The development of a reducing gas from a special and somewhat complicated apparatus is not absolutely necessary. The whole process of reduction, from beginning to end, may take place in a single tube by any of the following processes:—(1) The sulphide is mixed with oxalate of soda (a salt which contains no water of crystallisation), and the dry mixture is transferred to a suitable tube, sealed at one end. An arsenical mirror is readily obtained, and, if the heat is continued long enough, no arsenic remains behind—an excellent and easy method, in which the reducing gas is carbonic oxide, in an atmosphere of carbonic anhydride. (2) The sulphide is oxidised by aqua regia, and the solution evaporated to complete dryness. The residue is then dissolved in a few drops of water, with the addition of some largish grains of good wood charcoal (which absorb most of the solution), and the whole carefully dried. The mass is now transferred to a tube closed at one end, a little charcoal added in the form of an upper layer, and heat applied first to this upper layer, so as to replace the air with CO2, and then to bring the whole tube gradually to redness from above downwards. In this case also the whole of the arsenic sublimes as a metallic mirror.
There are various other modifications, but the above are trustworthy, and quite sufficient. Brugelmann’s method of determining arsenic, elsewhere described, would appear to possess some advantages, and to promise well; but the writer has had no personal experience of it with regard to arsenic.
§ 747. Conversion of Arsenic into Arsenious Chloride (AsCl3).—This process, first employed by Schneider and Fyfe, and afterwards modified by Taylor, differs from all the preceding, since an attempt is made to separate by one operation volatile metallic chlorides, and to destroy the organic matter, and thus obtain two liquids—one a distillate—tolerably clear and free from solid particles, whilst the mass in the retort retains such metals as copper, and is in every way easy to deal with.
Schneider and Fyfe employed sulphuric acid and common salt; but Taylor recommends hydrochloric acid, which is in every respect preferable. As recommended by Taylor, all matters, organic or otherwise, are to be completely desiccated before their introduction into a retort, and on these dried substances sufficient pure hydrochloric acid poured, and the distillation pushed to dryness. Every one is well aware how tedious is the attempt to dry perfectly the organs of the body (such as liver, &c.) at any temperature low enough to ensure against volatilisation of such a substance as, e.g., calomel. This drying has, therefore, been the great stumbling-block which has prevented the general application of the process. It will be found, however, that drying in the ordinary way is by no means necessary. The writer cuts up the solid organ (such as liver, brain, &c.) with scissors into small pieces, and transfers them to a retort fitted by an air-tight joint to a Liebig condenser; the condenser in its turn being connected with a flask by a tube passing through an india-rubber stopper dipping into a little water. Another tube from the same flask is connected with india-rubber piping, which is connected with a water-pump, the fall tube of which terminates in the basement of a house over a gully. The distillation is now carried on to carbonisation; on cooling, a second quantity of hydrochloric acid is added, and the last fraction of the distillate examined for arsenic. If any is found, a third distillation is necessary. At the termination of the operation the retort is washed with water, the solution filtered, and this solution and the distillate are each separately examined for arsenic. If properly performed, however, the second distillation brings over the whole of the arsenical chloride,[794] and none will be found in the retort. With the above arrangement there can be no odour, nor is there any loss of substance. In the distillate the arsenic can hardly be in the form of arsenious chloride, but rather arsenious acid and hydrochloric acid; for the chloride easily splits up in the presence of water into these substances. It is best to convert it into the trisulphide. Taylor[795] recommends evolving arsine in the usual way, and passing the arsine (AsH3) into solution of silver nitrate, finally estimating it as an arseniate of silver. Objections with regard to the impurity of reagents should be met by blank experiments. Kaiser[796] has proposed and practised a modification of this method, which essentially consists in the use of sulphuric acid and sodic chloride (as in Schneider and Fyfe’s original process), and in passing the distillate first into a flask containing a crystal or two of potassium chlorate, and thence into an absorption bulb; in the latter most of the arsenic is found in the form of arsenic acid, the chloride having been oxidised in its passage. The apparatus is, however, complicated in this way without a corresponding advantage.[797] Lastly, E. Fischer[798] has shown that it is a considerable advantage to add from 10 to 20 c.c. of a saturated solution of ferrous chloride before distilling with HCl. In this way all the arsenic, whether as arsenic or arsenious acids, is easily converted into chloride.
[794] Dragendorff asserts to the contrary; but we may quote the authority of Taylor, who has made several experiments, in which he obtained all the arsenic as chloride. The writer has performed the process many times, each time carefully testing the mass in the retort for arsenic; but the result proved that it had entirely passed over.
[795] Principles of Medical Jurisprudence, vol. i. p. 267.
[796] Zeitschr. f. anal. Chem., xiv. pp. 250-281.
[797] Selmi (Atti dell. Accademia dei Lincei, Fasc. ii., 1879) proposed a modification of Schneider’s process. The substances are treated with hot, pure sulphuric acid, and at the same time the liquid is traversed by a stream of hydrochloric acid gas. The resulting distillate is tested for arsenic by Marsh’s process. Selmi states that, operating in this way, he has detected 1⁄400 of a mgrm. of As2O3 in 100 grms. of animal matter.
[798] Scheidung u. Bestimmung d. Arsens; Liebig’s Annalen d. Chemie, Bd. ccvii. p. 182.