produced, containing a dark nucleus of undissolved metallic arsenic. (Guy.) f. Conversion into arsenic acid. The arsenical stain disappears completely when gertly warmed with a drop or two of nitric or nitro-muriatic acid. By evaporating to dryness, a slight residue of arsenic acid is left, recognisable by its ready solubility in a drop of water, so as to form a solution in which nitrate of silver produces a brick-dust-red turbidity.

(136.) MODIFIED MARSH'S PROCESS.

Marsh's apparatus may also be used for generating arsenetted hydrogen when it is intended to decompose the gas by extraneous heat. In this case the jet of the stopcock is unscrewed and replaced by a metal elbow, to which a horizontal piece of narrow hard glass tubing is adapted. But some such apparatus as that shown in fig. 47 is on the whole to be preferred. It consists of a

small flask furnished with a cork through which passes a long funnel tube, and a short wide rectangular tube, loosely plugged with a little cotton wool, and connected by means of a perforated

cork with a long horizontal piece of narrow hard glass-tubing, bent downwards at its extremity so as to dip into a solution of nitrate of silver. The apparatus is charged with pure zinc and dilute sulphuric acid, so as to cause an evolution of hydrogen gas, which passes through the cotton wool, where it deposits any mechanical impurities, along the horizontal tube, and into the silver solution. During the transmission of the gas, the flame of a spirit lamp is applied steadily to some particular part of the horizontal tube, and if after a little time there is not produced any deposit within the tube, or precipitate in the silver solution, the materials are known to be pure, and the arsenical liquid may be introduced.

a. Decomposition by heat. The resulting arsenetted hydrogen is decomposed in its passage through the heated portion of the tube, and deposits a steel-grey ring of metallic arsenic at some little distance beyond the flame. The ring may be identified by its appearance, its position at a little distance beyond the flame, its volatility, its conversion into arsenious anhydride by repeated sublimations, and its conversion into arsenic acid by treatment with nitro-muriatic acid and evaporation to dryness.

B. Decomposition by nitrate of silver. Any arsenetted hydrogen that may escape decomposition by the flame, or that may be purposely allowed to escape, is arrested by the nitrate of silver solution, with formation of a black deposit of metallic silver. On the termination of the experiment, the excess of silver may be precipitated with hydrochloric acid, the filtrate evaporated to dryness, and the residue of arsenic acid dissolved in water and tested with the usual reagents.

(137.) REINSCH'S TEST.

This test is particularly useful for the detection of arsenic in organic liquids or solids. The suspected liquid is simply acidulated with about one-eighth of its bulk of pure hydrochloric acid, and boiled. The solid tissue is cut up into very small pieces and boiled for some time in a mixture of about one part of hydrochloric acid with six of water.

a. Deposit on copper. A small piece of clean copper foil, or preferably of fine gauze, is introduced into the hot acidulated liquid, and the boiling continued for a period varying from a few minutes to a quarter of an hour or longer. Should the copper acquire a grey metallic discolouration, other pieces thereof may be added from time to time, and the supply continued so long as the last added piece assumes any perceptible alteration in colour.

6. Character of deposit. The pieces of coated foil or gauze are removed from the liquid, washed in water, and dried between folds of bibulous paper. The deposit unless very thick adheres firmly to the subjacent copper, presents a well-marked metallic lustre, and has a dark steel-grey colour, or, if very thin, a somewhat bluish tint. On the application of heat it disappears entirely, while the copper resumes its ordinary appearance.

y. Crystalline sublimate. A piece of the coated copper, held between the fingers, is warmed over a flame, coiled up into a small bulk, and introduced into an ordinary reduction tube (fig. 45). The heat of a small spirit flame is then carefully applied, at first a little above the coil, and afterwards to the coil itself, whereby the arsenic is volatilised, oxidised, and condensed in the cool part of the tube as a crystalline sublimate. If necessary, several pieces of coated copper may be thus heated successively in the same tube until a sufficiently obvious sublimate of arsenious anhydride is produced, which, when examined by a lens or the low power of a microscope, will exhibit highly iridescent octahedral forms. A small piece of tubing open at both ends, one of which is drawn out to a long, almost capillary, termination, as

