specimen itself deflagrates upon charcoal from the presence of a nitrate or chlorate, in which case a slight explosion results from the reaction of the cyanide and oxisalt.

The specimen having been intimately mixed with five or six times its bulk of flux, a small portion of the resulting powder, sometimes moistened with water so as to make it cohere, is heated strongly on charcoal. The mixed mass should fuse readily before the blowpipe, so that any minute globules of reduced metal may run together. If not readily fusible, a fresh mixture must be taken with a larger proportion of flux. When the reduced metal volatilises at the temperature employed, its vapour becomes oxidised outside the flame, and is deposited upon the charcoal as a more or less abundant, white or coloured, incrustation. Silver gives no incrustation, and tin scarcely any lead gives a yellow, and bismuth a brownish-yellow, incrustation; while antimony gives an abundant easily volatile bluish-white, and cadmium a comparatively fixed brown-red incrustation. Metallic antimony vaporises rapidly; while cadmium is so volatile that its reduction and vaporisation are simultaneous, wherefore no globule, but only an incrustation, is producible with it. Zinc compounds, heated with reducing flux, behave in this respect like cadmium, furnishing no globule of metal, but only an incrustation, which is yellow when hot, white when cold; but zinc will have been previously detected by its reaction with nitrate of cobalt (y). Reduced arsenic and mercury are so volatile that they can only be obtained in the form of sublimates by performing the reduction in tubes as already described (a).

The different metals usually identified by their reduction on charcoal, exhibit the following characters:-Silver yields a bead of white, moderately hard and malleable metal, with no incrustation. Tin, which is less easily reducible, yields a bead of white malleable metal, softer than silver, with very slight, if any, incrustation. Lead yields a bead of soft bluish-white metal, with yellow incrustation; bismuth, a bead of brittle yellowish-white metal, with brownish-yellow incrustation; antimony, a bead of brittle bluish-white metal, with an abundant

bluish-white incrustation; and cadmium, no metallic bead, but only a reddish-brown incrustation.

In addition, copper, iron, nickel, and cobalt are reducible, though they do not afford incrustations, and are not easily obtainable in the form of beads. But on crushing the fused mass in a mortar, and washing away the lighter portions, copper may often be recognised in the form of red spangles, and iron, nickel, and cobalt as heavy powders affected by the magnet.

Fig. 35.

Ε. Compounds of those metals which tinge the borax bead usually leave dark-coloured infusible residues when heated alone on charcoal. In testing by the borax bead, a piece of platinum wire is bent into a single or double loop, as shown of the actual size in fig. 35, the loop dipped into powdered borax, and the adhering borax heated to redness, when it first undergoes a highly characteristic intumescence, and afterwards, when more strongly heated, sinks into a colourless transparent bead. To this bead is then attached a minute quantity of the substance under examination, and the whole strongly heated in the blowpipe flame, when in some cases the fused bead dissolves the specimen, and thereby acquires a more or less definite colour, the depth of which may be increased by adding more of the specimen and again heating strongly.

The metals iron and copper, which form two classes of salts, also form beads of two colours. Thus in the oxidising flame we have a blue cupric and a yellow-brown ferric bead, while in the reducing flame we have a cuprous bead of an almost colourless or opaque reddish aspect, and a ferrous bead of a sea-green colour. The chromium bead has an emerald green, and the cobalt bead a sapphire blue colour. Manganese, when free from iron, imparts an amethystine tint, and nickel a deep sherry hue, which becomes amethystine when the bead is heated with a

fragment of nitre. The borax may be replaced by microcosmic salt, or even by ordinary glass. In these several reagents we have, after ignition, an excess of melted boric, phosphoric, or silicic anhydride, which at the temperature of the blowpipe flame combines with the various metallic oxides to form coloured fusible salts.

