field whose 

 intensity is less than that which exists exactly between the 

 plates, and a number of ions, depending on the Telocity of 

 flow of the gas, may have readied the plates before they 

 reach the maximum field. An increase of potential might 

 then give rise to an increase of current on account of the 

 velocity of the ions becoming greater in this weaker part of 

 the field, and general recombination therefore becoming less. 



In fig. 3 I have plotted ex. and X ray current-voltage curves. 

 The curve Q was plotted from aluminium-chamber reading.-. 

 The part of the curve corresponding to low potentials has 

 been plotted because eaeh electrode was of the same material 

 on the outside, and the disturbing effect of the Yolta difference 

 of potential therefore small. 



Curves D and E were plotted from lead and uranium 

 chamber readings respectively, but were not plotted for low 

 potentials because here the Volta couples consisted of lead- 

 aluminium and lead-brass respectively , and the curves there- 

 fore were not comparable with the curve Q or with one 

 another at low voltages. 



The lack of saturation of current corresponding to any 

 point on the a ray curve, indicated by the curve's slope, is 

 very marked. The curve D (from lead chamber) has a slightly 

 greater slope than the curve E, which is probably due to the 

 ionization in the chamber not being uniform. 



The initial recombination was next studied in some gases 

 and mixtures of a gas and air. 



Since no difference appeared in the behaviour of primary 

 and secondary X rays with respect to the initial recombina- 

 tion of the ions made, secondary X rays have been used 

 throughout these experiments. 



The uranium chamber used in the previous experiments 

 (fig. 2), being provided with two tubes communicating with 

 the interior, was used as the adjustable chamber by fixing an 

 aluminium plate with a suitable lead guard to the bottom. 



