228 
shows that all these different conditions may occur in different 
parts of the same healthy leaf. 
As I have already said, the red colour of by far the greater 
number of leaves is due to two substances belonging to 
Group B, which are related to one another in a very simple 
manner. One of these may be obtained from the red skin of 
the leaf-stalks of the common rhubarb, from the very young 
leaves of the red beech, or from the flowers of Calceolaria and 
some species of Dianthus. It is im an unoxidized state, but 
when dissolved in sulphuric acid diluted with an equal bulk 
of water, and carefully oxidized by means of a little solution 
of hypochlorite of soda, the absorption seen in the spectrum 
is lowered from about the red end of the green to the red end 
of the yellow, whilst the colour is changed from orange to 
pink. This alteration is permanent, and is not reversed by 
the addition of deoxidizing reagents. 
The other kind of erythrophyll is met with in so very many 
leaves that it may be looked upon as the characteristic 
substance of the red tints of autumn. This is naturally in 
an oxidized state, and its spectrum exactly corresponds with 
that of the other kind which has been oxidized artificially. 
Both are completely decolorized by the addition of an excess 
of the hypochlorite, unless, indeed, some member of the 
chrysotannin group be present, which may give rise to a 
more or less deep yellow or orange. 
Now, I find that when the unoxidized modification is kept 
in a dry state, exposed to the air, it slowly, but gradually, 
passes into the oxidized, and a similar change takes place natu- 
rally in leaves. Thus, for example, in early spring, the very 
young shoots of the common bramble contain the unoxidized 
colour, but later in the year it is replaced by the oxidized. 
This change occurs at a very early period in some plants, so 
that, as in the case of the bilberry, the very young leaves, 
and also the flowers, appear to contain a mixture of the two 
modifications, and the very youngest red leaves of the haw- 
thorn and of the sycamore are coloured by the oxidized kind 
alone. It is, therefore, apparent that a certain amount of 
oxidizing action may occur in young leaves; but still it is 
quite limited in its amount. Probably, in some cases, it 
extends so far as to destroy the erythrophyll altogether, but 
does not produce any important effect on the colours of any 
other group so long as the leaves are healthy. Though 
much evidently depends on the species or variety of the 
plant, yet, on the whole, it seems to me that we might easily 
explain the partial or entire disappearance of the erythro- 
phyll from the red leaves of early spring, and its reappear- 
