20 THE DETERMINATION OF LITHIUM. 



moistened with a little sulphuric acid and water and evaporated 

 again. The organic matter is destroyed by repeated very gentle 

 ignitions, the residue being washed down with water and a drop or 

 two of sulphuric acid after each ignition, and the water evaporated 

 off. When the organic matter is removed, the dish and contents 

 are gently ignited to constant weight, which is usually accom- 

 plished about as soon as the organic matter is removed, as the last 

 trace of it is hardly weighable, although it may be visible. 



The sulphates of sodium, potassium, and lithium are dissolved in 

 a small quantity of water and filtered through a small filter and the 

 dish and filter carefully washed. The paper with contents is burned 

 in the dish and the weight of dish and ash subtracted from the 

 previous weight to obtain the weight of sodium, potassium, and 

 lithium sulphates formed from the clilorids dissolved by the amyl 

 alcohol. To the filtrate are added a few crystals of ammonium 

 phosphate, and when they are dissolved a few drops of ammonia 

 are added, the solution stirred, and allowed to stand overnight. 

 If a small precipitate forms, it is collected, ignited, and weighed 

 as magnesium pyrophosphate. The equivalent weight of magnesium 

 sulphate is calculated and subtracted from the weight of the com- 

 bined sulphates obtained from the amyl alcohol extraction. The 

 figures given by Gooch are used to calculate the correction for the 

 solubility of sodimn and potassium chlorids in amyl alcohol. Gooch 

 found the solubihty of sodium clilorid in the boiling amyl alcohol to 

 be 0.0041 gram in 100 cc, and for potassium chlorid 0.0051 gram in 

 100 cc, Wlien calculated to sulphates, these become 0.0050 gram of 

 sodium sulphate and 0.0059 gram of potassium sulphate, or 0.0109 

 gram of both. In his experiments Gooch found that the anhydrous 

 amyl alcohol used for washing the insoluble residue dissolved no 

 appreciable amount of sodium or potassium chlorid, so that no cor- 

 rection was necessary for the volume of amyl alcohol used in the 

 washing. For every 10 cc of alcohol in the filtrate (exclusive of the 

 amount used for wasliing) from which the chlorids are precipitated 

 there is subtracted from the weight of sulphates of lithium, sodium, 

 and potassium 0.00109 gram. In many cases the corrections amount 

 to the total weight obtained within the Hmit of the weighing error 

 of the work. In such cases the amount of lithiima may be deter- 

 mined by the use of the spectroscope. 



It is evident that for the determination of very small amounts of 

 lithium no method depending on the precipitation of a litliium com- 

 pound is entirely satisfactory, because the solubility of the least 

 soluble compound ma}^ be so great as to give no precipitate at all. 

 It seems equally unreasonable to expect complete extraction of very 

 small amounts of litliium from large amounts of other salts either 

 by ether-alcohol or pyridin. However, in the extraction by am}'! 



