233 
The affinity for oxygen of the various kinds of chrysophyll 
and tannic acid, and the rate at which they are changed by 
exposure to the air, vary in most cases with the intensity of 
the alkaline reaction of their solutions. They remain un- 
changed for a long time when acid, are very slowly altered 
when neutral, but rapidly when very alkaline ; and, therefore, 
it is possible that the amount of free acid of the natural solu- 
tion in leaves may sometimes have an important influence in 
causing the chrysotannin colours to remain for a long time 
unaltered, whilst, if this acid became neutralized or were 
destroyed, these oxidizing processes would commence and 
progress rapidly. ‘Though these principles must certainly be 
taken into account, yet alone they do not appear sufficient to 
explain why the chrysotannin substances are so permanent 
in living and so soon oxidized in dead leaves. 
On the whole, we may, I think, express the connection 
between the various tints of leaves at different seasons of the 
year, and the vital and chemical changes, in a very simple 
manner. Chlorophyll is formed when the condition of the 
leaves is such that, under the influence of light, the affinity 
of oxygen is not only resisted, but actually overcome—car- 
bonic acid decomposed, and oxygen evolved. Xanthophyll, 
erythrophyll, and the chrysotannin substances are formed 
when the vital and chemical forces are more nearly equal, so 
that certain chemical changes occur which may be imitated 
artificially, along with others depending on the presence of 
vitality and light. Ata later period, when the leaves have 
died, the affinity of oxygen meets with no resistance, and the 
changes are entirely of that kind which we can imitate 
artificially by acting on dead compounds without the instru- 
mentality of light. 
As I have already explained, the finer autumnal tints 
mainly depend on the complete disappearance of chlorophyll, 
so as to allow the bright colours of the xanthophyll and 
erythrophyll to be seen unimpaired. [If little or no erythro- 
phyll be present, we then have clear yellow foliage, but, if it 
had been developed in considerable quantity, we have bril- 
liant scarlet and red. The production of these striking and 
beautiful tints therefore depends on special conditions, subject 
to considerable variation in different countries and in different 
years. Apparently, the vital powers of the leaves must be 
more or less paralysed, but no active oxidizing processes set 
up, or else the red colour would be destroyed, and the yellow 
completely hid by the darker and duller phaiophyll. Early 
frosts, occurring whilst the daylight is still strong, and before 
wet weather has set in, appear to me to be just what would 
