dissolve, while the silica remains behind, sometimes with a portion of the gypsum. The earth is precipitated after filtration by oxalate of ammonia; the oxalate of lime is then collected on a filter, and, after being washed, treated with soda on charcoal, or with microcosmic salt upon the platinum wire. Should the mineral contain traces of alumina, this earth can be precipitated by caustic ammonia, before the lime is separated by the oxalate, and afterwards tested with a solution of cobalt, provided there is no sesquioxide of iron present to prevent the pale-blue reaction, upon charcoal. To obtain the silica quite free from sulphate of lime, the silicious residue on the filter must be treated with a very weak hydrochloric acid solution, and filtered. This second filtrate must not be added to the first, because if it is not properly diluted, a precipitate of sulphate of lime may occur on the addition of ammonia. Silicates,-including furnaced ores, scoriæ, et cetera, with several bases, are sometimes not at all decomposed by bisulphate of potassa. This, however, is effected, if the process be conducted similar to that given under Baryta for furnaced ores and slags. For the decomposition of silicates, a mixture of carbonate of soda and potassa will be found the best.

If the solution obtained after melting the silicate with soda and borax, et cetera,-given under Baryta, contains protoxide of iron, it must be converted into the sesquioxide with a few drops of nitric acid; then add chloride of ammonium, and precipitate the alumina and sesquioxide of iron by a slight excess of ammonia. The method of separating these two constituents will be given under Alumina, when the silicious compounds are noticed. The lime which is in solution, combined probably with portions of magnesia, protoxide of manganese, and oxide of cobalt, is examined as follows. Sulphide of Ammonium is added to the solution, until all the manganese and cobalt are precipitated. The liquid is then filtered, and the residue washed with water, containing some of the precipitant. After affusing for some time, the filter containing the sulphides is burnt, the residue pulverized, and examined with microcosmic salt on platinum wire for cobalt, and with carbonate of soda, to which a small quantity of saltpetre

has been added, on the platinum foil, in the oxidating flame, for manganese.

To the filtrate is added a few drops of hydrochloric acid, and the whole is treated over the lamp, until no sulphide of hydrogen escapes. The menstruum is then filtered from the sulphur, the solution saturated with ammonia, the lime precipitated as an oxalate, and separated by filtration. The magnesia is precipitated by microcosmic salt as ammonia-phosphate, but before adding the microcosmic salt it must be ascertained whether all the lime has been abstracted. When no manganese or cobalt is present, there will only be two filtrations, to separate the lime from the magnesia. If the silicate contains traces of protoxide of manganese only, the precipitated magnesia must be examined. for them with soda and saltpetre, as will be given under Manganese.

The protoxide of manganese is converted into a higher oxide by the nitre which combines with the alkali, forming a clear limpid glass, which is driven from one part of the platinum foil to another, while the magnesia remains in one spot upon the support, and is readily seen.

The examination of silicates, which, besides lime, may contain yttria and protoxide of cerium, is given under Yttria.-When solutions of lime are so very dilute, that sulphuric acid causes no precipitate, alcohol should be added, when a deposit of sulphate immediately appears.

DOBEREINER has given an excellent method for the separation of lime and magnesia, which is nearly as follows :-Evaporate the hydrochloric acid solution of the two earths to dryness, in a large porcelain crucible, and then ignite to expel the excess of acidAdd chlorate of potassa in small crystals, until the eliminated chlorine gas is perceived. When the dry mass is affused with water, chlorides of calcium and potassium dissolve, and a residue of pure hydrated magnesia remains. The liquid can now be filtered, to separate the magnesia, and oxalate of ammonia added, to precipitate the lime.

§ 8. MAGNESIA-Mg O-Presence in the Mineral Kingdom.

b. In Aluminate, as also in Aluminate of MagnesiaSpinel-[Mg O, Ale 03]; and as an Aluminate of the Protoxide of Iron and Magnesia, in Zeylanite-Pleonaste — [Mg 0} Al2 03];

Fe

c. In Brucite-Condrodite, Maclureite [Mg Fl + 3 (3 Mg O, Si O2)];

Mg
o

d. With Carbonic Acid, in Magnesite e C O]; and

Mn O

with Carbonic acid and Carbonate of Lime, in many Carboniferous Spars, as Dimeric in Dolomite, Isomeric in Tarandite, Americ in Brown Spar, et cetera.

e. With Sulphuric Acid and Water, in Sulphate of Magnesia [Mg O, S 03 + HO+6 aq]; and with Sulphuric Acid, Sulphate of Potassa, and Sulphate of Lime, in Polyhallite.

f. With Boracic Acid, in Boracite [Mg 0, 2 B 03, + 2 (Mg 0, B 03) ];

g. In Silicates:

:

1 In which Magnesia forms the principal base; exempli gratia, in Speckstein [6 Mg 0, 5 Si O2 + 4 aq]; Meerschaum [Mg 0, Si Oaq]; Picrosmine-Parallolite-[3 Mg 0, 2 Si 02]; Serpentine-hydrate of magnesia with subsilicate of magnesia[3 Mg 0, 6 H 0 + 2 (3 Mg 0 + 2 Si O2)]; Olivine [3 Ca O, Si O2 + [30 Si O,] et cetera. Mg Fe

