The crude sulphide of arsenic is oxidized by means of nitric acid, the acid completely removed by evaporation, and the residue (to which a few drops of caustic potassa may be added), mixed with about six or eight parts of carbonized cream of tartar. The whole is made into a thick paste with a little water, and the mass shaped into small cylinders; these are then reduced, after perfect dessication, in hydrogen gas, as described in the preceding paragraph. I can assure, from my own experience, that by operating after this method, a pure mirror is obtained from a very minute quantity of arsenic, provided the destruction of the organic matters has been complete. It is always best to employ a perfectly pure sulphide of arsenic, prepared from the crude sulphide, in the manner described § 13. All that is true regarding the volatilization of the arsenic in the process of Berzelius, is also true in this process.

A more serious objection than the one just mentioned may be raised against the methods of Berzelius, and Duflos and Hirsch. It is this: that the compounds of antimony afford also a mirror, owing to the volatility of metallic antimony in a current of hydrogen gas; and it is on this account, more especially, that the method of Fresenius and Babo deserves the preference. The antimony-mirror may, however, be distinguished from that of arsenic by the reactions given in § 26. No spots of antimony can be obtained on porcelain during reduction; it is only on trying to drive the metal into the narrow part of the tube (an operation much more difficult to perform in this case than in that of arsenic), that these spots are formed; they are readily distin

guished from arsenic-spots by the tests given in $ 27.

[$ 52. The process of H. Reinsch is founded on the observation that, from a dilute and acid solution of arsenious acid the arsenic is deposited, in the metallic state, on metallic copper, which is placed in the liquid. This deposition takes quickest place, when the solution is boiling and air has free access. The details of the process are as follows: the substances are cut into small fragments and mixed with hydrochloric acid, to the amount of about a tenth of the whole mixture, and more, if the subject of analysis is decayed or ammoniacal, so that there may be a decided excess of acid. The mixture is boiled gently for an hour, or until all soft solids are either dissolved or broken down into fine flakes or grains; and then strained through calico. The filtered fluid is again brought to the boiling point, and some bright copper leaf or fine copper gauze placed in it. The copper becomes covered immediately, or after some ten or fifteen minutes, with a thin, brittle, and steellike coating of metallic arsenic. To prove that the deposit is really arsenic, the copper is removed from the liquid, dried with a gentle heat, and cut into small shreds. These shreds are heated in a glasstube, narrowly drawn out at one end, when a sublimate of arsenious acid will be formed; this may be dissolved in caustic potassa, and the solution used for further experiments. Or the copper-leaf is heated to redness in a current of hydrogen gas when arseneted hydrogen is formed, which may be tested in the usual manner.

This method, though highly recommended by some

toxicologists, is open to many serious objections. On heating the copper with access of air, arsenite of copper may be formed, and some arseniuret of copper remain undecomposed, when, possibly, so minute a quantity of arsenious acid may sublime that it is insufficient for the tests. Arseneted hydrogen is not so easily formed under those circumstances, as usually stated, and brown spots may be deposited on metallic copper, in presence of hydrochloric acid and organic matters, even if no arsenic is present. If the arsenic occurs as sulphide, it escapes detection altogether, and if not arsenic, but another metal was the toxic substance, its presence is not indicated by this process. The last-mentioned objection is of great weight, the more so, as another poisonous metal-copper-is brought into the liquid under examination.]

[§ 53. An additional difficulty in the detection of arsenic in legal cases, occurs when the hydrate of sesquioxide of iron, which has been given as antidote, is not free from arsenic. In all cases, therefore, where this substance has been administered, it should be the first care of the analyst to subject a portion of it to Marsh's test. If this caution has been neglected, and the hydrate of iron is subsequently found to contain arsenic, the time and labor bestowed on the investigation is lost and valueless. If arsenic is found in the antidote, the contents of the stomach and the intestines have to be repeatedly washed with water, in order to effect as complete as possible a separation of the oxide of iron from other matters of less or greater density. If, during this process, particles of solid arsenious acid are found, which may have escaped the influence of the iron, the case offers

no further difficulties. But, if the analyst is not so fortunate to discover any arsenic in substance, it becomes very difficult and, frequently, impossible, to decide by chemical analysis, whether the arsenic detected in the body was introduced before, or only with the antidote, in other words, whether poisoning has been attempted or not. At all events, the liquid portion of the mass should be separated from the solids by filtration, and examined separately; if arsenic is found here, it is much more likely that the poison was administered alone, than that it was introduced with the antidote. A negative result, however, would by no means justify the conclusion that no poisoning has taken place, since the antidote may have fixed the whole of the arsenious acid. If any vomited matters, ejected before the administration of the antidote, are at the disposal of the analyst, an examination of these may lead to a satisfactory conclusion. But, if these also are wanting, there remains nothing to be done but to make a comparative quantitative analysis of the oxide of iron collected from the stomach and the oxide administered as antidote. From the difference in the amount of arsenic, found in either case, some conclusions may be drawn as to the nature of the case.

Without entering into the details of an investigation of this kind, which, in all probability, will but rarely be required, because it will rarely lead to a satisfactory result, I shall add a few remarks as to the best mode of procedure.

A portion of the oxide of iron found in the stomach, is repeatedly washed with water, then dried, and the weight ascertained.

The destruction of the adhering organic matter, and the conversion of the arsenious into arsenic acid, is effected by means of hydrochloric acid and chlorate of potassa, as described in § 9. The resulting liquid is mixed with tartaric acid in sufficient quantity to prevent the precipitation of sesquioxide of iron by ammonia, which is added in excess. To the ammoniacal solution a mixture of chloride of ammonium and sulphate of magnesia is added, when the whole of the arsenic acid is precipitated as arsenate of magnesia-ammonia (2 Mg O. NH* O), As O3+HO; the precipitate is collected on a filter, washed with ammoniacal water, dried at 212°, and weighed. 100 parts of the salt correspond to 52.10 parts of arsenious acid.

A portion of the hydrate of iron, used as antidote, is treated in the same manner.]

54. The following list, which contains the various apparatus and materials necessary for a legal analysis, may be found convenient for those who have not a well-furnished chemical laboratory at their disposal:

Large and small-sized porcelain-dishes (Berlin porcelain).

Small and very thin porcelain crucibles, to be used over a spiritlamp.

Beaker-glasses, or Florence flasks, of various sizes.

Apparatus and materials for the evolution of sulphureted hydrogen, with washing bottle.

Several apparatus for the evolution of gas, of various sizes, for the tests of Marsh, and Fresenius and Babo.

Glass-tubes, especially the kind known as reduction-tubes of hard glass which is free from lead).

Spirit lamps with Argand burner.

Pure zinc, pure hydrochloric, sulphuric, and nitric acids, pure chlorate of potassa, nitrate and carbonate of soda, ammonia, etc. Strong