DETECTION AND ESTIMATION OF TANNINS. 39 
thrown out and a more dilute solution of the tannin 
material taken. When precipitation has taken place 
satisfactorily, the mixture should be rapidly poured on 
two counterpoised filters placed one iuside the other, the 
precipitate washed with just sufficient warm water to re¬ 
move alum, dried at 110°, and weighed. For gallotannic 
acid 54 per cent, of the precipitate should be considered 
tannin. With the amount of alum above mentioned it 
is possible to add the gelatin solution in slight excess 
without dissolving any of the precipitate. 
According to Jolmnson 100 parts of gelatin precipi¬ 
tate 120 parts of dry tannin. 
Schulze recommended the use of ammonium chloride 
to cause coagulation of the precipitate; his solution 
contained ten grammes of gelatin in one litre saturated 
with ammonium chloride, and Lowenthal used a satu¬ 
rated solution of common salt with one-tenth its volume 
of hydrochloric acid (specific gravity 1.120) for the same 
purpose, but, if the proportion of alum (10 grammes 
to 1 litre) recommended above be used and the details 
carefully carried out, it will be found preferable. 
Powdered or rasped raw-hide was originally sug¬ 
gested by Davy and more recently used in the process 
of Hammer. This method consisted in preparing a 
solution of the tannin material, taking its specific 
gravity, then digesting it for a few hours with hide to 
four times the supposed amount of tannin present, 
which latter should be present to the extent of from 3 
to 5 per cent., filtering, and again taking the specific 
gravity. The loss in specific gravity on account of di¬ 
gesting with hide represented the tannin, a loss of .0040 
indicating one per cent. 
