This particular view of the indications of the liquid tests, however obvious it may seem, has been often overlooked by the numerous chemists and medical jurists who have written for and against them. The antagonists of the tests have been content with proving how so many fallacies lie in the way of each, that no dependence can be put in any one of them: They have not considered that the fallacies attached to one are obviated by the conjunct indications of the others. I am of opinion therefore that the analysis for arsenic by liquid reagents has been unjustly neglected in the present day. It is an exceedingly convenient method, and one of extreme delicacy, because by using small tubes it is easy to operate with precision on very minute portions of a suspected fluid. It is also perfectly conclusive, so far as chemical knowledge now goes. On a remarkable trial a few years ago in this country, a distinguished chemist, who, as witness for the prisoner, was made by counsel to throw discredit on the liquid tests individually, nevertheless admitted to the counsel for the prosecution, that no other substance in nature but arsenic could produce the same effects as it with the whole three tests in succession. _Reduction process._—The process by reduction of arsenic to the metallic state, as applied to the poison in a state of solution, consists in separating the whole arsenic by a liquid test in such a state as to admit of the precipitated compound being subjected to the process of reduction and sublimation. The best method of the kind is a modification of one described by me in 1824.[523] This consists in throwing down the whole arsenic in the form of sulphuret by means of hydrosulphuric acid, converting the sulphuret by the process of reduction to the metallic state, and oxidating the metal thus procured. The hydrosulphuric acid is preferred to other liquid reagents, because the precipitate it forms, while possessing a very characteristic colour, is also more bulky than those caused by the other tests, and is therefore more easily collected,—and because its action is not liable to be prevented or obscured by so many disturbing causes. The steps of the process are the following:—