apply for this purpose the above-described mixture of both, because this mixture requires a far lower degree of heat for fusion than either of its two components, and thus enables us to conduct the operation over a Berzelius lamp, or over a simple gas lamp. The fusion with alkaline carbonates is invariably effected in a platinum crucible, provided no reducible metallic oxides be present.

$ 76.

2. HYDRATE OF BARYTA (Ba O, H O).

Preparation. The crystals of baryta prepared in the manner directed § 32, are heated gently in a silver or platinum dish, until the water of crystallization is completely expelled. The residuary white mass is pulverized, and kept for use in a well-closed bottle.

Uses.-Hydrate of baryta fuses at a gentle red heat without losing its water. Upon fusing silicates together with about 4 parts of hydrate of baryta, a basic silicate of baryta is formed, and the oxides are liberated. If the fused mass is treated with hydrochloric acid, the solution evaporated to dryness, and the residue digested with hydrochloric acid, the silicic acid is left behind, and the oxides are obtained in solution in the form of chlorides. We use hydrate of baryta as a flux when we wish to test silicates for alkalies. This reagent is preferable as a flux to the carbonate or nitrate of baryta, since it does not require a very high temperature for its fusion, as is the case with the carbonate, nor does it cause any spirting in the fusing mass, arising from disengagement of gas, as is the case with the nitrate. The operation is conducted in silver or platinum crucibles.

§ 77.

3. FLUORIDE OF CALCIUM (Ca Fl).

Take fluor-spar as pure as can be procured, and more particularly free from alkalies, reduce to fine powder, and keep this for use.

Uses.-Fluoride of calcium applied in conjunction with sulphuric acid, serves to effect the decomposition of silicates insoluble in acids, and more especially to detect the alkalies which they contain. Compare Section III. Silicic acid, § 150.

$ 78.

4. NITRATE OF SODA (Na O, NO,).

Preparation. Neutralize pure nitric acid with pure carbonate of soda exactly, and evaporate to crystallization. Dry the crystals thoroughly, triturate, and keep the powder for use.

Tests. A solution of nitrate of soda must not be made turbid by solution of nitrate of silver or nitrate of baryta, nor precipitated by carbonate of soda.

Uses.-Nitrate of soda serves as a very powerful oxidizing agent, by yielding oxygen to combustible substances when heated with them. We use this reagent principally to convert several metallic sulphides, and more particularly the sulphides of tin, antimony, and arsenic, into oxides. and acids; and also to effect the rapid and complete combustion of organic substances; for the latter purpose, however, nitrate of ammonia

is in many cases preferable; this latter reagent is prepared by saturating nitric acid with carbonate of ammonia.

II. BLOWPIPE REAGENTS.

$79.

1. CARBONATE OF SODA (Na O, C 02).

Preparation.-See § 44.

Uses.-Carbonate of soda serves, in the first place, to promote the reduction of oxidized substances in the inner flame of the blowpipe. In fusing it brings the oxides into the most intimate contact with the charcoal support, and enables the flame to embrace every part of the substance under examination. It co-operates in this process also chemically by the transposition of its constituents (according to R. Wagner, in consequence of the formation of cyanide of sodium). If the quantity operated upon was very minute, the reduced metal is often found in the pores of the charcoal. In such cases, the parts surrounding the little cavity which contained the sample are dug out with a knife, and triturated in a small mortar; the charcoal is then washed off from the metallic particles, which now become visible either in the form of powder or as small flat spangles, according to the nature of the particular metal or metals present.

Carbonate of soda serves, in the second place, as a solvent. Platinum wire is the most convenient support for testing the solubility of substances in fusing carbonate of soda. A few only of the bases dissolve in fusing carbonate of soda, but acids dissolve in it with facility. Carbonate of soda is, moreover, applied as a decomposing agent and flux, and more particularly to effect the decomposition of the insoluble sulphates, with which it exchanges acids, the newly formed sulphate of soda being reduced at the same time to sulphide of sodium; and to effect the decomposition of sulphide of arsenic, with which it forms a double sulphide of arsenic and sodium, and arsenite or arsenate of soda, thus converting it to a state which permits its subsequent reduction by hydrogen. Finally, carbonate of soda is the most sensitive reagent in the dry way for the detection of manganese, since when fused in the outer flame of the blowpipe together with a substance containing manganese, it produces a green opaque bead, owing to the formation of manganate of soda.

$ 80.

