OF ARTS AND SCIENCES. 



193 



concerns the parting of the alkalies from magnesium alone, — a 

 matter wliich is by no means perfectly simple, — and experiments (38) 

 to (41) touch upon this topic. 



The chlorides of sodium and potassium were weighed, as before, in 

 solution; (he magnesium chloride was obtained by dissolving in hydro- 

 chloric acid the oxide specially prepared and weighed as such. The 

 process of treatment was identical with that just clescribed for the 

 separation of the chlorides of potassium and sodium from lithium 

 chloride. 



The residues of experiments (38) and (39), in which the separation 

 was made by a single precipitation, carried traces of magnesia ; those 

 of (40) and (41), in which two precipitations were introduced, were 

 found to contain in the one case no magnesia, and in the other an 

 unweighable trace. These results point out a method by which the 

 chlorides of sodium and potassium may be obtained free from magnesia, 

 while the small amounts of the former which pass into solution with 

 the magnesium chloride are capable of accurate estimation ; and there 

 seems to be no reason why the separation of these alkaline chlorides 

 from magnesium chloride and lithium chloride occurring together 

 should not be effected in one operation, and the parting of the latter 

 salts brought about by the familiar method of precipitating the mag- 

 nesia in the cold as ammonium-magnesium phosphate. 



Experiments (42) to (48), upon the separation of sodium and potas- 

 sium from calcium by the action of amyl alcohol on the chlorides, 

 yielded the figures of the following table. The mode of treatment 

 was identical with that of the experiments with magnesia, just de- 



VOL. XXII. (n. 8. XIV.) 13 



