110 Bumstead and Wheeler — Radio-active Gas. 



total), the value of />& obtained from equation (5 7 ) when substi- 

 tuted in (4) will not lit the observations. On the other hand, 

 if it be assumed that between 50 and 60 per cent of the total 

 activity is due to the gas, a reasonably good fit can be secured ; 

 and this value of A is also indicated by a special experiment 

 under better conditions in which a larger quantity of radium 

 emanation was admitted rapidly ; but even under the best con- 

 ditions with the electrometer used, readings could not be 

 obtained oftener than every 2-J minutes, and thus the principal 

 weight in the selection of a value for A must be given to the 

 diffusion curve. It is to be noted that the constant, a in (4), 

 as determined from experiment, depends on the value chosen 

 for A . 



In the following table the second column is calculated from 

 equation (4) with A = 56, X = l*36, /a =0*58, which value gives 

 a better lit than either 0*57 or 0'59. The third column gives 

 the experimental values for the radium emanation as deter- 

 mined from the curve, fig. 7. 



Time in hour^. 



Cale. 



Curve. 



Diff. 



1 



6T-5 



66*0 



+ 1-5 



2 



40'2 



41-2 



— 1-0 



3 



23-0 



237 



-0-7 



4 



12 9 



12-0 



— 0-9 



The differences are considerably less than the differences be- 

 tween the experiments for the earth-gas and the radium emana- 

 tion ; for the former, /a=0*56 appears to give the closest agree- 

 ment. If we take 0*57 as the value from the two experiments 

 and assume that Graham's law may be applied, we find that 

 the density of the gas is 4*1 times that of carbon dioxide, which 

 would* give it a molecular weight of 180. 



As our apparatus was conveniently arranged for use with 

 radio-active gases, we attempted to determine the properties 

 of the active gas recently obtained by Strutt* from metallic 

 mercury. We were, however, unable to obtain any evidence 

 of radio-activity from air which had been circulated through 

 hot mercury for fourteen hours ; and an increase of ten per cent 

 over the '"spontaneous" ionization in the cylinder could cer- 

 tainly have been detected. This would seem to indicate that 

 the gas observed by Strutt was due to some radio-active impurity 

 in the mercury which he used. 



In connection with the evidence of the existence of minute 

 quantities of radium in the ground in this vicinity, it is of 



* Strutt, Phil. Mag., July, 1903. 