2. CYANIDE OF POTASSIUM (K Cy).

Preparation.-See § 52.

Uses.-Cyanide of potassium is an exceedingly powerful reducing agent in the dry way; indeed it excels in its action almost all other reagents of the same class, and separates the metals not only from most oxygen compounds, but also from sulphur compounds: this reduction is attended in the former case with formation of cyanate of potassa, by the absorption of oxygen, and in the latter case with formation of sulphocyanide of potassium. By means of this reagent we may effect the reduction of metals from their compounds with the greatest possible facility; thus we may, for instance, produce metallic antimony from antimonious acid or from sulphide of antimony, metallic iron from sesquioxide of iron, &c.

The readiness with which cyanide of potassium enters into fusion facilitates the reduction of the metals greatly; the process may usually be conducted even in a porcelain crucible over a spirit lamp. Cyanide of potassium is a most valuable and important agent to effect the reduction of arsenites and arsenates, and more particularly of tersulphide of arsenic (see § 131). Cyanide of potassium is equally important as a blowpipe reagent. Its action is exceedingly energetic; substances like binoxide of tin, bisulphide of tin, &c., the reduction of which by means of carbonate of soda requires a tolerably strong flame, are reduced by cyanide of potassium with the greatest facility. In blowpipe experiments we invariably use a mixture of equal parts of carbonate of soda and cyanide of potassium; the admixture of carbonate of soda is intended here to check in some measure the excessive fusibility of the cyanide of potassium. This mixture of cyanide of potassium with carbonate of soda, besides being a far more powerful reducing agent than the simple carbonate of soda, has, moreover, this great advantage over the latter, that it is absorbed by the pores of the charcoal with extreme facility, and thus permits the production of the metallic globules in a state of the greatest purity.

$81.

3. BIBORATE OF SODA (Borax) (Na O, 2 B 0,, crystallized + 10 aq.).

The purity of commercial borax may be tested by adding to its solution carbonate of soda, or after previous addition of nitric acid, solution of nitrate of baryta or of nitrate of silver. The borax may be considered pure if these reagents fail to produce any alteration in the solution; but if either of them causes the formation of a precipitate, or renders the fluid turbid, recrystallization is necessary. The pure crystallized borax is exposed to a gentle heat, in a platinum crucible, until it ceases to swell; when cool, it is pulverized and kept for use.

Uses.-Boracic acid manifests a great affinity for oxides when brought into contact with them in a state of fusion. This affinity enables it, in the first place, to combine directly with oxides; secondly, to expel weaker acids from their salts; and, thirdly, to predispose metals, sulphides, and haloid compounds to oxidize in the outer flame of the blowpipe, that it may combine with the oxides. Most of the thus produced borates fuse readily, even without the aid of a flux, but far more so in conjunction with borate of soda; the latter salt acts in this operation either as a mere flux, or by the formation of double salts. Now, in the biborate of soda we have both free boracic acid and borate of soda; the union of these two substances renders it one of our most important blowpipe reagents. In the process of fluxing with borax, we usually select platinum wire for a support; the loop of the wire is moistened or heated to redness, then dipped into the powder, and exposed to the outer flame; a colourless bead of fused borax is thus produced. A small portion of the substance under examination is then attached to the bead, by bringing the latter into contact with it, either whilst still hot or having previously moistened it. The bead with the sample of the substance intended for analysis adhering to it, is now exposed to the blowpipe flame, and the phenomena to the manifestation of which this process gives rise are carefully observed and examined. The following points ought to be more particularly watched:-(1) Whether or not the sample under examination dissolves to a transparent bead, and whether or not the bead retains its transparency on cooling; (2) whether the bead exhibits a dis

tinct color, which in many cases at once clearly indicates the individual metal which the analysed compound contains, as is the case, for instance, with cobalt; and (3) whether the bead manifests the same or a different deportment in the outer and in the inner flame. Phenomena of the latter kind arise from the ensuing reduction of higher to lower oxides, or even to the metallic state, and are for some substances particularly

characteristic.

$ 82.