Fig. 48.

shown in fig. 48, is convenient for volatilising a very slight deposit. The coated foil having been introduced and tilted down

to the shoulder, the tube is sealed by the blowpipe at the point a, and the resulting cylindrical bulb containing the foil heated in a spirit-flame, from its capillary shoulder backwards to its sealed extremity, whereby a crystalline ring becomes condensed in the capillary projection at b.

8. Reactions of the sublimate. By means of a couple of file-marks, the short length of tubing containing the sublimate may be broken off from the two ends of the tube, and the sublimate itself be acted upon by reagents. Moistened with sulphide of ammonium solution and dried in a water-bath, it yields a yellow residue of orpiment. Moistened with a mixture of nitric and a little hydrochloric acid and evaporated to dryness, it leaves a slight residue of arsenic acid, which produces a red turbidity when treated with a drop of nitrate of silver solution.

(138.) IMPEDIMENTS TO REINSCH'S TEST.

a. Influence of oxygenants. Reinsch's process is not applicable in the presence of oxidising bodies, which moreover enable dilute hydrochloric acid to dissolve metallic copper. But the majority of such compounds may be reduced by the action of sulphite of sodium upon the acidified liquid, while any excess of sulphurous acid from the decomposition of the salt may be got rid of by ebullition before introducing the foil or gauze.

B. Purity of the acid. This may be ascertained by diluting a sufficient quantity of the hydrochloric acid with about four times its bulk of water, and boiling a very small piece of foil or gauze in the diluted liquid for a period of twenty minutes or half an hour.

7. Purity of the copper. As even in the most satisfactory performance of Reinsch's test, there is always some, although but an extremely minute quantity of the copper dissolved, and as commercial copper is rarely quite free from arsenic, it is important that the foil or gauze employed in the experiment should be specially tested as to its purity. If, however, the

solution of four or five grains of the copper does not yield any evidence of arsenic, the metal is quite pure enough for the purpose, even though a trace of arsenic should be detected in a larger quantity of it. A few grains of the copper cut into fine pieces are placed in a small tube-retort, or in a bulb-tube, such as that shown in fig. 14, with not less than twice their weight of precipitated peroxide of iron, and an excess of hydrochloric acid. The mixture is then distilled to dryness, great care being taken at the last to prevent spurting. Any arsenic originally contained in the copper is in this manner carried over in the form of chloride of arsenic, and may be condensed in a little water with the excess of aqueous hydrochloric acid. The resulting liquid may then be tested for the presence of arsenic by boiling in it a fresh piece of clean copper gauze or foil.

d. Modified processes. The peroxide of iron mentioned above may be replaced by an equivalent quantity of perchloride of iron. Indeed it is better to dissolve the peroxide in excess of hydrochloric acid, and then employ the residue left on evaporating to dryness, which will be free from any trace of arsenic the peroxide itself may have originally contained. Moreover, oxide or chloride of copper may be substituted for the peroxide or perchloride of iron, though not with advantage. Or the copper may be dissolved in hydrochloric acid alone, without the addition of any special oxygenant, by moistening the metal with the acid and exposing both to the air for several days. The addition of a few drops of a solution of perchloride of iron or chloride of copper to the acid greatly facilitates this solution by exposure to air. The hydrochloric solution, no matter how obtained, is eventually distilled to dryness, and the distillate tested for arsenic.

(139.) OTHER FORMS OF ARSENIC.

a. Orpiment and realgar. These sulphides of arsenic are yellow or orange-coloured compounds, which volatilise unchanged upon the application of heat. Mixed with soda flux and heated in reduction tubes, they give rise to sublimates of metallic