178 PROCEEDINGS OF THE AMERICAN ACADEMY 



cient care, trilithium phosphate free from sodium ; but the point was 

 made, that the phosphate is apt to be contaminated with lithium cai'- 

 bonate when sodium carbonate is employed to bring about alkalinity. 

 Mayer therefore modifies the method of Berzelius by substituting 

 sodium hydrate for the carbonate ; and, proceeding, evaporates to dry- 

 ness, treats the dry mass with as much' water as is needed to dissolve the 

 soluble salts with the aid of heat, adds a drop or two of sodium hydrate 

 if necessary to restore alkalinity and then ammonia in volume equal 

 to that of the water already added, sets aside at a gentle heat, filters 

 only after twelve hours, and washes with a mixture of ammonia and 

 water in equal parts. From the filtrate and first washings a small 

 amount of the lithium phosphate is to be recovered by evaporation 

 and the repetition of the former treatment. According to Mayer, 

 the precipitation of the phosphate may be effected with equal com- 

 pleteness by boiling the solution, prepared as before, instead of evapo- 

 rating it ; but the objection to this mode of proceeding is the tendency 

 of the liquid carrying the precipitate to bump explosively. Careful 

 washing, somewhat prolonged, is essential to secure the complete 

 removal of salts of sodium and potassium, and it is remarked that the 

 purity of the precipitate is shown by its failure to cake when strongly 

 ignited. 



This is the mode of proceeding by which Mayer separates lithium 

 from sodium and potassium, isolating it as presumably pure tri-lithium 

 phosphate and weighing it as the anhydrous salt. In dealing with 

 mixtures of the chlorides in which the proportion of the lithium salt 

 is relatively small, the removal of the greater part of sodium and 

 potassium chlorides by a preliminary treatment with absolute alcohol 

 is recommended. The following table comprises the results of Mayer's 

 test analyses of lithium carbonate in the first seven, of lithium sulphate 

 in the last two, recalculated with the use of the number 7 — the figure 

 now generally accepted — as the atomic weight of lithium. 



