It is not liable to any material fallacy. The salts of cadmium yield with it precipitates nearly of the same colour: but they are exceedingly rare; and the precipitate, unlike sulphuret of arsenic, is insoluble in ammonia.—The salts formed by selenic acid, if decomposed by another acid, also yield yellow precipitates; but these salts are extremely rare.—The salts of peroxide of tin give a dirty grayish-yellow precipitate; which however ammonia turns brown.—A lead solution acidulated with hydrochloric acid gives at first a yellow precipitate; but this becomes brownish-black when more gas is transmitted.[517] The contents of the human intestines sometimes yield a yellowish precipitate though no arsenic be present; and it is dissolved, like sulphuret of arsenic, by ammonia.[518] The tartrate of antimony and potash (tartar-emetic) does not form, as was once thought, any source of fallacy, the antimonial precipitate having always a tint of orange-red; besides it is not, like sulphuret of arsenic, soluble in carbonate of ammonia.—Other fallacies exist, unless the test be used with the precautions mentioned above. But these need not enumeration here. _Ammoniacal nitrate of silver_ is prepared by precipitating the oxide of silver by means of ammonia, from a solution of nitrate of silver or lunar caustic in ten parts of water, and then redissolving the precipitate nearly, but not entirely, by adding gradually an excess of ammonia. When thus prepared, it causes, even in a very diluted solution of the oxide of arsenic, a lively lemon-yellow precipitate of arsenite of silver; which passes to dark brown under exposure to the light.—The action of this test is prevented by nitric, acetic, citric, or tartaric acid in excess, particularly by the first and last. It is also prevented by an excess of ammonia; and in very diluted solutions by the nitrate of ammonia. These facts will suggest the necessity of certain obvious precautions. Its action is obscured by the co-existence of various salts, which singly cause a white precipitate with nitrate of silver; for the yellow colour is then much lessened in intensity. The only one of these requiring special notice, because it occurs in very many of the fluids which are likely to be subjected to the researches of the medical jurist, is common sea-salt, the chloride of sodium. The best way of getting rid of the difficulty is to use in the first instance, not the ammoniacal nitrate, but the simple nitrate of silver, as long as any white precipitate falls down, to add a slight excess of that test, and then, after subsidence, to drop in ammonia. No arsenic is thrown down by the first steps of this process; but if any be present, it is subsequently thrown down in the form of the yellow arsenite of silver, on the addition of ammonia. This simple mode of getting rid of chloride of sodium was first proposed by Dr. Marcet.[519]—Ammoniacal nitrate of silver is of no use as a test for a moderately diluted solution of the