308 Mr. Schweitzer on an Analysis of the 



its weight did not aher; it gave 84<"301 grs. The earthy car- 

 bonates were now dissolved very carefully in nitric acid, 

 great care being taken that no loss was sustained during the 

 effervescence, which is best obviated by covering the vessel 

 with a convex glass, having in its centre a very small aper- 

 ture for admitting the dilute acid to drop gradually into the 

 vessel. The solution thus obtained was evaporated in a 

 water-bath to dryness, (no action upon the glass that covered 

 the platina vessel was perceptible, manifesting the absence of 

 fluates,) moistened with nitric acid, and left in that state for 

 twelve hours, when it was re-dissolved. The substance which 

 did not dissolve was lixiviated and boiled with hydrochloric 

 acid, which left a perfectly white silica [a), weighing 1*509 

 gr*. The acid fluid from a was then carefully neutralized 

 with ammonia, and, under the well-known precautions, pre- 

 cipitated by succinate of ammonia ; the precipitate was first 

 lixiviated by very dilute ammonia, and afterwards with pure 

 water, then carefully ignited with free exposure to the air, 

 when 0*105 gr. of oxide of iron [b) was obtained. The fluid 

 derived from b was precipitated by hydrosulphate of ammonia ; 

 the precipitate, dissolved in hydrochloric acid and boiled with 

 an adequate quantity of carbonate of soda, gave a precipitate, 

 which, after lixiviation with boiling water and ignition, weighed 

 0*059 gr., and was red oxide of manganese (c). The acid 

 fluid obtained from D was mixed in a closed vessel with am- 

 monia, and after the precipitate had subsided, the supernatant 

 fluid was quickly filtered, protected from the atmospheric air, 

 and the precipitate (E), lixiviated and ignited, weighed ir819 

 grs. The menstruum from which this precipitate was ob- 

 tained was only slightly acid, so that but a small amount of 

 an ammonia-salt could be formed, and hence a larger amount 

 of magnesia was precipitated, as will be hereafter seen. The 

 precipitate (E) was dissolved in hydrochloric acid ; it left 

 0*103 gr. silica {d). The filtered fluid was mixed with bicar- 

 bonate of potassa, when a precipitate (F) was formed. This 

 precipitate was dissolved in hydrochloric acid, and the solu- 

 tion boiled with caustic soda; the precipitate (G) thus formed, 

 after lixiviation, was redissolved in hydrochloric acid and pre- 

 cipitated by ammonia, when 0*244 gr. oxide of iron {e) was 

 obtained, which dissolved in hydrochloric acid without re- 

 sidue and gave no indications of manganese. The menstruum 

 from which this oxide of iron was precipitated gave with hy- 

 drosulphate of ammonia a few minute flakes of sulphuret of 



* Very often the silica cannot be easily freed from traces of iron : in such 

 cases it is requisite to fuse it with three or four times of its weight of car- 

 bonate of soda, and to dissolve it afterwards in hydrochloric acid. 



