QUANTITATIVE SEPARATION OF LITHIUM. 11 



BuNSEN * in the analysis of some alkali residues from lepidolite 

 determined lithium as well as rubidium indirectly. He said that the 

 determination by extraction of the mixed chlorids hy ether-alcohol 

 was as inaccurate as the determination as phosphate. His indirect 

 method has not l)een used, however, by many other investigators. 

 Like most indirect methods, there is a nuiltiplication of error on the 

 smaller constituent when one of the substances to be determined is 

 present in much smaller quantity than the other. In addition, as 

 Gooch ^ notes, there is the uncertainty in the total weight due to the 

 behavior of the lithium chlorid when heated to dryness and the 

 possi])ilily of precipitation of lithium with the potassium as a double 

 cliloritl of lithium and potassium.' 



Fresenius * repeated some of the work of Mayer ^ and found the 

 objections of Rammelsberg ^ to be groundless. He evaporated to 

 dryness with sodium phosphate and sodium hydroxid, the solution 

 of the lithimn chlorid with small quantities of sodium and potassium 

 chlorids, separated from the greater part of the other alkaline chlorids 

 by alcohol. The residue, taken up with a Uttle water, was allowed 

 to stand 12 hours after the addition of a little ammonia and the 

 precipitate then separated by filtration and washed with equal parts 

 of strong ammonia and water. He found on evaporating the filtrate 

 and wash water and taking up the residue that a weighable precipi- 

 tate of lithium phosphate remained even on the fourth evaporation. 

 These analyses led him to agree with Mayer that it consisted of 

 tri-basic lithium phosphate (LigPO^) without any sodium. The 

 precipitate as formed contained water, but dried at 100° C. or gently 

 ignited gave the lithium phosphate. He stated^ that if the lithium 

 phosphate left a residue insoluble in dilute hydrochloric acid, this 

 should be separated, weighed, and its weight subtracted from that of 

 the phosphate precipitate, and that if the solution moderately diluted 

 gave a precipitate with ammonia, this should be separated and its 

 weight taken from the original weight of lithium phosphate. This 

 method, given in the text-book of Fresenius, was in use in the Fre- 

 senius laboratories at Wiesbaden in 1908.^ 



Kraut, Nahnsen, and Cuno ^ made analyses of phosphates of 

 lithium and sodium obtained by precipitating after the manner of 

 Berzelius ^^ from solutions containing varying proportions of lithium 



' Ann. Chem. Pharm., 18G2, 1S2: 347; 354. 

 ' Loc. cit. 

 'Jenzsch, loc. cit. 

 ♦ Zts. anal. Chem., 1862, /.• 42. 

 6 Loc. cit. 



» Pogg. Ann., 1857, 102: 441-3. 

 'Zts. anal. Chem., 1876, IS; 227. 



8 Investigation of Gertrudis-Spnulels, by 11. Fresenius. .fahrhiifher dPs nassaiiisohen \"erein.s ffir 

 Naturkimde, 1908, p. 61. 

 « Ann. Chem. Pharm., Is7ti, 18t- 165. 

 w Loc. cit. 



