252 ^- CHIKASHIGE. 



ignition directed to be used in § 153, and would entnil seriou« loss of 

 alkali chlorides by spirting and volatilisation. But to place the in- 

 accuracy of the statement in the footnote beyond doubt, 1 have made 

 a few simple trials of the method. 



Magnesia, 0.5 gram, was dissohed in a little hydrochloric acid ; 

 to the solution were added two dro))s of dilute sulphuric acid (1 to 10 

 Witter by volume), a solution of about two grams of ammonium 

 chloride, and ammonia in sniall excess ; and the whole w;is then 

 evaporated to dryness ;ind all ammonium salts expelled at a barely 

 red heat. Again, a solution of two grams of ammonium chloride was 

 added, and the evaj)(~)ralion and ignition re])eated. The I'csidue was 

 dissolved in dilute hydrochloric acid and mixed with barium chloride, 

 which gave a precipitate. On comparing this with that tlu'own down 

 by barium chloride from two drops of the same dilute sulphuric acid 

 in about the same volume of water, the ignited salts wei'e tbund to 

 have lost but very little, if any, of their sulphuric acid. The ex].)eri- 

 ment was repeated three times with fresh magnesium diloride. and the 

 same results obtained. 



Similar experiments were made with sodiiun chloride, and no 

 Ijettei" removal of the sulphuric acid effected than when magnesium 

 chloride was taken. 



It is thus quite evident that other steps must be taken to i-emove 

 sui])huric acid before it is allowable to calculate the weighed alkali 

 s:dts as chlorides or to resort to ways of separating magnesi;i from tlie 

 alkalis which require the absence of sulphates. 



