3. A ten per cent. solution of iodide of potassium. Dissolve 0.1 to 0.2 gramme of the volatile oil in 10 cubic centimeters of chloroform, and add first 15 cubic centimeters of the iodine-mercuric chloride solution; let stand three or four hours, and, in case the mixture gets discolored, add a few more centimeters of solution. Now add 10 to 15 cubic centimeters iodide of potassium solution, dilute with 150 cubic centimeters of water, and titrate with hyposulphite till the mixture remains clear for about a minute. The iodide of potassium solution must be added before the water, and the relative proportions between this solution and the iodine-mercuric chloride solution must be 15 to 20 cubic centimeters. The quantity of iodine solution consumed is calculated to iodine for 100 parts and the figure thus obtained is designated as the "iodine number." Barenthin has in this manner determined the iodine number of several volatile oils; other experimenters, however, for instance, Kremel and Davies,[4] have found different numbers for the same oils, so that this method requires further thorough examination before it can be classed as available. [4] Pharm. Centralh. 1888, S. 482 u. 555; 1889, S. 133. VI. A. Kremel has endeavored to utilize _titration or saponification with alcoholic potash lye_ for the examination of volatile oils. In his experiments he was guided by the following points: A series of volatile oils contains partially free organic acids, like oils of bitter almonds and cinnamon, and partially aldehydes or other combinations. Now it seems not impossible, that up to a certain limit, the quantities of these combinations in the separate volatile oils remain constant, thus presenting the opportunity of testing the respective oils as to their quality and purity by saponification. In some cases these combinations are the chief bearers of the specific odor, and hence the determination of the "_saponification number_" becomes of double value. It is, of course, self-evident that not every volatile oil can be saponified, and Kremel admits that, even where saponification takes place, it is not in every case a sure test.