52. ESTIMATION OF TANNIN. 43 



satisfactory results in some few cases only, because, even if the 

 solution be mixed with chloride of ammonium, it is difficult to 

 ascertain when a sufficient quantity of the reagent has been 

 Added, and because some of the tannin precipitates so produced 

 are rapidly decomposed. Some tannins (rheo-tannic acid) are not 

 precipitated at all by tartar emetic. 



V. Ammoniacal Solution of Acetate of Zinc. This reagent 

 should, according to Terreil, 1 Carpene", 2 and Barbieri, 3 be used 

 in the following manner for the estimation of tannic acid. The 

 liquid to be precipitated is brought to the boiling-point, an excess 

 of the zinc solution added, and, after concentration by evapora- 

 tion, the mixture is cooled and filtered. The precipitate is then 

 dissolved in sulphuric acid, and the tannin estimated by titration 

 with permanganate of potassium. I must admit that some tannic 

 acids may be determined in this manner, but I must also draw 

 attention to the fact that all tannins occurring in vegetable 

 substances do not exercise the same influence on permanganate of 

 potassium ; that is, one tannic acid may differ from another in 

 the amount of permanganate a given quantity can decolourize, 

 and this value of the tannin in terms of permanganate must in 

 many cases be first determined. It is partly on this account that 

 the estimations of the tannin in wine, made according to this 

 method, are of but little value. 



VI. Ferric Acetate, in conjunction with acetate of soda, has been 

 used by Handtke 4 for the estimation of tannic acid in oak-bark, 

 valona, divi-divi, sumach and catechu. He found the reagent un- 

 suited for the precipitation of the tannin present in Rheum, various 

 species of Filex, coffee and other plants ; and even with the first- 

 named substances it was only when the concentration was such 

 that the precipitate contained 45 -8 per cent, of oxide of iron that 

 the estimation yielded satisfactory results. 



Still less feasible is Wildenstein's 5 colorimetric examination, 

 which is based upon the intensity of the colour produced by the 

 solution on paper impregnated with ferric citrate. 



VII. Titration with Permanganate of Potassium.. Monier, 6 Cech, 7 



1 Zeitschr. f. anal. Chem. xiii. 243 (1874). 



2 Ibid. xv. 112 (1876). 



3 Ibid. xvi. 123 (1877). See also Kathreiner, ibid, xviii. 113 (1879). 



4 Jovirn. f. pr. Chem. Ixxxi. 345. 



5 Zeitschr. f. anal. Chemie, ii. 137 (1863). 



6 Compt. rend. xlvi. 447. 7 Loc. tit. 



