DETECTION AND ESTIMATION OF TANNINS. 37 
Terreil in 1874, by Carpene in 1876, and by Barbieri in 
1877: While the details of the different authors varied 
somewhat, the following covers most of the methods. 
The reagent was added to a boiling solution of the tan¬ 
nin, and the mixture concentrated to one-half, when it 
was cooled and filtered. The precipitate was dissolved 
in dilute sulphuric acid, and the tannin estimated by 
titration with potassium permanganate, This process 
was especially recommended for determining the tannin 
in wine, but Dragendorff found it of little value. 
Handtke used ferric acetate in the presence of sodium 
acetate, and found it available for the estimation of the* 
tannins of oak-bark, valonia, dividivi, sumac, and cate¬ 
chu, when the concentration was such that the precipi¬ 
tate yielded 45.8 per cent, of ferric oxide by ignition; 
but with a number of tannins it did not yield satisfactory 
results, and the process is apparently not used. 
3. Methods of estimating tannins by means of gela¬ 
tin, isinglass, and hide or hide powder have been sug¬ 
gested in great number, with as many or more modi¬ 
fications by others until the list is a formidable one. 
Apparently the first paper on the quantitative determi¬ 
nation of tannin was by George Biggin in 1799, who 
collected, washed, dried, and weighed the precipitate 
produced by gelatin on tannin. He did not attempt to 
arrive at the exact amount of tannin present, as he 
made no estimation of the value of the precipitate in 
tannin. It could, therefore, be used only to compare 
one material with another, and was available only to 
the tanner who wished to compare the value of a new 
substance with that of a Avell-known one he was using. 
Davy in 1803 extracted six parts of bark with 
4 
