Jaksch speaks highly of _benzopurpurin_ as a test. Filter-paper is soaked in a saturated aqueous solution of benzopurpurin 6 B (the variety 1 or 4 B is not so sensitive), and the filter-paper thus prepared allowed to dry. On testing the contents of the stomach with the reagent, if there is more than 4 parts per 1000 of hydrochloric acid the paper is stained intensely blue-black; but if the colour is brown-black, this is from butyric or lactic acids, or from a mixture of these acids with hydrochloric acid. If the paper is washed with pure ether, and the colour was due only to organic acids, the original hue of the paper is restored; if the colour produced was due to a mixture of mineral and organic acids, the brown-black colour is weakened; and, lastly, if due to hydrochloric acid alone, the colour is not altered by washing with ether. Acid salts have no action, nor is the test interfered with by large amounts of albumins and peptones. A. Villiers and M. Favolle[98] have published a sensitive test for hydrochloric acid. The test consists of a saturated aqueous solution of colourless aniline, 4 parts; glacial acetic acid, 1 part; 0·1 mgrm. of hydrochloric acid strikes with this reagent a blue colour, 1 mgrm. a black colour. The liquid under examination is brought by evaporation, or by the addition of water, to 10 c.c. and placed in a flask; to this is added 5 c.c. of a mixture of equal parts of sulphuric acid and water, then 10 c.c. of a saturated solution of potassic permanganate, and heated gently, conveying the gases into 3 to 5 c.c. of the reagent contained in a test-tube immersed in water. If, however, bromine or iodine (one or both) should be present, the process is modified as follows:--The hydracids are precipitated by silver nitrate; the precipitate is washed, transferred to a small flask, and treated with 10 c.c. of water and 1 c.c. of pure ammonia. With this strength of ammonia the chloride of silver is dissolved easily, the iodide not at all, and the bromide but slightly. The ammoniacal solution is filtered, boiled, and treated with SH₂; the excess of SH₂ is expelled by boiling, the liquid filtered, reduced to 10 c.c. by boiling or evaporation, sulphuric acid and permanganate added as before, and the gases passed into the aniline. The process is inapplicable to the detection of chlorides or hydrochloric acid if cyanides are present, and it is more adapted for traces of hydrochloric acid than for the quantities likely to be met with in a toxicological inquiry. [98] _Comptes Rend._, cxviii.