2 In Silicates of Lime and Magnesia, with Silicates of Alumina, without water; as Anorthite [3 Mg O, Si 02 + 2 (3 Ca O, Si O2) + 8 Al2 03, Si O2] or [3 Mg O, Si 02 + 2 Al2 03, Si O2 + 2 (3 Ca O, Si O2 + 3 Alo 03, Si 0o) ]—BERZELIUS, and according to ABICH [3 R 0, 2 Si O2 + 2 (R2 O3, 2 S 02) + 6 (3 Ca O, Si O2) + 3 (Al2 03, Si 02) ];

3 In Silicates of Potassa or Soda, with Silicates of Magnesia, Alumina, and Protoxide of Iron, with and without water, in which the alkalies are often replaced by Magnesia and Protoxide of

Iron; exempli gratia, in Prismatic Magnesian Mica, to which Potstone, Green Earth, Chlorite, and Tale belong;

4 In Silicates of Lime and Magnesia, in which the alkaline earths are more or less replaced by Protoxides of Iron and Manganese, and the Silicic acid sometimes by Alumina; to these belong Amphibole and the Magnesia-Pyroxenes;

5 In Silicates of Lime and Magnesia with Alumina, in which the alkaline earths are sometimes replaced by Protoxide of Iron, and the Alumina by the Sesquioxide; exempli gratia, in Magnesia-Idiocrase; in the Black Garnet of Arendal; in Soapstone [3 Mg 0, 2 Si 02 + Al2 03, 2 Si O2 +6 aq]; in Blue and Red Dichroite-Iolite, Cordierite-3 Fe O Si O2 + Alo 03, 2 Si 0o]

A12
Fe

[s Mg

3 Mn

and [3 Mg 0, 2 Si O+ {30} Si Oo];

6 In Silicates with Phosphates; exempli gratia, in Sordawalite, seemingly a mixture of [2 Mg O, P 05] with [3 Mg 0, 2 Si 02 + Al2 03, 2 Si 0o + 2 (3 Fe 0, 2 Si O2 + Al2 03, 2 Si 0°) ];

7 In Silicates with metallic Fluorides; as in Chondrodite [3 (Mg Fl) + 4 (3 Mg O, Si O2) ]-Magnesia is met with in Boltonite, Batrachite, Nemalite, Villarsite, Steatite, Aphrodite, Gymnite, Antigonite, Pennine, Monradite, Piotine, Leuchtenbergerite, Spadaite, Schillerspar, Dermatine, Praseolite, Esmarkite.

If some of the above mentioned minerals occur as well in the gangues as in the ores, the ores prepared in the dry way, and also the recrements and scoriæ obtained therefrom, generally contain a small quantity of Magnesia, which can readily be determined.

Examination for Magnesia.

a. Aluminate, Spinel, and Zeylanite are readily recognized before the Blowpipe, as they give characteristic glasses, owing to the metallic oxides they contain, but magnesia per se, is not detectable with certainty. To arrive at this earth, the finely powdered mineral must be heated with two parts of soda and three parts of borax, upon charcoal, the fused mass dissolved in

hydrochloric acid, the solution evaporated to dryness; again dissolved in water, and the liquid filtered, to separate the silica. Nitric acid must then be added to the filtrate, to convert the protoxide of iron into sesquioxide, and two spoonfuls of chloride of ammonium introduced; the iron, alumina, and any oxide of chromium must then be precipitated by ammonia, and separated from the supernatant liquor. The filtrate must now be boiled, the magnesia precipitated by microcosmic salt as ammoniaphosphate. This precipitate must be collected on a filter, edulcorated with water, and then treated on charcoal. It fuses to a white crystalline translucent pearl, and when treated with a few drops of cobalt solution, assumes a violet color, which appears reddish in the flame of the lamp. When the Spinel dissolves in borax with effervescence, it is a sign that there is some carbonate of lime between the interstices. Should this occur, the lime is to be precipitated by oxalate of ammonia, before the addition of the microcosmic salt. The aluminous precipitate can be examined with borax or microcosmic salt upon the platinum wire, for iron and chromium, or else treated with potassa, which dissolves the alumina, and leaves the sesquioxide of iron as a reddish brown, and the oxide of chromium as a dirty green deposit. Another method of analysing a mixture containing alumina, sesquioxide of iron, and oxide of chromium, is as follows:-To a cold hydrochloric acid solution of the compound, add solution of potassa, which will precipitate the sesquioxide of iron as a reddish brown powder, and filter. Boil the solution, which should contain an excess of alkali; the chromium will precipitate of a green color, and must be separated by filtration. To the filtrate add chloride of ammonium, which will throw down the alumina. This is an excellent method, especially for qualitative analysis.

If the mineral contains iron and chromium, and the residue only appears ferruginous, the mass may be smelted upon the platinum wire with nitrate of potassa, and the fused mass treated with water. The chromate of potassa formed, and the undecomposed nitrate, dissolve, while the iron remains behind. The chromic acid can be detected in the solution, according to the method given under Chromium.