18 THE DETERMINATION OF LITHIUM. 



obtained. Of this amount the Hthium is usually not a very consid- 

 erable part and so any danger of loss of lithium with the barium sul- 

 phate precipitate may be safely neglected. 



Note. — In some experiments on the recovery of small amounts of lithium from a 

 mineral water the barium sulphate precipitate was extracted for 40 hours with 100 cc 

 of water and 10 drops of hydrochloric acid (specific gra\'ity 1.1). The liquid was 

 filtered off, evaporated to dryness, and taken up in 10 cc of water. WTien examined 

 with the spectroscope the lithium line was barely visible, indicating tlie presence of 

 about 0.00002 mg of litliium per cc, or 0.0002 mg in all. These precipitates were 

 from waters containing from 0.0004 to 0.0060 mg of lithium with no perceptible varia- 

 tion in the amount of lithium from the different samples. 



In the case of a water sample on which it is not desired to deter- 

 mine the sulphate the barium chlorid may be added to the original 

 water moderately acidified and the whole quantity evaporated to 

 dryness, taken up with water and a little hydrochloric acid, and 

 filtered. 



The filtrate from the barium sulphate is evaporated to dryness 

 and gently ignited to remove ammonium salts. The residue is 

 then taken up in from 50 to 200 cc of water, according to the amount 

 of alkali salts present, and milk of lime added in decided excess, the 

 amount of excess being unimportant, as the solution can not be 

 more than saturated with calcium hydroxid. The solution is boiled 

 and placed on the steam bath for half an hour to allow the magnesium 

 hydroxid to precipitate. This precipitate should not contain much 

 lithium in ordinary cases. With waters very liigh in alkalis it is 

 sometimes advisable to dissolve the magnesium hydroxid in hj^dro- 

 chloric acid and reprecipitate with milk of lime in order to be sure 

 that all the sodium and potassium are in solution. 



To the filtrate, having a volume of about 200 cc, ammonium 

 hydroxid and carbonate are added to precipitate the calcium and 

 barium. The flask is placed on the steam bath until the flocculent 

 precipitate becomes crystalline; then the liquid is filtered off into 

 a platinum dish and the precipitate w^ashed only once or twice and 

 dissolved into the flask with hydrochloric acid, using little more than 

 enough to dissolve tlie whole precipitate. The barium and calcium 

 are precipitated as before and the filtrate and washings added to 

 the first filtrate for evaporation. By gentle ignition the ammonium 

 salts are driven off the residue from evaporation of the fihrates. 

 The alkalis are taken up in a little w^ater and filtered through a 

 small filter mto a 50 cc beaker, a little ammonium carbonate, some 

 ammonia, and one or two crystals of oxalic acid added, and the 

 beaker placed on the steam bath for an hour or more. The last 

 traces of calcium and barium left in solution on account of the large 

 volumes used and the possible action of carbon dioxid on the pre- 

 cipitate are thus precipitated out and filtered off. If more than a 



