Mr. 0. D. Allen on Casium and Rubidium, 191 



small admixture of sulphates, of sodium, lithium, potassium, 

 rubidium, and caesium. This quantity of salts, subjected to 

 Bunsen's process of fractional precipitation with bichloride of 

 platinum, furnished 132 grammes of the platinochlorides of 

 caesium and rubidium, in which no potassium could be detected 

 with the spectroscope. The platinochlorides were very gently 

 heated in a current of hydrogen gas until complete reduction of 

 the platinum took place, and the chlorides were then extracted 

 with water. 



The percentages of caesium and rubidium obtained from the 

 mineral by this process were calculated from the amount of 

 chlorine contained in these mixed chlorides. 



0*5825 grm., dissolved in water and precipitated with nitrate 

 of silver, gave 0*5835 grm. of chloride of silver, which represents 

 0*1439 grm. of chlorine. 



These numbers furnish the following equations : — 



Rb + Cs = 0*5825 -0-1439, . . . . (1) 



Rb Cs 01439 



85*36* + 123*35* ~ : 35*5 ' * ' ' * **} 



which give Cs = 0*3002 and Rb = 0*1384. According to these 

 proportions, the 132 grammes of platinochlorides contained 

 31'1969 grammes of csesium, 14*3826 grammes of rubidium, 

 which numbers respectively correspond to 0*3 per cent, and 0*14 

 per cent, of the mineral employed. 



It appears, therefore, that it is practicable to extract almost 

 one-half per cent, of the two metals from the Hebron lepidolite, 

 even when operating on a large scale, and in a somewhat crude 

 manner. In separating the platinochlorides of csesium and rubi- 

 dium from the platinochloride of potassium, a not inconsiderable 

 amount of these metals went into solution with the potassium- 

 salt, thus materially diminishing the quantity obtained. Much 

 the larger proportion of this loss was rubidium, due to the greater 

 solubility of its platinochloride. On comparing these results 

 with Cooper's analyses f of the Rozena lepidolite, it appears that, 

 although not quite so rich in rubidium, the Hebron mineral is 

 remarkably rich in caesium. The lepidolite from Rozena and 

 Zinnwald contain, according to the published analyses, only an 

 unweighable trace of csesium, while that from Hebron contains 

 more than three-tenths of 1 per cent. 



Experiments in separating Casium and Rubidium. — The process 

 described by Bunsen for separating the new alkalies appeared to 

 be so troublesome, requiring for the preparation of pure rubi- 



* Combining proportions of csesium and rubidium determined by Bunsen 

 (Poggendorff's Annalen, vol. cxiii. p. 339). 

 t Journ. Prakt. Chem. vol. lxxxv. p. 125. 



