264 Kreider and Breckenridge — Sejparation and 



even in the presence of considerable perchloric acid and at the 

 boiling point of the mixture. 



The strongest aqueous solution of hydrochloric acid, how- 

 ever, is inapplicable, as was proved by several experiments in 

 which as much as 0-010 grms. of sodium perchlorate, dissolved 

 in only 5 cm^ of 97 per cent alcohol, failed to be revealed by the 

 addition of the strongest aqueous solution of the acid, added 

 in quantities varying from a single drop to 10 cm^ — not the 

 slightest turbidity being produced. By substituting for the 

 aqueous solution of the acid, a saturated solution of hydro- 

 chloric acid in 97 per cent alcohol, of which 5 cm^ were used for 

 the precipitation, quantities of from 0*002 to 0'003 grms. of 

 sodium perchlorate dissolved in 5 cm^ of 97 per cent alcohol 

 could be detected with certainty ; but this could not be made 

 sufficiently delicate. Concentrated sulphuric acid was also 

 applied as the precipitant and was found reliable for quantities 

 of about 0*003 grms. of sodium perchlorate dissolved in lOcm^ of 

 97 per cent alcohol, when a single drop of the acid was added, 

 but an excess of the acid caused the precipitate to redissolve. 

 Gaseous hydrochloric acid proved most effectual. The dehy- 

 drating effect of the acid upon the alcohol greatly increases the 

 insolubility of sodium and secures a remarkable delicacy, as is 

 evident from the results recorded in Table I. 





Table I. 





3104 taken. 



NasO equivalent. 



97^ alcohol. 



Indication. 



grms. 



grms. 



cm3. 





0-0100 



0-00250 



10 



very strong 



0-0050 



0-00125 



(( 



strong 



0-0040 



0-00100 



a 



(( 



0-0030 



0-00075 



a 



u 



0-0030 



0-00075 



ie 



good 



0-0020 



0-00050 



a 



ii. 



0-0020 



0-00050 



it 



(I 



0-0010 



0-00025 



(( 



(( 



0-0005 



0-00012 



(C 



•trace 



0-0003 



0-00006 



(( 



C( 



0-0001 



0-00003 



cc 



none 



0-0000 



0-00000 



ii 



li 



0-0010 



0-00025 



40 



distinct 



The sodium employed in these determinations was in the 

 form of perchlorate, prepared by evaporating sodium chloride, 

 free of potassium and ammonium, with perchloric acid until 

 tests for hydrochloric acid proved a complete conversion to the 

 perchlorate, when the excess of acid was volatilized by heating 

 over a drying cone. Two grams of this purified salt were dis- 

 solved in 200 cm'' of 97 per cent alcohol and served for our 

 standard solution for those tests in which quantities greater 

 than O'OOi grms. of sodium perchlorate were used. When it 



