352 /. W. Mallet on the Atomic Weight of Lithium. 



Authority. 



Salt analyzed. 



At. Weight of Li. 



ArfVedson 1st experiment, 



Sulphate. 



129-30 



2nd 





135-22 



3rd 



Chlorid. 



126-71 



Vauquelin, 



Sulphate. 

 <( 



116-17 



Gmelin 1st " 



135-54 



" 2nd " 



Sulphate conv. into Carbonate. 



108-22 



Stromeyer, 



Sulphate. 



122-74 



Kralovanszky, 





131-40 



Hermann 1st " 



Carbonate. 



75-97 



2nd 



Sulphate. 



75-82 



3rd 





75-67 



Berzelius 1st " 



Carbonate conv. into Sulphate. 



80-89 



2nd " 



Sulphate. 



83-01 



Hagen, 





82-41 



Of these sufficiently discrepant numbers, those of Arfwedson, 

 Vauquelin, Gmelin, Stromeyer, and Kralovanszky are at once 

 to be rejected, for the reason already mentioned, that the sub- 

 stance analyzed by each of these chemists was not in reality a 

 pure salt of lithia, but a mixture of salts of lithia and soda. 

 The determination by Hermann of the amount of carbonic acid 

 in the carbonate was not conducted in a manner likely to yield 

 the most accurate results, and probably the same objection may 

 be made against Berzelius's first experiment, which was made 

 with the same salt. 



We find then but three numbers apparently deserving of much 

 confidence, namely those deduced from the analysis of anhydrous 

 sulphate of lithia by Hermann, Berzelius, and Hagen — one ex- 

 periment by each of these chemists. Hermann's result differs 

 very considerably from the other two, and, as is observed by M. 

 Oudemans, should have less importance attached to it, since the 

 original weighings are not recorded, but merely the per-centage 

 of sulphuric acid deduced therefrom by the analyst. Berzelius's 

 result has been generally taken of late years as the true one, and 

 with it that of Hagen agrees pretty well. 



We have thus the atomic weight of Lithium as the result of 

 but two experiments, agreeing, it is true, fairly with each other, 

 but both made by the same process — 'the precipitation of sul- 

 phate of lithia by a salt of baryta, washing, igniting, and weigh- 

 ing the BaO, SO 3 produced. And as it is well known that sul- 

 phate of baryta is washed with extreme difficulty, an excess of 

 the barytic salt used for precipitation obstinately adhering to the 

 precipitate even after a very large quantity of hot water has been 

 filtered through it — this fact rendering the exact determination 

 of the sulphuric acid difficult ; and as the quantity of sulphate 

 of lithia submitted to analysis in each of these experiments was 

 but small, and therefore the effect of any trifling error in the 

 estimation of the sulphuric acid would be more seriously felt — 

 it seemed that the equivalent of the metal might be redetermined 



