DATOLITE. 145 



which latter is wanting only in the olivine of meteoric 

 iron. These are precipitated by sulphide of ammo- 

 nium, an excess of which is, as far as possible, to be 

 avoided. The precipitate is not to be filtered oft* until 

 it has separated so as to leave the solution perfectly 

 clear ; it may then be washed with very dilute sulphide 

 of ammonium. If both metals are present only in 

 small quantities, the sulphide of manganese may then 

 be separated from the sulphide of nickel by treating it, 

 upon the filter, with very dilute hydrochloric acid, in 

 which the sulphide of nickel is, practically, insoluble. 

 The small quantity of sulphide of nickel is then ignited 

 in the air, and weighed as protoxide. The manganese is 

 precipitated from the solution by carbonate of soda, at 

 a boiling heat. 



The liquid filtered from the precipitate produced by 

 sulphide of ammonium, is mixed with ammonia, and 

 the magnesia precipitated by phosphate of soda (No. 6). 



In the analysis of olivine, the iron, when converted 

 into sesquioxide, may also be separated from the other 

 bases by succinate of ammonia (No. 21). 



78. DATOLITE. 

 (3 CaO, 3HO, B0 3 ) Si0 2 . 



For the determination of water, a weighed quantity 

 of the mineral is heated to bright redness. 



If the unignited mineral, in a finely-powdered state, 

 be digested with moderately strong hydrochloric acid, 

 it becomes a gelatinous mass. If the mixture be heated 

 to boiling, and filtered while hot, boracic acid separates 

 from the solution in crystals. 



The finely-powdered mineral is decomposed by di- 

 gestion with hydrochloric acid, and the mass evaporated 

 to dryness, when a great part of the bqracic acid is 



