Gerland. A method based upon the precipitation of tannin by tartrate of antimony and potassium, in the presence of ammonium chloride. Chem. News, 1863, 54; Zeitsch. f. anal. Chem., 2, 419.

Gintl, W. This chemist rejected Pribram's method on account of great variations in the results. Zeit. Chem., 1868, 144; Jsb.d. Chem., 21, 888.

Grassi, E. A modification of Maumene's method. The tannin is precipitated by barium hydrate; the precipitate is washed with ammonium chloride or nitrate solution, neither of which dissolves the tannate of barium. This latter is dissolved in dilute sulphuric acid and titrated with standard per-manganate. Ber. chem. Ges., 1875, 257; Jsb. d. Chem., 1875, 989.

Guenez, E. A volumetric method employing a standard tartar emetic solution, with Poirier's green, 4 J. E., as an indicator. The tannin solution is kept at the boiling point during the titration. It is claimed that gallic acid offers no resistance in this method. Compt. Rend., 1890, 532.

Hammer. This method employs a piece of animal skin to abstract the tannin from solution. Dr. A. B. Prescott (Organic Analysis, 473) says Hammer's method gives untrustworthy results. Journ. f. prakt. Chem., 81, 159; notes on this method, by Muntz and Ramspacher, Compt. Rend., 79, 380; also, Zeitsch. f. anal. Chem., 13, 462.

Handtke. A method based on the reaction of ferric acetate with tannin. Standardize the iron solution by means of a tannin solution of known strength. In practice add the iron solution to the tannin solution until the precipitate settles and the liquid becomes clear. Method not applicable to coffees and teas. Journ. f. prakt. Chem., 82, 345; Zeitsch. f. anal. Chem., 1, 104.

Johanson, E. A modification of Davy's method, employing 5 to 8 drops of a solution of chromic sulphate (one part in twenty-five parts water) per 100 cc of the glue solution. Chem. News, 50, 176. Jackson, R. This method employs lead carbonate as a precipitant of tannin. The percentage of tannin is calculated from the density of the solution before and after the precipitation. An error is possible from the partial solution of the lead by a free organic acid. Chem. News, 50, 179; Zeitsch. f. anal. Chem., 24, 271. Jean, F. A volumetric method based upon the fact that tannic and gallic acids combine with iodine, in the presence of sodium carbonate, in definite proportions. Filter paper covered with starch powder is used in testing for excess of iodine. The presence of gallic acid necessitates a second titration after the removal of the tannin by an appropriate precipitant. Ber. d. chem. Ges., 9, 730; Zeitsch. f. anal. Chem., 16,123. Jean, F. A colorimetric method, in which a tannin solution of known strength (1 per cent) is added to a ferric chloride solution containing free hydrochloric acid, until a piece of white paper placed under the beaker on black cloth is no longer visible. A similar titration is made with a solution of the sample to be assayed. Two titrations are necessary, one before and one after the removal of the tannin by gelatin. Bul. Soc. chim. de Paris., 44, 183; Zeitsch. f. anal. Chem., 28,108. Katreiner. This chemist objects to the methods of Carpene and Barbieri, and states that the probabilities of error are great. Zeitsch. f. anal. Chem., 18, 112. Lowe. In this method the infusion is evaporated with a few drops of acetic acid, and residue is then exhausted with alcohol, which dissolves the tannin only. The tannin is then estimated in the alcoholic solution. Jsb. d. Chem., 18, 743. Lehman. Titration with a glue solution containing ammonium chloride. Jahresber. d. Chem., 1881, 1206; Russ. Zeit. Pharm., 1881, 321. Lowenthal, J. This method is one of the most important yet devised for tannin. It is given in considerable detail, since a modification of this method is usually employed in modern practice. See method as modified by Procter, Conncler, and Schroeder, page 890, this report. The method devised by Lowenthal is based upon the oxidation of tannin by permanganate of potassium. Four solutions are employed: (a) permanganate of potassium, standard solution; (b) 6 grams indigo, with 50 cc concentrated sulphuric acid per liter; (c) 25 grams white glue, dissolved in water and the solution saturated with sodium chloride, then made up to 1 liter; (d) a saturated solution of sodium chloride, with 25 cc sulphuric acid per liter. The sample is extracted in boiling water, the solution is cooled and made up to 1 liter. (1) 10 cc of this extract are diluted to 800 or 1,000 cc, 25 cc indigo solution added, and a titration is made with the permanganate. (2) 25 cc indigo solution are diluted to 800 or 1,000 cc and titrated with permanganate. This burette reading is subtracted from that under (1) to obtain the net reading, i. e., the permanganate required for the tannin and gallic acid and other oxidizable matter in 10 cc of the extract. (3) 100 cc of the extract are now treated with 50 cc glue solution, and, after shaking, 100 cc salt solution are added. After several hours the solution is filtered. To 50 cc of the filtrate 25 cc indigo are added and another titration is made. Under these conditions, again determine the permanganate required tor the Indigo and deduct the burette, rending as before. In the titration under (3) a volume of the solution equal to 20 cc of the extract was taken; hence divide the net burette reading by 2. This gives the permanganate required to oxidize the substances other than tannin in 10 cc of the extract. Deducting this last amount of permanganate from that required tor all the oxidizable matter in the extract, and the remainder is that required for the tannin in 10 cc extract.

The permanganate is standardized with oxalic acid. According to Neubauer, 63 grams crystallized oxalic acid correspond to 41.57 grams purified gallotannic acid.

Zeitsch. f. anal. Chem., 16,33; 20, 91.

Macagno. A method based on the use of mercuric nitrate as a precipitant. Ber. d chem. Ges., 7, 1, 360.