50. MAGNESIUM. 63 



walls of the tube rubbed here and there with the rod, the 

 precipitate is deposited along these lines, producing the 

 appearance of white streaks on the glass. 



Even in concentrated solutions containing magnesium 

 and ammonic chloride and sodic phosphate, the whole of 

 the ammonio-magnesic phosphate is not deposited until 

 after long standing ; hence, if the first precipitate pro- 

 duced on adding the reagent is filtered out, and the clear 

 filtrate stirred and set aside, a fresh precipitation will take 

 place, and partly on the walls of the tube in the manner 

 described above. 



Quantitative estimation. — Magnesium is usually deter, 

 mined as pyrophosphate, Mg^P^O^. 



a. To the solution of the magnesic salt add a consider- 

 able quantity of ammonic chloride, and then ammonia in 

 slight excess ; if this ammonia causes the formation of a 

 precipitate, add enough more amnionic chloride to dis- 

 solve it ; then add hydric disodic phosphate, as long as a 

 precipitate is formed, stir the mixture well, with care not 

 to touch the sides of the beaker with the rod, cover the 

 beaker carefully, and let it stand with its contents 12 

 hours without ajiplying heat ; decant the clear liquid 

 through the filter, rinse the contents of the beaker into 

 the filter with portions of the first filtrate, and wash the 

 contents of the filter with a diluted ammonia water 

 containing one part of ammonia water of 0.96 Sp. Gr. 

 and three of water, until the last five drops of the wash- 

 ings give no opalescence with very dilute nitric acid con- 

 taining argentic nitrate. 



Ignite the precipitate and filter separately. Rose rec- 

 ommends to ignite the precipitate for a short time in a 

 porcelain crucible over the blast-lamp ; in this way it is 

 obtained quite white. 



Add 0.002 grm. to the residue of magnesic phosphate 

 for every 110 c.c. of the filtrate from the precipitate (but 