4. PHOSPHATE OF SODA AND AMMONIA (Microcosmic Salt)
(Na O, N H, O, H 0, P 0,, crystallized + 8 aq.).

4

Preparation.-Heat to boiling 6 parts of phosphate of soda and 1 part of pure chloride of ammonium with 2 parts of water, and let the solution cool. Free the crystals produced of the double phosphate of soda and ammonia by recrystallization from the chloride of sodium which adheres to them. Dry the purified crystals, and pulverize them for use. Uses. When phosphate of soda and ammonia is subjected to the action of heat, the ammonia escapes with the water of crystallization, and readily fusible metaphosphate of soda is left behind. The action of microcosmic salt is quite analogous to that of biborate of soda. We prefer it, however, in some cases, to borax as a solvent or flux, the beads which it forms with many substances being more beautifully and distinctly colored than those of borax. Platinum wire is also used for a support in the process of fluxing with microcosmic salt; the loop of the wire must be made small and narrow, otherwise the bead will not adhere to it. The operation is conducted as directed in the preceding paragraph.

§ 83.

5. NITRATE OF PROTOXIDE OF COBALT (Co O, NO,, crystallized + 5 aq.).

Preparation.-Fuse in a Hessian crucible 3 parts of bisulphate of potassa, and add to the fused mass, in small portions at a time, 1 part of well roasted cobalt ore (the purest zaffre you can procure) reduced to fine powder. The mass thickens, and acquires a pasty consistence. Heat now more strongly, until it has become more fluid again, and continue to apply heat until the excess of sulphuric acid is completely expelled, and the mass accordingly no longer emits white fumes. Remove the fused mass now from the crucible with an iron spoon or spatula, let it cool, and reduce it to powder; boil this with water until the undissolved portion presents a soft mass; then filter the rose-red solution, which is free from arsenic and nickel, and mostly also from iron, and remove the copper, &c., from the filtrate by means of hydrosulphuric acid. Filter again, and evaporate the filtrate, with addition of some chlorine water, until it is much concentrated. Mix the concentrated filtrate now with a hot saturated solution of binoxalate of potassa, and let the mixture stand at a gentle heat until the fluid appears colorless. Wash the precipitated oxalate of protoxide of cobalt thoroughly, dry, and heat to redness in a covered platinum or porcelain crucible. This decomposes the oxalate into water and carbonic acid, which escapes, and metallic cobalt, which is left behind. Dissolve a portion of the latter in nitric acid, taking care to avoid a large excess of the solvent; evaporate the solution in the water-bath to dryness, and dissolve 1 part of the residue in 10 parts of water for use.

Tests.-Solution of nitrate of protoxide of cobalt must be free from other metals, and especially also from salts of the alkalies; when precipitated with sulphide of ammonium and filtered, the filtrate must, upon evaporation on platinum, leave no fixed residue.

Uses.-Protoxide of cobalt forms, upon ignition with certain infusible bodies, peculiarly colored compounds, and may accordingly serve for the detection of these bodies (oxide of zinc, alumina, and magnesia; see Section III.).

$ 84.

6. CHLORIDE OF SILVER (Ag Cl).

Preparation. Precipitate solution of nitrate of silver with hydrochloric acid; wash the precipitate, mix it with water to a thick pulp, and keep in a small bottle for use.

Uses.-Chloride of silver has lately been recommended by Gericke as a means of making the colorations more distinct and lasting which certain bodies, upon exposure to the inner blowpipe flame, impart to the outer flame. I can from my own experience confirm the results arrived at by Gericke. The action of the chloride of silver is owing to the circumstance that this compound loses its chlorine only gradually upon exposure to heat, and gives rise accordingly for a certain time to the formation of metallic chlorides, which it is well known impart more distinct colorations to flame than any other class of salts. As platinum wire would speedily be rendered unfit for use by the reduced silver, thin iron wire is employed; every operation requires a new loop.

SECTION III.

ON THE DEPORTMENT OF BODIES WITH REAGENTS.

$ 85.

I STATED in my introductory remarks that the operations and experiments of qualitative analysis have for their object the conversion of the unknown constituents of any given compound into forms of which we know the deportment, relations, and properties, and which will accordingly permit us to draw correct inferences regarding the several constituents of which the analysed compound consists. The greater or less value of such analytical experiments, like that of all other inquiries and investigations, depends upon the greater or less degree of certainty with which they lead to definite results, no matter whether of a positive or negative nature. But as a question does not render us any the wiser if we do not know the language in which the answer is returned, so, in like manner, will analytical investigations prove unavailing if we do not understand the mode of expression in which the desired information is conveyed to us; in other words, if we do not know how to interpret the phenomena produced by the action of our reagents upon the substance examined.

Before we can therefore proceed to enter upon the practical investiga