648 SODIUM. 



of acid to 1 grain of soda. It is transferred to a stock bottle. 

 The remainder of the original dilute acid is diluted with water 

 to an equal extent, in the same instrument, and added to the 

 bottle. The density of this acid is 1.0.995 or 1.0998,, which is 

 sensibly the same as 1.1. By a curious coincidence, strong oil 

 of vitriol diluted with 11 times its weight of water,, gives this 

 test acid exactly ; but as oil of vitriol varies a little in strength, 

 it is better to form the test acid in the manner described, than 

 to trust to that mixture. Twenty-one measures of the test acid 

 should neutralise 100 grains of cr. carbonate of soda; and 58.5 

 measures, 100 grains of pure anhydrous carbonate of soda. 



(3) In applying the test acid, it is poured from the alkali- 

 meter^ as before, upon 100 grains of the soda- salt to be tested, 

 dissolved in two or three ounces of hot water. The salt contains 

 so many grains of soda, as it requires measures of acid to neu- 

 tralise it ; and therefore so much alkali per cent. If the soda- 

 salt is mixed with insoluble matter, its solution must be filtered 

 before the test acid is added to it. In examining a soda-salt 

 which blackens salts of lead, and contains both carbonate of 

 soda and sulphuret of sodium, 100 grains are tested as above, 

 and the whole alkali in both salts thus determined. A neutral 

 solution of chloride of calcium is also added in excess to the so- 

 lution of a second hundred grains, by which the carbonate of 

 soda is converted into chloride of sodium, while carbonate of 

 lime precipitates. The filtered liquid is still alkaline, and con- 

 tains all the sulphuret of sodium ; the quantity of soda corres- 

 ponding with which is ascertained by means of the test acid. 

 The difference between the quantities of alkali observed in the 

 two experiments is the proportion of soda present, as carbonate. 



Borax also may be analyzed by the same test acid, and will 

 be found when pure to contain 16.37 per cent of soda. The 

 carbonates of potash may also be examined by the same means, 

 but the per centage of alkali must then be estimated higher than 

 the measures of acid neutralised, in the proportion of the equiva- 

 lent of soda to that of potash, which are to each other very 

 nearly as 39 to 59. The test paper employed in alkalimetry 

 must be delicate. It should be prepared on purpose, by dipping 

 good letter-paper several times in a filtered infusion of litmus, 

 and drying it after each immersion, till the paper is of a deep 

 purple colour. A test paper prepared with cudbear in the same 

 way answers still better, but the latter colouring matter is not 



