36 THE ROYAL SOCIETY OF CANADA 



Assuming the water to contain no radioactive matter we may 

 suppose the ionisation in the electrometer to be given by 



(I - x) = do - x) e ~ 1'"^ 



where /■« is the coefficient of absorption of the penetrating radiation, 

 d the thickness of the water-screen and x the ionisation in the electro- 

 meter which is not due to the penetrating radiation. We have then 



(5-82 — x) = (8-05 — x) e ~" -" ^^'^ 



For soft gamma rays Rutherford* gives the value • 0558 for fi. and 

 for harder rays the value f • 0472. The measurements of Hess at Vienna 

 on the absorption or gamma rays by air also lead to a value for [i for 

 water of -0413. 



Applying these values we find the value of x to be as follows: 



fi x (ions per cc. per sec). 



•0558 5-48 

 •0472 5-33 

 •0413 5^19 



In so far as these measurements go then, it would appear that 5^19 

 ions per cc. per sec. is the lowest value one should have expected to get 

 for the ionisation in the zinc vessel when all the penetrating radiation 

 was cut off. 



In view, however, of the fact that 4 ions per cc. per sec. were ob- 

 tained experimentally the above experiments can only be considered 

 as giving a roughly approximate result. 



They serve to shew, however, that unless there is present at the 

 surface of the earth a radiation of a much higher penetrating power 

 than any yet observed, there must be a residual ionisation in air con- 

 fined in a zinc vessel (of about 4 ions per cc. per sec.) when it is not 

 traversed by any radiation from an external source. 



It would seem, therefore, that this residual ionisation must be due 

 either to i-adiation from a minute quantity of radioactive matter still 

 remaining in the zinc of which the receiver is made or to a process 

 going on in the gas which results in what may be called a spontaneous 

 dissociation of the molecules. 



If the residual effect were due to active matter in the walls of the 

 receiver one should expect that the value obtained for "q" would in- 

 crease if the volume of the receiver were diminished. Since the value 



* Rutherford, Radioactive Substances and their Transformations, p. 266. 

 fHess. Phys. Zeit. 12 p. 998, 1911. 



