﻿Radium Content of Secondary Rock*. 



283 



Comparison of Electroscopes. 



! Scale of Scale of 

 1st Elec. in '2nd Elec. in 

 scale-divs. , scale-divs. 

 per hr. per hi*. 



By means of a granite repeatedly pre 

 viously tested, yielding closely agreeing 

 results ■•■ 



By means of the fusion of a granite 

 powder - 



By means of the fusion of a granite 

 powder 



By means of the introduction of the 

 same volume of radioactive gases 

 drawn from below and near to the 

 surface of the soil 



By means of a Cretaceous Chalk 



By means of a Cretaceous Chalk 



3] 

 35 



28 



21 



133 



12-5 



48 

 24 

 24 



21 

 164 



M 1 



Constant of 



2nd Elec. 



assuming 



1st to be 



06. 



044 



0-02 

 0-70 



060 

 0-38 

 0-3 



Mean 0\56 



In these experiments, which on the whole are in fair agree- 

 ment, the mean value of the constants appears to be about 06 ; 

 the tendency both by direct and comparative experiment being 

 slightly above that figure in the case of one electroscope 

 (0 = 0*66) and slightly below it in the other (0 = 0*53). 



The mean result of six comparative experiments on the 

 assumption that 0*6 represented the constant of the first 

 electroscope yielded for the second the figure 0'56. It was 

 therefore considered that by taking 0*6 x 10 ~ 12 gram radium 

 as equivalent to a gain of one scale-division per hour for 

 both instruments, there would be no error on the side of 

 excess, and in all probability this figure is as nearly correct 

 as certain risks of error pointed out by Prof. Joly as un- 

 avoidably incurred in the use of uraninite as a standard will 

 permit. It is, however, satisfactory to know that this was 

 the figure arrived at for the constant of the electroscopes in 

 numerous determinations on the standard aqueous solution of 

 Prof. Rutherford. As additional testimony to the truth of this 

 figure and of the improbability of any considerable source of 

 error dependent upon u personal equation," we have the mean 

 value of nine experiments performed by Prof. Joly on an 

 electroscope of equal dimensions and identical in construction. 

 This mean comes out as 0*64. Finally, the reliability of the 

 electroscopes in their very slight diurnal variation renders it 

 practically certain that the following results are in the great 

 majority of cases within 10 per cent, of their true value, and 

 in view of the number of results from which the mean value 

 is calculated, the percentage error is in this case in all 

 probability even below this figure. 



