DETECTION AND ESTIMATION OF TANNINS. 33 
cator. This method failed to give satisfaction, on ac¬ 
count of the difficulty in determining the end of the 
reaction. Clark used a modification of this for the 
estimation of tannin in tea. His improvement con¬ 
sisted in adding an excess of the cinchonine sulphate, 
filtering, and estimating the excess of the alkaloid by 
titrating with potassio-mercuric iodide. A series of 
trials, by the author of this volume, in which the 
method was tried on gallotannic acid, failed to give 
concordant results. 
F. Musset suggested using an excess of standard so¬ 
lution of iodine, and, after twelve hours’ standing, 
titrating the free iodine with standard sodium thiosul¬ 
phate, which was added in excess and titrated with 
decinormal iodine solution and starch. 
2. The methods of precipitation with metallic salts 
are very numerous and comprise many modifications. 
It is desirable sometimes to obtain the value of a tan¬ 
nin by this means, therefore the following more impor¬ 
tant processes will be given. 
Pribram in 1866 detailed a method involving pre¬ 
cipitation by lead acetate, washing, drying, and weigh¬ 
ing the precipitate; but, as there are at least three 
salts of lead tannate, according as the lead acetate or 
tannin is in excess, and these salts are readily decom¬ 
posed by washing, it is evident that such a method can 
only give approximations. 
Allen originated a modification of this for assaying 
the tannin in tea, in which the latter in hot concen¬ 
trated solution was titrated by a volumetric solution of 
the lead acetate; the end of the reaction was ascertained 
by filtering a small portion of the liquid and testing with 
