ATOMIC WEIGHTS 



151 



From the AgCl ratio, Eb = 85.446. 



From the Ag ratio, Eb = 85.426. 



And Ag: CI:: 100: 32.847. 



The values for the AgCl ratio combine as follows : 



Bunsen 84.253, ± .031 



Piccard 84.290, ± .0105 



Godeffroy 84.3345, ± .0051 



Archibald 84.3485, ± .0014 



General mean 84.3433, ± .0013 



Heycock's single value for the Ag ratio, reduced to the usual standard, 

 becomes Ag: EbCl:: 100: 111.926. It is not worth while to combine it 

 with Archibald's values, for its influence would be quite negligible. In 

 the AgCl ratio the older determinations coimt for something, but the 

 general mean falls witliin the range of variation of Archibald's series. 



The bromide analyses by Archibald are as follows : 

 Weight RbBr. Weight AgBr. Weight Ag. AgBr ratio. A g ratio. 



2.68170 3.04578 1.74930 88.047 153.301 



2.07280 2.35401 1.35230 88.054 153.280 



2.10086 2.38589 1.37061 88.053 153.278 



2.61044 2.96462 1.70300 88.053 153.285 



3.84082 4.36215 2.50590 88.049 153.272 



3.77852 4.29084 2.46502 88.061 153.287 



4.34299 4.93210 2.83340 88.056 153.278 



Mean, 88.0533, 153.283, 



± .0012 ± .0024 



From tlie Ag ratio, Eb = 85.442. 

 From the AgBr ratio, Eb = 85.444. 

 And Ag:Br:: 100: 74.080. 



Heycock's mean for the Ag ratio, reduced, becomes Ag : EbBr : : 100 : 

 153.238, ±.0300. Its probable error is so high that combination with 

 Archibald's data would be useless. 



There are now four ratios from which to compute the atomic weight 

 of rubidium : 



(1). AgrRbCl: : 100: 112.0545, ± .0016 

 (2). AgCl : RbCl :: 100: 84.3433, ±: .0013 

 (3). Ag: RbBr:: 100: 153.283, ±.0024 

 (4). AgBr:RbBr: :100:88.0533, ± .0012 



Eeducing these ratios with Ag= 107.880, ±.00029, CI = 35.4584, ± 

 .0002, and Br = 79. 9197, ±.0003, we have— 



From ratio 1 Rb = 85.426, ± .0018 



" 2 85.438, ± .0019 



" 4 85.442, ±.0026 



" 3 85.444, ± .0023 



General mean, Rb = 85.436, ± .0010 



