576 Ionization of Ait in Closed Vessels and its Causes. 



same zinc vessel was used, but this time it was lined with a 

 thin sheet of carefully polished tin. The mean leak was 

 now 0*281 ; the lining was removed, treated with strong 

 hydrochloric acid, washed, dried, and replaced. The leak 

 was found to have increased to 0*366 — an increase of more 

 than 25 per cent. Repolishing reduced the leak again to 

 0*315. The acid was in each case tested for radioactivity, 

 but without success. Now by no known process can radio- 

 activity be produced, and it seems extremely improbable that 

 the treatment of metals with acids is any exception to 

 this general rule. Another possible explanation suggests 

 itself, however. In a paper communicated to the Phil. 

 Mag. recently, Mr. Beilby* has gone very carefully into 

 the changes produced in the surface of a metal by polishing. 

 He concludes that a film is drawn over the surface, filling 

 up the interstices between the crystals of the metal. This 

 film is removed by treatment with acids. It seems not 

 improbable, therefore, that this surface-film hinders the 

 escape of emanation from the metal, and so reduces the 

 leak in the vessel. The action of acids, on the other hand, 

 dissolves off this film, and so allows the emanation to have 

 free access to the air in the vessel. An extension of those 

 experiments is at present being carried out. Further indirect 

 evidence of the existence of an emanation is supplied by 

 some experiments of McLennan f, confirmed in the course 

 of the present research. A metal vessel was filled with fresh, 

 dry, dust-free air and the leak measured immediately. 

 Measurements of the leak were also taken subsequently at 

 short intervals, and it was found that the leak at first fell and 

 then rose to a value higher than its initial value, at which it 

 remained constant. The initial decrease was attributed to 

 the decay of emanation drawn from the air, and it is difficult 

 to account for the subsequent increase of the leak on any 

 other hypothesis than the escape of emanation from the metal, 

 especially as it differed both in time-constant and amount for 

 different metals. All this evidence is of course indirect, but 

 taken in conjunction with the argument founded on the 

 agreement of the results obtained by different observers for 

 the ionization produced by the various metals tried, it points 

 strongly to the conclusion that all matter is radioactive and 

 that the disintegration processes going on in radium and the 

 other radioactive elements are going on also, though to a 

 very much smaller extent, in all matter. 

 Cavendish Laboratory, Cambridge. 



* Phil. Mag. August 1904. t Phys. Rev. iv. 1903. 



