Brouardel and Bouting (Compt. Rend. 92, 1056) propose the reducing action of ptomaines as a distinction between them and vegetal alkaloids. The solution in weak acid is added to a dilute mixture of ferric chloride and potassium ferricyanide: the latter, if a ptomaine be present, is reduced to ferrocyanide, and Prussian blue is thereby precipitated. But Spica (Gazetta, 11, 486) has shown that strychnia, brucia, morphia, and some others produce this reaction readily, and Beckurts (Arch. Pharm. 3, 20, 104) adds aconitine and others as producing it slowly. Hence the distinction is delusive. See also Husemann (Arch. Pharm. 3, 16, 169; also 3, 20, 270), Tauret (Compt. Rend. 92, 1163). The discovery of these bodies has certainly raised a new difficulty for toxicologists, and suggested a new and plausible defence, as it must be confessed that at present there is no general method of distinguishing between “cadaveric” and vegetal alkaloids. Yet this mainly affects the “physiological” tests—on frogs and other small animals—for there is no ptomaine yet discovered which gives _all_ the reactions of strychnia, morphia, &c. If a chemist be asked, “Could any other substance produce these reactions?” he can only answer, “I do not know of any”; he cannot aver the impossibility. Then the circumstantial evidence must decide. In conclusion, the following principles should be noted:— 1. The quantity of poison found is generally only a small fraction of the quantity taken. The vomit and evacuations are frequently lost, and much may be decomposed by vital actions in the body, or by putrefaction. That which has caused death is probably thereby either decomposed or so combined as to be rendered undetectible: that which is found is merely the surplus beyond the fatal dose. This would account for the frequent non-discovery in the tissues when a small amount has been given, or much time has elapsed. To metallic poisons this does not apply, as, unless eliminated, they can always be found. See further under Strychnia. 2. The symptoms will differ according to the dose, the form (solid or solution, pure or admixed), habit, or idiosyncrasy, the state of health, &c. 3. In the post-mortem examination, appearances common to dead bodies generally are often mistaken for the effects of poison. See Casper’s Handbook, vol. I., _et passim_. 4. Unhealthy or improper food or acute disease may cause suspicious symptoms. This is the commonest solution of suspected poisoning. 5. In experiments on animals, it may be objected that they are inconclusive as to man. This is not strictly true. But if a _recent vomit_ proves poisonous to an animal, _with the same symptoms as in the man_, that is almost conclusive evidence. 6. When a poison is not found by analysis, it does not follow that it has not caused death. Unequal distribution, uncertainty of tests, improper securing of the samples (Palmer case), decomposition or elimination of the poison, may hinder discovery.