THE FUNCTION OF TANNIN. 
9 
young conn shows increase in the quantity of tannin, though 
not to any marked extent. Its distribution is unchanged. 
In the conn of Crocus the tannin is not, therefore, like 
starch, transferred from the part whose function is ended to 
that where its utility can be continued. 
Similar evidence is given by Arum maculatum , and the 
bulb of Narcissus Pseudo-narcissus. 
jEsculus Hippocastanum. April 13tli.—Shows abundant 
tannin in the cortex of twigs in scattered cells. All cells of 
phellogen, and most of plielloderm full; a few of the bast ray 
cells full, as are many cells of the soft bast; a few cells of 
the wood rays; scattered cells in the pitli-crown full, and 
some cells in central pith. In the leaf-traces of previous 
year’s leaves, the whole of the parenchyma of the vascular 
bundle is full. The cells underlying the cork layer on the 
scar are full; none in the cork layer itself; but some of the 
cells of the parenchyma of the leaf-stalk, which still adhere 
outside the scar, contain tannin. 
On this and the succeeding plants the method of investi¬ 
gation adopted was to enclose branches in dark frames, with 
provision for ventilation, and thus leave them to grow, but 
not to assimilate. 
The experiments closed on July 21st. The tannin had 
undergone no diminution whatever in the stem of the last year, 
although the starch was quite exhausted. This year’s stem, 
however, showed an almost complete absence of starch, but 
tannin perhaps more than in the normally grown stem of the 
year. The tissues of the bud and pith are especially full of it. 
On Pavia rubra (smooth-fruited red-flowered horse chest¬ 
nut), 1 experimented also with flowering axes. The flowering 
and fruiting proceeded apparently almost normally ; but when 
the fruits were approaching half-an-inch in diameter, most 
indeed much sooner, they fell off. They, and all the tissues 
of the peduncle, this year’s and last year’s stems showed 
tannin, but no diminution in the last year’s stem. 
Similar evidence is given by Quercus pedunculata, of 
which, however, I could not quite exhaust the starch, 
Pyrus Aucuparia, and Syringa vulgaris. 
I investigated also Ilex aquifolium and Phododendron as 
types of the evergreen, but did not succeed in exhausting the 
supplies of starch, notwithstanding the free flowering and 
subsequent active growth of the latter species. 
Pinus sylvestris (April 13tli), in which plant Schell had 
seen the diminution of tannin in the spring, I carefully 
investigated. Last year’s stem contains pretty abundant 
tannin in the outer cortex, and the cells around the resin 
