Air in Closed Vessels and its Causes. 573 



The ebonite was replaced by cork, as this latter was a 

 sufficiently good insulator for that part o£ the apparatus, and 

 was less apt to charge than the ebonite. The wall of the 

 vessel Avas carried up in the form of a tube and the insulation 

 put at the mouth of the tube, the earthed brass tube projecting 

 through the cork down into the vessel, and so screening the 

 electrode. If the vessel after being charged was now earthed 

 and heated, there was no apparent leak in any direction. 

 Xevertheless, when the experiments were repeated, it was 

 found that if the vessel was heated a distinct increase of the 

 leak could be observed. The next point was to ascertain 

 whether this increase was due to an emanation. A current 

 of air from outside was drawn through (a) a glass-wool plug 

 to free it from dust ; (6) a drying-tube rilled with calcium 

 chloride ; (c) a tube containing granulated tin ; and, lastly, 

 through a vessel in which the leak could be measured. A 

 rapid, steady current of air was passed through the apparatus 

 and the leak measured in the tin vessel. The tube containing 

 the granulated tin was now heated with a bunsen, and the 

 ionization in the tin vessel showed a distinct increase. The 

 following are the figures for one experiment. The numbers 

 given in the column headed " Ionization '' are the voltages 

 to which the electrode system was charged up in ten minutes 

 from the time indicated in the first column. The numbers 

 in the second column are the potentiomer readings. 



Time. Poter 



9.2(H 

 9.32| 

 0.44 

 9. 55 

 10.6^ 



Glass tube containing the granulated tin gently heated 10.18^ to 10.21. 



10.18 730 0844 



Glass tube heated to the melting-point of tin 10.29 to 10. 36. 



10.29 



10.10 



10.61| 



11.21" 



Glass tube heated again to the meltin 

 11.14 

 11.27 

 11.48 



It will be observed that in each case after heat was applied, 

 the leak in the measuring-vessel showed a. distinct increase. 

 This increase only lasted lor a very short time, but it must 

 be remembered that the' correal of air was passing all the 



leter Readings. 



Ionization. 



710 



0-830 



750 



0-857 



750 



0-857 



630 



0-773 



685 



0-813 



880 







0-936 



650 







0-788 



750 







0-857 



770 







0-870 



elting 



-point 



of tin 



11.14 to 11.17. 



900 







0-947 



780 







0-876 



760 







0-804 



