46 SUBSTANCES SOLUBLE IN ALCOHOL. 



indicator. But the majority of those who have worked the process 

 have failed to obtain good results. Almost all of them have found 

 that the assumption that rosaniline would not colour the liquid until 

 all the tannin had been precipitated by the cinchonine, was true 

 of certain tannins only, and not of all. It has been shown that, 

 with some tannins, the appearance of a red tinge in the solution, 

 which is said to indicate the end of the reaction, may be noticed 

 long before all the tannic acid has been precipitated. In many 

 cases better results might be obtained with cinchonine if the 

 tannic acid were precipitated by an excess, the liquid filtered and 

 the excess of cinchonine in the filtrate determined by titration 

 with potassio-mercuric iodide. This method has been adopted 

 by Clark in estimating the tannic acid in tea. (See 65.) 



XI. Gelatine and Hide. The behaviour of gelatine and hide 

 to tannic acid is often made use of in the estimation of 

 tannin. The estimation may be made either by determining 

 the increase in weight of a piece of hide, previously freed from 

 substances soluble in water and petroleum spirit by digestion in 

 those solvents, when allowed to lie for some time in the solution 

 of tannin, or by ascertaining the specific gravity of the solution 

 before and after the absorption of the tannic acid, and calculating 

 the amount from the difference. Hammer 1 has constructed a 

 table for gallo-tannic acid, from which the amount of tannin can 

 be directly read off. If this method of estimation is to be adopted, 

 a similar table would have to be constructed for other important 

 tannic acids, showing the relation between the difference in specific 

 gravity and the amount of tannin present. 



XII. Gelatine : Gravimetric Process. Precipitation of the tannin 

 by gelatine, and calculation of the amount present from the weight 

 of the precipitate, has also been tried. But the disadvantages 

 which present themselves here are that these precipitates are 

 neither sufficiently insoluble nor constant enough in composition 

 to allow of their being made the basis of a gravimetric estimation ; 

 especially in washing the precipitate with pure water, considerable 

 quantities of tannic acid are removed. 



It is, therefore, most advantageous to apply the precipitation 



1 Journ. f. pract. Chem. clxxxi. 159. See also Lowe, Zeitschr. f. anal. 

 Chem. iv. 365 (1865), and Hallwachs and Cech (foe. cit.}. Davy has already 

 estimated tannic acid gravimetrically by employing hide (Chem. News, 1863* 

 p. 54, and Zeitschr. f. anal. Chem. ii. 419). 



