31 
solutions to sunlight. It is, however, very clear that the manner in 
which It changes depends very much on the conditions of the case. 
Thus, if chlorophyll is exposed to sunlight dissolved in bisulphide of 
carbon, a reddish-colored product is formed, and, though this differs 
very greatly from the red pigment met with in many autumnal leaves, 
It seems probable that under some conditions the chlorophyll in 
leaves is changed by the action of light into a red substance. By 
taking green sorrel leaves and keeping them somewhat fresh by stick- 
ing the stalks into moist ground, I found that those exposed to the 
sun with the under side upwards turned to a bright red, whereas 
those kept in the shade did not develop any fine coloring. We may 
often see that partially broken leaves or twigs undergo this change 
when all other parts of the tree remain green, and this and various 
other facts lead me to conclude that the change of chlorophyll into a 
red product depends on a certain amount of reduced vitality as well 
as on little-understood conditions varying in different kinds of plants, 
inough I fully admit that there are some facts not easy to under- 
stand, yet on the whole it seems to me that these principles fairly well 
explain why certain leaves turn red in autumn. Slight frosts reduce 
their vitality in such a manner that the chlorophyll is changed by the 
action of the light into a red product. Thus, according to the char- 
acter of the season and the nature of the plants, the first effect of the 
reduced vitality in the leaves is that the chlorophyll is removed so as 
to show their normal yellow color, or is changed into a red pigment, 
or IS altered into a comparitively stable dull brown-green product. 
Ihese are the three extreme changes, but in many cases intermediate 
niixed results give rise to such less perfect and well-marked tints as 
^^^^y yellows and reds. 
The next series of changes is best studied in the case of those 
'eaves which in the first instance turn to a bright yellow, and it 
appears to me that they depend mainly, if not entirely, on the pro- 
^^^ction of deeply-colored pigments by the oxidization of tannic acid 
and other more or less colorless substances. The difference in the 
j;esulting tint seems to depend on the nature these substances. Thus, 
J^^r exemple, the tannic acid in the yellow oak leaves changes into a 
Drown substance, whereas the quinotannic acid in yellow beech 
eaves changes into the fine orange-brown color which makes those 
rees so ornamental in autumn. On the contrary, the bright yellow 
poplar leaves rapidly pass to a dark dirty brown by the alteration of 
another constituent. Other kinds of leaves give rise to tints of an 
JJterinediate and less well marked character. In many cases it is 
th ^k '^^P'^ssible to draw the line between the color of this stage in 
l»e change and the final dark and dirty bro%vns of dead and decaying 
aves. For fine effect very much depends upon the production of 
ach special tint in a fairly pure state, so as to show bright yellows, 
^s and browns. This seems to be influenced by the character of 
j^^ weather. It is also, of coarse, important that the half dead 
aves should hang long on the trees, so as to develop their full 
loring before being blown off by the wind, 
all ti ^ 1 -^ ^'^^^^ ^^^ ^^^^ f^^^s into consideration, it appears clear that 
^'le bright and beautiful tints of autumn are merely the earliest 
