56 RESEARCHES UPON ATOMIC WEIGHTS. 



shown evidence of great purity. For details concerning the purification the 

 papers referred to should be consulted. 



Sample A was employed in a determination of the atomic weight of 

 iodine.' This specimen had been twice precipitated as chloride and once 

 electrolyzed. 



Sample B was used in experiments upon the atomic weight of iodine ^ and of 

 manganese.' It was precipitated once as chloride, electrolyzed once, and finally 

 precipitated as metal with ammonium formate. 



Sample C also was employed in a determination of the atomic weight of man- 

 ganese, and was purified by recrystallizing silver nitrate, 7 times from nitric 

 acid and 5 times from aqueous solution. Finally the silver nitrate was reduced 

 by means of ammonium formate. 



Sample D was prepared for the determination of the atomic weights of cad- 

 mium ^ and manganese, by precipitation as chloride, precipitation with am- 

 monium formate, and electrolysis. 



Sample E was first purified in part by precipitation as chloride, in part by 

 precipitation with ammonium formate. The combined material was then 

 subjected to two electrolyses. 



In all cases the electrolytic crystals were fused in a boat of the purest lime, 

 contained in a porcelain tube, in a current of electrolytic hydrogen. After the 

 buttons had been cleansed with dilute nitric acid and dried at 200°, they were 

 cut into fragments of from 4 to 8 gm. either by means of a clean chisel and anvil 

 or with a fine jeweller's saw. The latter method was employed in the case of 

 samples D and E, because it proved easier completely to free the silver from 

 surface contamination with iron by etching the fragments with nitric acid, than 

 when a chisel was used. The cleansing process with nitric acid was repeated 

 until the solution thus obtained, after precipitation with hydrochloric acid and 

 evaporation, proved free from iron. That every trace of iron could be removed 

 by this treatment was proved by testing for iron the evaporated filtrates from 

 several of the analyses subsequently recorded in this paper. Negative results 

 were obtained in all cases. 



After thorough washing with water and drying at 100°, the pieces of metal 

 were heated to about 400° in a vacuum, and were preserved over solid potas- 

 sium hydroxide in a desiccator. 



* Baxter: Proc. Amer. Acad. ,40, 420 (1904); Jour. Amer. Chem. Soc, 26, 1578; Zeit. anorg. 

 Chem., 43, 15. (See page 92.) 



2 Baxter: Ibid., 41, 79 (1905); Jour. Amer. Chem. Soc, 27, 881; Zeit. anorg. Chem., 46, 

 42. (See page 108.) 



' Baxter and Hines: Jour. Amer. Chem. Soc, 28, 1560 (1906); Zeit. anorg. Chem., 51, 207. 

 (See page 37.) 



* Baxter and Hines: Jour. Amer. Chem. Soc, 28, 772 (1906); Zeit. anorg. Chem.^g, 417. 

 (See page 6.) 



