DETECTION AND ESTIMATION OF TANNINS. 35 
The precipitate with it, like that of the corresponding 
lead salt, is not of constant composition; therefore the 
precipitation must be made with moderately concen¬ 
trated solutions, and the resulting precipitate rapidly 
washed, dried, and weighed. Fleck precipitated the 
tannin by copper acetate, and washed the precipitate 
with ammonium carbonate to remove gallic acid. The 
precipitate was then washed, dried at 100°, weighed, 
and ignited. Wolff found 100 parts of cupric oxide 
represented 130.4 parts of tannin. 
Pavesi and Rotondi found 100 parts represented 145 
parts tannin. This difference among different investi¬ 
gators Schiff explained by supposing they used tannins 
of variable degrees of purity, and he found 136 parts 
of pure digallic acid to be represented by 100 parts of 
cupric oxide. 
J. Persoz recommended dissolving eight grammes of 
ammonium chloride in one litre of water, and in this 
dissolving some stannous chloride. Ten grammes of 
pure dry tannin were then dissolved in one litre of 
water, and 100 cc. of this solution precipitated with 
100 cc. of this stannous chloride solution, the mixture 
allowed to stand ten or twelve hours in a graduated 
cylinder, and the volume of the precipitate read. Then 
ten grammes of the tannin material were boiled for 
one-half hour with a half-litre of water, filtered, and 
the residue washed with sufficient water to make one 
litre; 100 cc. of this solution in a similarly graduated 
cylinder were then precipitated with 100 cc. of the stan¬ 
nous chloride solution, and after standing ten or twelve 
hours the volume of the precipitate was compared with 
that from the pure tannin. 
