CESIUM. 25 



Johnson and Allen : 133 (0 = 16). 



Determined by four experiments on the precipitation of 

 ctesium chloride with argentic nitrate. The mean result 

 was Cs = 133.036 ; the extreme difterence, 0.842. Ag = 

 107.94; CI = 35.46. Caesium and rubidium were separated 

 by partial crystallization of their bitartrates. The cgesium 

 bitartrate was converted into chloride by precipitation with 

 platinum chloride, reduction and solution. The nitrate 

 formed on the precipitation of the caesium chloride with 

 silver was reconverted into caesium chloride and redeter- 

 mined, and so on. The purity of the salt w^as tested spec- 

 troscopically. {Silliman's Amer. Joiirn., (2,) oS, 1863, 96.) 



R. W. BuNSEN: 133 (0 = 16). 



Determined by three experiments on the precipitation of 

 caesium chloride with argentic nitrate. The mean result 

 was 132.99; extreme difterence, 0.02. Ag = 107.94; CI 

 = 35.46. In order to prepare pure chloride, a mixture of 

 caesium and rubidium salts was converted into carbonates, 

 a little more tartaric acid was added than was necessary to 

 form acid tartrate with the rubidium and neutral tartrate 

 with the cfBsium, and the mixture was exposed on a filter 

 to the action of a saturated atmosphere of aqueous vapor. 

 The caesium salt is deliquescent, and gradually passes 

 through the filter, while the rubidium salt is unafl:ected. 

 The caesium tartrate was turned into pure chloride b}^ re- 

 peated precipitation with platinum chloride, reduction in 

 hydrogen and solution. The determinations were made on 

 the product of successive purifications, and only those were 

 taken into consideration which were made after analysis 

 showed a constant composition. The spectroscope was em- 

 ployed to test the purity of the salt. [Poggend. AnnaL, 119, 

 1863, 5.) 



— . Mercer : 133 (0 = 16). 



The fact of this determination, without details, is men- 

 tioned by Frankland. {Chem. Nexus, 8, 1863, 18.) 



R. GoDEFFROY : 132.557 (0 = 16). 



Derived from the mean of four analyses of caesium chlo- 

 ride with argentic nitrate, the extreme difterence being 

 0.185. CI = 35.5 ; Ag = 108. The caesium was separated 

 from the other alkalies by the fractional crystallization of 

 their alums continued until the caesium compound was 



