For the determination of phenols the method of shaking out with dilute sodium hydroxide solution has proven satisfactory.

1) Pharm. Rundsch. (Mew York) 12 (1894), 162; Arch, der Pharm. 282 (1894), 655.

2) Absolute alcohol is preferable to such containing water, since the reduction is more complete and the sodium alcoholate formed is kept better in solution. Beckmann, Journ. f. prakt. Chem. II. 55 (1897), 18.

It was first applied by Gildemeister1) to oil of thyme, is readily carried out and for practical purposes is sufficiently accurate. In as much as all phenols yield water-soluble compounds with the alkalies, the method is quite generally applicable for volatile oils. However, as has been pointed out by Schimmel & Co.,2) the concentration of the alkali3) has to be chosen with reference to the phenol to be determined. For thymol or carvacrol-containing oils (oils of ajowan, thyme, Cretian origanum oil) a five percent, solution is employed as originally suggested by Gildemeister (loc. cit). Eugenol-containing oils (oils of clove, clove stems, pimenta, bay and cinnamon leaves), however, are to be shaken out with a three percent, solution. If in the latter case a stronger solution is used, the results obtained are too high, for the excess of alkali together with the eugenol alkali exercice a solvent capacity on the nonphenol constituents, more particularly those containing oxygen. It may happen that oils with a high eugenol content are completely dissolved in the alkali. For this reason Umney's4) proposition to use a 10 to 12 p. c. caustic potash solution is to be rejected, for the results obtained are always too high. The idea might readily suggest itself always to use a three percent solution. However, it has been shown that thymol and carvacrol are not quantitatively removed by such a solution.

The assay is conducted in the following manner: To 10 cc. of oil contained in a cassia flask that has a capacity of at least 100 cc. (fig. 75, p. 583) enough three or five percent. caustic soda solution is added to fill the flask about four-fifths. The mixture is thoroughly and repeatedly shaken. That portion of the oil which has not entered into reaction is driven into the neck of the flask by the addition of more lye. Gentle tapping or rotation of the flask are resorted to, if necessary, to cause any oily drops that adhere to the glass walls to rise into the neck. After all of the nonphenols have risen into the neck, a process that usually requires several hours, their volume is read off by means of the calibration. Substracting this number from 10 yields the amount of oil dissolved and multiplication of this number with 10 the phenol content of the oil expressed in volume percent.

1) Hager, Fischer and Hartwich, Kommentar zum Arzneibuch fur das Deutsche Reich. 3. Ausgabe. Berlin 1892. 1. Ed., Vol. II, p. 377.

2) Report of Schimmel & Co. April 1907, 118.

3) It should be specially mentioned that it makes no difference whether soda lye or potash lye be used. Because of the higher molecular weight of the latter, somewhat more solution, however, will have to be used.

4) Pharmaceutical Journ. III. 25 (1895), 951.

In the case of oil of cloves, the mixture is heated in a water bath for 10 minutes. The object of this is to saponify the aceteugenol, which is of equal value in the assay of the oil, and to determine it together with the free eugenol.

In place of the cassia flask a large burette can be used. In as much as these burettes, as a rule, have a capacity of but 60 cc, they should be used only where a five percent lye is applicable. Otherwise it might occur that the amount of alkali should prove insufficient for the complete solution of the phenol. In the case of oil of cloves a cassia flask will be used for the reason that the mixture is to be heated on a water bath.

If for the purpose of identification the phenol is to be isolated, the phenol-containing lye is separated from the oil, filtered through a moist, hardened filter, and the phenol regenerated by means of dilute sulphuric acid. The liberated phenol is shaken out with ether, the solution transferred to an evaporating dish, and the ether evaporated. The residual phenol can be identified by means of suitable derivatives. Thymol is identified readily since it solidifies upon cooling either by itself or when inocculated with a fragment of thymol.

Other methods for the quantitative determination of phenols have been worked out by Schryver, by Verley and Boelsing, and by Hesse. However, they are not used as commonly as the one described. Under special circumstances, however, the one or the other method may prove useful, more particularly if only little oil is available.

Schryver1) makes use of the property of sodium amide, NaNH2, to react with the phenols in such a way that the phenol hydrogen is replaced by the sodium and ammonia is set free. The latter is combined with acid and titrated.

The process is carried out in the following manner: About 1 g. of sodium amide is reduced to a fine powder, washed two or three times by decantation with benzene, and transferred to a 200 cc. flask with a wide neck, which is provided with a separating funnel and a condenser. 50 to 60 cc. of benzene free from thiophene are then transferred to the flask and boiled on a water bath. At the same time air free from carbon dioxide is drawn into the flask by means of a suction pump and through the separating funnel, the stem of which passes beneath the surface of the benzene. After boiling for ten minutes the last traces of ammonia adhering to the sodium amide have been drawn off. 20 cc. of normal sulphuric acid are then given into a receiver and this connected with the condenser. Now a solution of 1 to 2 g. of the phenol or of the volatile oil are allowed to flow slowly from the separating funnel into the boiling mixture of sodium amide and benzene. The separating funnel is washed with some benzene and air is sucked through the apparatus until all of the ammonia has been absorbed by the acid in the receiver. This usually requires one and a quarter hours. Finally, the excess of sulphuric acid in the receiver is titrated with sodium carbonate solution, methyl orange being used as indicator. The result is expressed as percent of phenol or, in the case of unknown phenols, as hydroxyl number. The "hydroxyl value" is defined by Schryver as the number of cc. of normal sulphuric acid necessary to neutralize the ammonia set free by 1 g. of substance if the conditions described above are followed.

1) Journ. Soc. chem. Industry 18 (1899), 553.