Dekker [Footnote: "De Looistoffen," vol. ii, p. 30 (1908).] was,
however, unable to detect diphenylmethane on distilling with zinc dust,
and did, therefore, not accept Nierenstein's views. In proposing the
formula--

O
||
HO ^ _ __C
| | |
| | }O
| | | __OH
| |____|_C_/ \OH
HO V \__/
OH OH OH

Dekker [Footnote: _Ber._, 1906, 34, 2497.] was enabled to account for
most of the details in the behaviour of tannin, viz.: (1) the
empirical constitution, C_14H_10O_9; (2) the almost complete
hydrolysis into gallic acid (the dotted line indicates the
decomposition of the molecule into 2 molecules gallic acid by taking
up water); (3) the formation of diphenylmethane as a result of
distillation with zinc dust; and (4) the electrical
non-conductivity. Since tannin on acetylating yields a considerable
amount of triacetylgallic acid, it should, according to Dekker,
contain at least six acetylisable hydroxyls.

Nierenstein [Footnote: _Chem. Ztg._, 1906, 31, 880.] objected to this
formula on account of its containing seven hydroxyl groups, whereas
Dekker found six, Nierenstein five, and Herzig still fewer hydroxyl
groups. The formula would also favour the conception of tinctorial