THE FUNCTION OF TANNIN. 
34 
leaf it is present in apparently the same quantity and with 
the same distribution in the fallen leaf, e.g., in JEsculus 
Hippocastanum, Salisburia adiantifolia, Cataipa bignonioides. 
On the other hand, these investigations have led me to look 
upon the complete emptying of the leaf of its starch cell- 
contents as the surest sign of approaching fall. A few ever¬ 
green shrubs which I have examined from the same point of 
view show the same more or less complete removal of starch 
from the leaf in winter, while, on the other hand, tannin 
may be present in abundance, e.g., Ilex aquifolium, Iihodo- 
dendron ponticum. 1 have examined about forty species of 
deciduous or evergreen plants, finding no essential divergence 
from these results. 
I have made no observations upon the relations of tannin, 
with the ripening of fruits. Mack* * * § and Haasf state that in 
the ripening of the grape the quantity of tannin diminishes. 
BuignetJ had previously suggested this is various fruits. 
Observations with fruits must, however, be made with some 
caution. Artificial selection has tended naturally to the pro¬ 
duction, for edible purposes, of fruits in which the natural 
tannin contents are either diminished, or localised in certain 
special parts, such as skin, testa, &c., and this localisation 
may take place during the processes of ripening. In the 
second place, the same artificial selection may have established 
in cultivated fruits a tendency to transfer their tannin to the 
stem in ripening. Many fruits, however, ripen normally 
when severed from the parent plant, and if in such dis¬ 
sociated ripening analysis showed a reduction in the quantity 
of tannin contents, substantive evidence would undoubtedly 
be provided. 
In the absence of any reliable method of separating tannin 
from free glucose, and further of separating tannin from the 
glucose with which it is loosely combined, micro-chemical 
investigations are prosecuted under considerable difficulties. 
No one would seriously dispute that in the processes of 
growth the free glucose could be separated out; nor is it in 
any degree improbable that in exhaustion experiments the 
“ tannin ” may be more completely reduced to the “tannic 
acid” state §; the point actually at issue is the decomposition 
of tannin in such way that tannic acid itself disappears by 
* Mack, Bot. Jahresb., 1877, p. 716. 
f Haas, Chem. Centralb., 1878, p. 700. 
I Buignet, Ann. Cbim. Phys., 1861, iii. Ser., Bd. 61, p. 281. 
§ Compare in tliis connection tbe different appearance and aggrega¬ 
tion of tannin masses in comparatively empty cells, as noted above. 
