[ 196 ] 



XXVI. On the Equivalent and Spectrum ofCasium. 

 By S. W. Johnson and 0. D. Allen*. 



IN the preceding article a method is described of separating 

 caesium from rubidium by fractional crystallization of the 

 bitartrates of these metals. The analyses of the bitartrate of 

 caesium there given, while perfectly according with each other 

 as regards carbon, disagree with the numbers deduced from 

 Bunsen's equivalent to such an extent that we have undertaken 

 to ascertain whether the salt was impure or Bunsen's equivalent 

 incorrect. 



From the great care used in preparing the bitartrate, and 

 especially from the fact that its spectrum remained unaltered 

 though the salt was repeatedly recrystallized, we were inclined 

 to suppose that Bunsen had not operated with a pure substance. 



This might easily happen, on account of the small quantity 

 of material at his disposal, without at all detracting from the 

 merit of this distinguished chemist. 



A quantity of bitartrate of caesium, purified by concentrating 

 its solution and recrystailization, as described in the paper re- 

 ferred to, and containing no foreign matters recognizable by the 

 spectroscope, except an inevitable trace of sodium and (to judge 

 from a certain red line) perhaps some lithium, was treated directly 

 with bichloride of platinum in quantity sufficient for complete 

 precipitation. This platinochloride of caesium, after thorough 

 washing, was reduced in hydrogen, the chloride of caesium dis- 

 solved from the platinum and evaporated to dryness with addi- 

 tion of a little hydrochloric acid. 



We thus obtained an amorphous mass of a pure white colour, 

 which, unlike Bunsen's chloride, was not perceptibly deliquescent 

 even in a very moist atmosphere. The spectrum of the chloride 

 thus prepared was identical with that of the original bitartrate. 

 Both salts gave a red line nearly coincident with the a line of 

 lithium. In order to determine whether this line was due to a 

 trace of lithium, or belongs to the spectrum of caesium, a por- 

 tion of chloride was again precipitated with a relatively small 

 quantity of bichloride of platinum, the precipitate was most 

 thoroughly washed, and from it a new sample of chloride of 

 caesium was prepared. The red line was no less apparent in 

 this than in the former preparations. The same process of 

 partial precipitation was repeated a third time without altering 

 the spectrum. 



Again, from a hot dilute solution of 15 grms. of chloride of 

 caesium, about 1 grm. of caesium was thrown down as platino- 

 chloride ; the product thus procured gave a spectrum identical 

 with that from the original bitartrate. 



* Communicated by the Authors : from Silliman's American Journal for 

 January 1863. 



