368 
entirely different colour. The character of this change agrees 
with what occurs when the colour of dark grapes is oxidized, 
as already described ; and both are good examples of what 
appears to be a general law. If not already oxidized, a 
certain amount of oxidization causes the absorption to ad- 
vance towards the red end of the spectrum to an extent 
varying according to the position of the original absorption, 
and then a still further oxidization causes the absorption to 
recede from the red end to considerably beyond the original 
position, sometimes so much so that all colour is lost. We 
might thus say that such substances may occur in an unoxi- 
dized, in an oxidized, and in a per-oxidized condition. These 
and similar facts seem also to show that there is an intimate 
relation between the chemical changes and the particular 
rays of light absorbed by such substances ; but the discussion 
of this very interesting question would lead me far away from 
the more immediate subject of this paper. 
On comparing the colour of sherry wine with that pro- 
duced by the action of the air on a solution in dilute alcohol 
of the substance in faded beech leaves soluble in water, I was 
unable to discover any difference, and hence I think we may 
conclude that it is formed by the oxidization of the yellow 
colour of the grapes ; but in some wines it is imperfectly 
oxidized, and they turn darker when exposed to the air. 
Adulterations of Wine. 
The only cases in which the spectrum method can be 
easily applied to detect adulterations in wines are when some 
colouring matter has been introduced to give a false appear- 
ance of age, or to bring the colour up to some desired 
standard. According to Payen’s interesting work , 1 logwood 
and Brazil wood were at all events employed a few years ago 
for this purpose ; and rhatany root and the berries of the 
so-called Virginian Poke ( Phytolacca decandra ) are some- 
times used. I mention these because I purpose to briefly 
describe how they may be detected, as 'an illustration of the 
methods which may be employed in such investigations. I 
must, however, confess that I have not detected these 
substances in any wines of commerce that I have yet ex- 
amined ; but then I have had very little opportunity for 
studying those likely to contain them. 
In order to detect logwood or Brazil wood, a small quantity 
of the wine should be agitated in a test-tube with an equal 
volume of ether, which rises to the surface in an almost 
colourless state if the wine be pure, but is tinged with a more 
1 ‘ Precis des Substances Alimentaires,’ note at p. 455. 
