398. The red matter of these tri-coloured leaves is 
soluble in water, and is dissipated by light, leaving 
a yellow resinous matter, in which acid abounds *. 
It is also rendered green by alkalis t> an observation 
which suggests the probability that its redness de- 
pends, as in other cases, on an excess of acid. 
Hence, then, the green colour of certain parts of these 
leaves would seem to depend on an excess of alkali, 
the yellow on a smaller portion of that substance, 
and the red on the predominance of acid matter : 
all which colours are, in succession, exhibited by au- 
tumnal leaves, and by certain vegetable infusions, 
when submitted to the action of acid and alkaline 
matter. Of the truth of these opinions, an observa- 
tion of M. Senebier affords additional evidence ; for 
he remarked, that the tri-coloured leaves of the ama- 
ranth, exposed under water to the sun, afford air 
from their green parts alone, while the red and yel- 
low portions yield no air J ; so that, in the same leaf, 
we see the green colour accompany the decomposi- 
tion of carbonic acid, and the yellow and red colours 
continue, when this acid is retained. 
399. The general causes which thus give rise to 
the colour of leaves, seem, likewise, to act in the 
colouration of flowers. We have already noticed 
(25.) the agency of light in the production of these 
Colours, by observing that roses, if kept in perfect 
darkness, are altogether deprived of colour. Blue 
hyacinths, says M. Senebier, become grey in dark- 
Mem* t. iii, p. 94. 95. f Ibid. p. 94, I Ibid. p. 99. 
