In case 20 or even 10 centigrammes of substance are at hand, it is best to convert the alkaloid into its hydrochlorate, and evaporate the solution of this salt to dryness. The residue, after being weighed, is dissolved in water, and a solution of sulphate of silver added. The precipitate of chloride of silver formed is collected and carefully weighed, in order to calculate the weight of the chlorine contained in the hydrochlorate and consequently the molecular weight of the alkaloid. The filtrate from the chloride of silver, which contains the alkaloid in the state of sulphate, is treated with hydrochloric acid, to remove the excess of silver present and the fluid then filtered. The filtrate is next shaken with potassa and ether. Upon decanting and evaporating the ethereal solution, a residue consisting of the alkaloid present is obtained, which is then purified by crystallization from alcohol. An elementary analysis of the alkaloid is now executed. _Certainty_ as to the presence of an individual alkaloid is attainable only when the execution of this confirmatory test is possible. The reactions previously described can be performed with fifteen centigrammes of substance, and this amount is sometimes contained in a cadaver. If but one or two centigrammes are at hand, it is still possible to detect the presence of an alkaloid; a conclusion, however, as to _which_ cannot be arrived at, especially if the substance found is a liquid or an amorphous body, and one that presents few distinctive properties. III. METHODS TO BE EMPLOYED, WHEN NO CLEW TO THE NATURE OF THE POISON PRESENT CAN BE OBTAINED. If poisoning has been caused by the administration of a mixture of numerous substances and these greatly differ in their properties, it is impossible to demonstrate in an incontestible manner the presence of each individual poison. This contingency fortunately but seldom arises; the criminal usually has recourse to one or two poisons, the detection of which is possible. It must not be imagined, however, that the presence of a poison in an organ can at once be detected with certainty by the mere application of a few tests; because, in searching for a substance which is absent, we may unwittingly destroy the one present, or, at least, transform it into combinations which would not allow of a definite conclusion as to its original condition.