^ATER OF THE DEAD SEA. ^| 



2. From another similar solution the lime was precipitated 

 by oxalate of ammonia; the magnesia was separated by- 

 heat in an open cmcible, and the common salt was obtained, 

 as before, by evaporation and exposure to a low red heat. 

 The result was satisfactory both as to the lime and magne- Themuriateof 

 si^; but as the separation of the latter could only be com- gj ^ subhma- 

 pleted by long continued heat, in an open vessel, I found tion. 

 the muriate of soda materially redttced by sublimation, and 

 was therefore obliged to abandon this mode of proceeding. 



3'. From a third artificial solution, the lime was preeipi- A 3d trial de- 

 lated by oxalate of ammonia, the magnesia by carbonate of *^'^''^^' 

 ammonia recently prepared, and the sea salt was obtained 

 as usual by evaporation and desiccation in a low red heat. 

 The object of this mode of operating was to supersede the 

 necessity of applying a red heat in the first instance. But 

 I was again disappointed ; for the magnesia was but imper- 

 fectly precipitated; and in order to separate the last portions 

 of this earth, it was necessary to calcine the last residue 

 containing the muriate of soda, which gave rise to the same 

 objections as in the former experiments. 



4. The last and most successful method consisted in di- Most success 

 viding the artificial solution into two portions. From one of ^"^ method'. 

 these the muriatic acid was precipitated by nitrate of silver, 

 and its quantity ascertained. From the other the lime was 

 separated by oxalate of ammonia, and the magnesia by 

 caustic potash* ; and the respective portions of acids belong- 

 ing to each of these earths being calculated, the quantity of 

 muriate of soda was inferred from the remaining quantity of 

 acid. 



This method afforded remarkably accurate results. The Only objectmu 

 only objection to it seems to be, that the muriate of soda ^^ ^* 

 being only estimated, and not actually obtained, if any er- 

 rour be made either in the estimation of the acid or in the 

 separation of the lime and magnesia, these errours must also 

 ultimately affect the Computation of the muriate of soda, 

 without allowing any immediate means of detecting them» 



• Or, by carbonate of ammonia. In this ease the precipitation of 

 Hiagnesia is not so perfect j but the precipitate fails down more quickly, 

 and' the separation of any remaining portion of thi? earth may be ulti. 

 raately completed by heat. 



This 



