big's Jahresbericht, 1851, p. 620; Fyfe, Journal für pract. Chemie. Bd. 55, p. 103.)

Different ways have been proposed for the further treatment of the distillate thus obtained. It is either to be introduced, without further preparation, into Marsh's apparatus, or to be treated with sulphureted hydrogen, or to be tested with nitrate of silver. Being strenuously opposed to introducing liquids containing hydrochloric acid into the apparatus of Marsh, I cannot recommend the first of these methods. The large quantity of hydrochloric acid which is present in the distillate requires, at all events, a very gradual addition. Volatile organic bodies, formed by the action of the hydrochloric acid on the organic matters, and which always occur in the distillate, contribute, besides, to influence the accuracy of the result. Nor is the proposal of Fyfe, to test the distillate with nitrate of silver, a very recommendable one, considering the large amount of nitrate of silver which has to be added to remove the chlorine, the consequent accumulation of nitric acid in the liquid, and the formation of a considerable quantity of nitrate of ammonia on neutralizing the free acid with ammonia. The precipitation of the arsenic by means of sulphureted hydrogen, appears to be, after all, the best way. But this is, according to my experience at least, not very easily accomplished, on account of the presence of the organic matters; the liquid assumes a dark yellow color, and, if only a minute quantity of arsenic is present, the sulphide is but very slowly deposited. This process, finally, possesses, in common with many other methods, the disadvantage of leading exclu

sively to the detection of arsenic, and not, at the same time, to the detection of other poisonous metals; it is only antimony and, perhaps, tin, that may also distill over, as chlorides, into the receiver.

§ 44. At an earlier period, it was recommended to treat the organic substances with caustic potassa, in order to extract the arsenic. The food, contents, etc., were boiled with a solution of caustic potassa, and the liquid afterwards acidulated with hydrochloric acid; the albuminous bodies are thereby precipitated in a finely divided state. After this method, an acid liquid is obtained, which is always very rich in organic matter, and in which sulphureted hydrogen produces a copious precipitate, even if not a trace of arsenic is present. It should also be borne in mind. that, on boiling albuminous substances with caustic potassa, sulphide of potassium is formed, which, on being decomposed with hydrochloric acid, gives rise to the formation of sulphureted hydrogen; this acts on the arsenious acid which has entered into solution and precipitates sulphide of arsenic. In an experiment, conducted after this method, I failed to detect. the arsenic. In the opinion of Pettenkofer, the sulphide of potassium becomes readily oxidized on exposure to air; still, he recommends, as the safer method, to treat the alkaline solution with some carbonate of lead, which removes the sulphur.

§ 45. Another method prescribed to boil the organic substances with a solution of caustic potassa until as complete as possible a solution was obtained, to acidulate the liquid with hydrochloric acid, and to pass chlorine-gas through it until the whole of the organic matters were destroyed; that is, until the

liquid, after twenty-four hours resting in a covered vessel, smells of free chlorine. The excess of chlorine, and the hydrochloric acid formed during the process, should then be driven off by heat, and the liquid be treated with sulphureted hydrogen. To this method it must be objected that the treatment with chlorinegas is an exceedingly laborious operation, which is very conveniently substituted by the treatment with chlorate of potassa, as given § 9, and that some arsenic may be lost in the evaporation of the acid liquid.

§ 46. Since sulphureted hydrogen precipitates the arsenic completely from its solution, even in presence of ammoniacal salts and organic substances, there is no longer any doubt entertained that sulphureted hydrogen gas is the best precipitant for arsenic from a solution, prepared after any of the previously mentioned methods. What speaks, moreover, in favor of the reagent is the circumstance that it throws down, at the same time, other poisonous metals which may be present, and which might have caused the death of the individual. The method of Valentin Rose, which I shall now notice, and in which lime-water is used for the precipitation of the arsenic, has, therefore, only a historical interest. The substances under examination are boiled with a very dilute solution of caustic potassa (one part of caustic potassa to forty-eight of water), the liquid is filtered through cloth, the filtrate heated to ebullition, and, while in this state, mixed with nitric acid, in small portions, until a fresh addition causes no longer a separation of solid matter, and until the liquid shows a strong acid reaction, and has become limpid and of a pale yellow color. When cold, the

liquid is filtered through paper, the filtrate mixed with carbonate of potassa, almost to the point of neutralization, and boiled to remove the carbonic acid. To the limpid liquid of feebly acid reaction, lime water is added in excess, and the whole heated to ebullition, to facilitate the deposition of the precipitate, which consists of arsenite and phosphate of lime, together with some organic matter. But the presence of organic matter, and of salts, especially sal ammonia, prevents the complete precipitation of the arsenite of lime, and hence it is that small quantities of arsenious acid cannot be detected by this method.

§ 47. Arsenate and arsenite of lime are very good materials for the production of arsenic-mirrors, but the precipitate obtained in the manner prescribed by Valentin Rose, is not fitted for this experiment, on account of the organic matter which it contains. The lime-salt must be precipitated from solutions which are free from such substances.

A liquid, well adapted for this purpose, may be obtained by repeatedly treating the crude sulphide of arsenic with nitric acid; or, better still, by carefully digesting the crude sulphide with hydrochloric acid, with occasional additions of chlorate of potassa. The resulting liquid is mixed with a solution of caustic potassa until it shows a weak alkaline reaction, and then precipitated with freshly prepared limewater; the whole is heated to ebullition, when the arsenate of lime becomes granular, and settles easily to the bottom.

If there is a sufficient quantity of the precipitate, it may be collected on a filter and washed; but, if

the amount is very small, it must be washed by decantation. In this case, the hot liquid with the precipitate is poured into a conical-shaped glass, and the precipitate allowed to settle; the clear liquid is then drawn off by means of a pipette, hot water, to which some lime-water has been added, poured on the sediment, and, after a while, drawn off again. If necessary, the process is once more repeated. The portion of the liquid which remains in the glass after the last decantation, and which contains the precipitate in suspension, is transferred into a porcelain dish, or a watch-crystal, and evaporated to dryness.

The reduction is effected in the following manner: the residue, or the precipitate from the filter, is mixed with some freshly ignited charcoal-powder, and the mass placed in the bulb of the reductiontube (Fig. 9, p. 59); the tube is cleaned in the manner described § 35, the mixture carefully dried, and then the bulb heated to redness, by means of the flame of a spirit-lamp and the blow-pipe.

If the mixture was properly dried, and the tube clean, a very excellent mirror is produced. The compounds of antimony afford, under the circumstances, no mirror; the appearance of the mirror is, therefore, a conclusive proof for the presence of arsenic. The mirror, after having been cut off, may serve for other experiments. A mixture of cyanide of potassium and carbonate of soda answers still better than charcoal; the reduction is then executed as taught § 35.

§ 48. Instead of precipitating arsenic acid with lime-water, it is much to be preferred to effect the precipitation of the acid by means of a soluble salt