Charglmj thvoiujh Ton Ahsorpt'ion. 591 



uhon a mixture of onlinarv air ^vitll positive and negative 

 ions moves along a tube. Every ion ^vhicll strikes the con- 

 tluctino- \vail is absorbed, this leaves an ion-free space within 

 molecular distances of the wall, into this space positive and 

 negative ions ditt'use by the ordinary laws of gas diffusion. 

 The greater the coefticient of diffusion the more ions will 

 reach this s])ace, and so a greater absorption of ions will take 

 })lace during the passage of the mixture along the tube. By 

 measuring the ratio of the quantity of positive ions in a given 

 mass of gas before and after it had passed along tubes of 

 different length, Townsend was able to calculate the rate of 

 diffusion of positive ions, similarly for negative ions. The 

 values of the coefficient of diffusion obtained by this method 

 agreed very well with Zeleny^s number for the " mobility.'''' 



Villari (Phil. Mag. [6] i. p. 535, 1901), contrary to Zelenj^ 

 found that metals exposed co a stream of ionized air became 

 positively or negatively charged according as the air streamed, 

 over them with great or little friction. 



Starting with the work carried so far, my first experiments 

 w^ere directed to finding an explanation for the apparent dis- 

 agreement bet^veen Zeleny^s and Villari's results. In order 

 to do this it was necessary to repeat their experiments, but 

 this was done by a slightly altered method. A straight tube 

 40 cms. long was connected to a box in which air could be 

 ionized by means of Rontgen rays and then passed along the 

 tube. 



X'Y^/'^'^X V//7Z^^ 



The tube was composed of four pieces : the first connected 

 to the box being of glass and 10 cms. long, the second a 

 metal tube 5 ems. long, the third a glass tube 20 cms. lono-, 

 and finally another metal tube 5 cms. long. It was then 

 found that when ionized air was passed through the tube the 

 metal piece nearest the box became charged negatively, the 

 one at the far end positively. This showed that for the first 

 few moments after leaving the place of ionization the air 

 gave up a negative charge, while later on it gave up a 

 positive charge. This agrees with Zeleny^s experiment, for 

 the metal he used was situated so near the place where the 

 air v\'as ionized that it was passed over during the time the 



