Bumstead and Wheeler — Radio-active Gas. 107 



diffuse more slowly than the radium emanation and later 

 it diffuses more rapidly. It is difficult to see how this could 

 be the effect of an admixture of either a heavier or a lighter 

 radio-active gas; and the close agreement between the total 

 amounts, diffusing out in four hours (89 per cent and 88 per 

 cent, respectively), together with the close similarity in the rates 

 of decay and of the production of induced activity, show that 

 if any other radio-active constituent is present in the ground 

 air it must be in very small proportion. 



Density of the Radium Emanation. 



The observations already recorded afford a value of the den- 

 sity of the gas (assuming that Graham's law may be applied) 

 provided the rate of diffusion of a gas of known density 

 through the porous plate is determined. For this purpose 

 carbon dioxide was introduced into the cylinder, allowed to 

 stand until thoroughly mixed with what air remained, and the 

 amount determined by drawing a certain fraction into a burette 

 and absorbing with caustic soda. The porous plate was then 

 exposed for 10 or 15 minutes and the bottom plate again 

 clamped on ; after allowing time for thorough mixture, the 

 amount of C0 2 was again determined. Three such observa- 

 tions, with different times of diffusion, gave the following 

 values for fi lf in the formula p=p e—, a i t (t measured in hours): 



M8 

 1-15 

 1-15 



1-16 



In the diffusion experiments with the radio-active gas, if 

 it were not for the presence of the induced activity, the curves, 

 figs. 6 and 7, might be expected to be exponential and it would 

 be a simple matter to obtain the corresponding constant, /*, for 

 the radium emanation; but the effect of the induced activity 

 must be taken into account, and this is the more difficult since 

 its rise and decay is not susceptible of a simple mathematical 

 expression. In fact, as Rutherford has pointed out, there are 

 probably two or three different kinds of induced activity, in 

 the case of radium, each being produced and decaying at a 

 different rate. In order to make the calculation somewhat 

 more manageable, it is assumed in what follows that the rise 

 and decay of the induced activity does follow the exponential 

 law ; it will be seen in the outcome that this assumption cannot 

 cause a very serious error in the determination of the diffusion 

 constant. 



