Liglit by the Recombination of Ions. 141 



50 ions may be drawn to a and 50 may recombine between 

 a and b ; 



(c) When a is uncharged, b is positive, and c is negative, 

 50 may recombine between a and b, 25 may be drawn to b 

 and 25 may recombine between b and c. 



We have here met all the conditions of Strutt's experiments 

 and yet have exactly the same number of ions recombining 

 between b and c when a current flows between those points 

 •as when there is no current. These conditions seem more 

 reasonable when one remembers that in the first case not all 

 o£ the negative ions passing a recombine between a and b, 

 while in the second case all that pass a must somehow re- 

 combine before they reach b, since none of them go to b and 

 none of them pass that point. 



But a further consideration of the subject shows that it is 

 not only possible that the rate of recombination is unaffected 

 by the passage of a current, but that in the greater part of 

 this region a condition approximating this must exist. When 

 h is positive and c is negative the field near b is very small, 

 as has been shown by myself (p. 908) and by Strutt (p. 94). 

 Since the positive ions are of the same mass as the mercury 

 atoms, this small field will give to the positive ions a velocity 

 very much smaller than that of the stream of vapour, a 

 velocity which has been estimated by Stark to be 28,000 cm. 

 per sec* The same thing is apparent when we remember 

 that the effect of the field on the positive ions must be 

 very much less than the effect on the negative ions, and yet 

 the effect on the negative ions is not sufficient to draw all of 

 them against the stream of gas. That they are not all drawn 

 out of the region is evident from the nature of the field. If 

 .a current of one milliampere were being carried by the 

 positive ions only, a field very many times greater than that 

 'existing would be necessary. 



The same number of positive ions pass b whether it is 

 charged positively or is uncharged. If the same number 

 passes this point and their velocity is approximately inde- 

 pendent of the field, then the number of positive ions per 

 unit volume is also approximately independent of the field. 



But the number of negative ions per unit volume is very 

 approximately the same as that of the positive ones, other- 

 wise the potential difference between b and a point 1 cm. 

 from it would be much more than that which the experiment 

 shows. If, for example, we were to assume a current of 

 06 milliamoere per sq. cm., a velocity of the positive ions 



* Phys. Zeit. iv. p. 440 (1903). 



