Secondly, it is thought that certain poisons, such as hydrocyanic acid, strychnia, alcohol, conia, and some others, produce their remote effects with a velocity, which is incompatible with any conceivable mode of action except the transmission of a primary local impulse along the nerves, and more especially incompatible with the poison having followed the circuitous route of the circulation to the organs which are affected by it remotely. Thus in regard to the hydrocyanic acid, Sir B. Brodie has stated,[8] that a drop of the essential oil of bitter almonds, which owes its power to this acid, caused convulsions instantly when applied to the tongue of a cat; and that happening once to taste it himself, he had scarcely applied it to his tongue, when he felt a sudden momentary feebleness of his limbs, so that he could scarcely stand. Magendie,[9] speaking of the pure hydrocyanic acid, compares it in point of swiftness of action to the cannon ball or thunderbolt. In the course of certain experiments made not long ago with the diluted acid by Dr. Freer, Mr. Macaulay and others,[10] to decide the true rapidity of this poison, several dogs were brought under its influence in ten, eight, five, and even three seconds; during an experimental inquiry I afterwards undertook for the same purpose,[11] I remarked on one occasion that a rabbit was killed outright in four seconds; and Mr. Taylor has more recently stated, that he has seen the effects induced so quickly in cats, that there was no sensible interval of time between the application of the poison to the tongue and the first signs of poisoning.[12] Strychnia, the active principle of nux-vomica, acts sometimes with a speed little inferior to that of hydrocyanic acid; for Pelletier and Caventou have seen its effects begin in fifteen seconds.[13] Alcohol, according to Sir B. Brodie,[14] also acts on animals with equal celerity; for when he introduced it into the stomach of a rabbit, its effects began when the injection was hardly completed. Conia, the active principle of hemlock, is not less prompt in its operation: when it was injected in the form of muriate into the femoral vein of a dog, I was unable, with my watch in my hand, to observe an appreciable interval between the moment it was injected and that in which the animal died;[15] certainly the interval did not exceed three or at most four seconds.