8 II.-SOLUTION AND PRECIPITATION.

(27.) Having made his examination in the dry way, by means of the blowpipe, the student must next bring his substance, by some means or other, into a state of solution, so that he may submit it to the action of liquid reagents. As a general rule, the substance to be dissolved should be in a finely divided state. This is particularly necessary in the case of bodies which are with difficulty soluble, such as many native oxides, sulphides, &c. Any substance having a decided colour, a hard structure, and an opaque aspect, whether earthy or lustrous, ought always to be pulverised very finely before being treated with solvents. The solution of the body under examination should be effected by preference in water; but, if insoluble in water, it may be acted upon with hydrochloric acid, or with nitro-muriatic acid, or with nitric acid.

A small portion of the powdered substance is to be placed in a test-tube, a moderate quantity of water* added, the whole agitated, and heated over a spirit or gas-flame. While heating, the tube should receive an occasional jerk, to facilitate mixture and avoid the sudden escape of vapour. If the substance, by this treatment, is obviously dissolved, the clear solution, filtered if necessary from any insoluble portions, can be submitted at once to the action of reagents. If the substance, however, is not obviously dissolved, a few drops of the liquid may be filtered on to a glass slip and gently evaporated to dryness. Should any

* By water is always meant pure or distilled water; but clean rain water may sometimes be employed as a substitute.

definite amount of residue remain upon the glass, the whole mix. ture must be thrown upon a filter, and the tests applied to the clear filtrate. There are many substances which, unless taken in very small quantity, do not disappear perceptibly when boiled in water, but yet are sufficiently soluble to afford an aqueous solution that can be successfully tested. Should a mere trace of residue, or none at all, be left upon the glass slip, as much of the water as possible is to be poured away from the insoluble substance, and replaced little by little with hydrochloric acid, warming between each addition. Should any obvious action occur, more hydrochloric acid may be added, if necessary, and the whole heated for some time until an available solution is formed. Should there be no obvious action, nitric acid must be added in the proportion of about one-fourth of the hydrochloric acid previously used, and heat again applied. By one or other of these means a solution will generally be effected. There are, indeed, a few substances which dissolve in nitric, but neither in hydrochloric nor in nitro-hydrochloric acid. There are also some substances which are quite insoluble in any ordinary menstruum; the consideration of these, however, is deferred for the present. The solution of the substance, whether in water or acid, to which no reagent has been added, is called in the tables and elsewhere the original solution. The freshly-made acid solution should generally be diluted somewhat freely with water, and filtered if necessary. The acidity, neutrality, or alkalinity of the aqueous solution should be ascertained by means of test-paper.

(28.) Different reagents are next to be added to the original solution in the order and manner described in paragraphs 30, 35, and 41, and in the Tables I. II. III. and V. These reagents produce in the solution certain effects, which are characteristic of the various substances dissolved. The effect most usually produced by a liquid reagent is to cause a precipitate or solid deposit of some insoluble compound of the substance sought for. Hence the formation or non-formation of a particular precipitate usually proves the presence or absence of a particular base or acid in the solution under examination. Precipitates differ much in their

colour, consistency, rapidity of formation, and solubility in different liquids, whence the student must make himself familiar with their various aspects and habitudes. As regards aspect, he must notice whether a precipitate is dense, crystalline, clotty, gelatinous, opaque, transparent, coloured or colourless, &c. As regards habitude, he will find that crystalline precipitates, unless thrown down from concentrated solutions, do not usually appear at once, but only after some little time. Their immediate formation, however, may be often determined by rubbing the liquid against the inside of the containing vessel with a glass rod. Again, many precipitates are characterised by their solubility in an excess of the precipitant, or in some other reagent.

Reagents and solvents should always be added gradually, except when special direction is given to the contrary. This rule is of great importance, and applies equally to the formation and solution of precipitates; in the latter case, the mixture should be agitated between each addition of the solvent. Many characteristic effects are occasionally overlooked through a neglect of this rule. The student must also bear in mind, when directed to employ an excess of any particular reagent or solvent, that every minute quantity more than sufficient to produce the desired effect is an excess.

In the tables, the word dissolved placed at the head of a column signifies either that the substances written under it have not been precipitated at all, or that, having been precipitated, they are now redissolved by an excess of the reagent, in any case that they remain in solution.