Mil. I.I OF POSITIVELY AND NEGATIVELY CHAI;<;K1> [ON8 



L'-.H 



consist in the breaking up of the neutral molecules into a certain number of pnsiti\ t - 

 ions and half as many negative inns, each carrying twice as large a charge as the 

 |i'>siti\e. It is plnin. therefore, that there must at any moment be a great excess of 

 the ions which have the greater distance to travel ; in other words, of the ions charged 

 v\ ith electricity of the same sign as that on the plate nearest the layer of air exposed 

 to the rays. The expansion may either U- made while this layer is ex post -d to the 

 rays, or the rays may l>e cut off l>efbre the expansion. If the interval, l>etweeii 



Fig. I. 



To Pump 



cutting off the rays and making the expansion, lies witlu'n certain limits, it is plain 

 that all the ions travelling to the plate next the ionised layer may have been removed, 

 while only a small proportion of those travelling towards the more distant pkte have 

 reached it before the expansion is made. In this way we would therefore expect to 

 get posit i\e or negative ions with almost complete absence of ions of the other kind. 



The method of producing sudden expansion of any desired amount was that which 

 I have described in a previous paper.* Such differences as there were in the detail-; 

 of the apparatus are sufficiently indicated in fig. 1. The glass cylinder and piston of 



* 'Camb. Phil. Soc. Proc.,' vol. 9, p. 333, 1897. 

 2 P 2 



