Air in Closed Vessels and its Causes. 575 



Thus in the second experiment there is no appreciable 

 increase of the leak as a result of heating the granulated tin. 

 This makes it probable that the increase previously observed 

 was due to free ions, and as a proof of the existence of an 

 emanation the experiment breaks down. All that one is 

 justified in concluding, however, is that the existence of the 

 emanation has not been proved ; as it is still perfectly open 

 to us to conclude that those free ions, which are the immediate 

 cause of the increase of the leak, have themselves been pro- 

 duced by the action of an emanation, the life of which is so 

 short that it is disintegrated before reaching the vessel. The 

 same experiments were performed on lead, and in this case 

 no increase could be found, either with or without the field. 

 Xow lead -is the very metal of all those tried in which one 

 would look with the greatest confidence for an emanation ; 

 because in the first place it is a heavy metal, and therefore 

 likely to be radioactive, and in the second place the intrinsic 

 radiation for lead was found to be greater than for any other 

 metal. 



Although no direct evidence of the emanation has been 

 found, the following experiments admit of a simple explana- 

 tion on the emanation hypothesis. On one occasion when a 

 lead vessel which had been in use was cleaned with nitric 

 acid and set up afresh, the natural leak was found to be 

 distinctly increased. If the surface of the lead was now 

 scraped and the vessel replaced, the leak was found to have 

 resumed its normal value. This increase could be produced 

 again and again ; after each treatment with acid, the vessel 

 was carefully washed in running water for half-an-hour or 

 more, and then thoroughly dried in a current of hot air, and 

 each time there was a marked increase in the natural leak. 

 This increase of course varied very much — in some cases 

 being as much as 300 per cent . and in no case being less 

 than 30 per cent., although occasionally as low as that. In 

 order to study this effect more carefully a piece of thin sheet- 

 lead was made into a lining for the zinc vessel. It was first 

 polished with fine emery and then placed inside the zinc 

 [. A mean of six observations of the natural leak in the 

 zinc vessel with this lining gave <>-27>\ The lead lining was 

 now removed and treated for a few minutes with strong 

 nitric acid. It was then carefully washed, dried, and replaced. 

 The mean leak was found to have increased to 0*328, showing 

 an increase of about 1 < s per cent. The polishing process was 

 now repeated, and the natural leak went back to ()-'22>>, while 

 after another treatment with nitric acid it rose again to 0*296. 

 The effect of hydrochloric acid on tin was next tried. The 



