Gooch and Eddy — Separation of Magnesium. 445 



Experimental. 



Schaffgotsch's reagent was prepared by dissolving 230 grm. 

 of ammonium carbonate in 180 cm3 of ammonium hydroxide, 

 diluting the solution to one liter and liltering. 



A solution of magnesium chloride was prepared by the 

 following process. Presumably pure magnesium oxide, 20 

 grm., was treated with nitric acid not quite sufficient in amount 

 to act upon all the oxide. The solution thus obtained was 

 diluted somewhat, and boiled and filtered to remove traces of 

 the elements of the iron group. From this nitrate the greater 

 part of the magnesium was precipitated by concentrated 

 ammonium carbonate, thrown upon a filter, washed, and nearly 

 dissolved in nitric acid. The solution, containing some undis- 

 solved magnesium carbonate, was first boiled and filtered to 

 remove traces of the elements of the calcium group, and then 

 treated again with ammonium carbonate to precipitate mag- 

 nesium carbonate. The insoluble magnesium carbonate was 

 washed and dissolved in the least possible amount of hydro- 

 chloric acid ; and the solution was suitably diluted, and 

 standardized by heating definite portions in a large platinum 

 crucible with sulphuric acid, evaporating, removing the excess 

 of sulphuric acid by careful heating over a radiator and bring- 

 ing to constant weight the residue of magnesium sulphate. 

 With the Schaffgotsch reagent and the standard solution of 

 pure magnesium chloride the following experiments were 

 made. 



In Table I are given results of experiments made to test 

 the solubility of the ammonium magnesium carbonate in an 

 excess of the precipitant of full strength or half strength. 

 Definite portions of the standard solution of magnesium chlor- 

 ide were evaporated in a beaker of 250 =m3 capacity nearly to 

 dryness, and the precipitant was added in the amount indicated. 

 The mixture, stirred vigorously until the flocky precipitate 

 had disappeared and the crystalline double carbonate had begun 

 to form, was allowed to stand at least twenty-four hours. The 

 precipitate, filtered off on asbestos in a perforated crucible, 

 the filtrate being used to effect the transfer without addition 

 of other liquid, was dried, ignited, and weighed as magnesium 

 oxide. The filtrate was treated with microcosinic salt, and, 

 after standing over night, was again filtered on asbestos in a 

 weighed perforated crucible, and the precipitate was washed 

 with water faintly ammoniacal, dried, ignited and weighed. 

 The increase in weight of the crucible was taken as magnesium 

 pyrophosphate from which the magnesium oxide was calcu- 

 lated. In the experiments of A, Schaffgotsch's reagent of full 

 strength was used as the precipitant. In B the same reagent 

 was used of half strength. 



Am. Jour. Sci.— Fourth Series, Vol. XXV, No. 149.— May, 1908. 

 30 



