38 THE ATOMIC WEIGHTS. 



discussion with the atomic weights deduced by Stas from 

 his own experiments only. His values are given under two 

 headings : one for oxygen = 16, the other for = 15.96. 

 As we have been using the figure 15.9633 for oxygen, here 

 is at the outset a discrepancy. Starting from this value we 

 found : 



Ag = 107.675, ±z .0096 



CI = 35-370, ±: .014 



Br = 79.768, it .019 ., 



I == 126.557, d= .022 



Na = 22.998, ± .011 



K = 39.019, ±: .012 



S = 31.984, =b .012 



If we assume 16 to be the true figure for oxygen, we get 

 the following results, which I have placed in a column par- 

 allel with the values found by Stas : 



T/ie A^eza Values. Stas. Differences. 



Silver 107.923 107.930 .007 



Chlorine 35-451 35-457 -006 



Bromine 79-951 79-95- -OO' 



Iodine 126.848 126.850 .002 



Sodium 23.051 23.043 .009 



Potassium 39-I09 39-137 -028 



Sulphur 32.058 32.074 .016 



These differences are insignificant. No other criticism 

 could more severely test the character of Stas' work, or more 

 definitely illustrate his magnificent accuracy of manipula- 

 tion. 



