Sunlight upon Aquatic Plants. 391 
Similarly living plants remain a normal green in bright 
diffuse daylight, whereas dead ones are bleached in a couple 
of days in the presence of oxygen, although if placed in 
well-boiled water, covered by a layer of oil, they retain their 
green colour for an almost indefinite length of time. Unless 
we admit that the decomposition and the reformation of 
chlorophyll proceed simultaneously when the living chloro- 
plastid is exposed to light, the above facts can only be 
explained by assuming that in the living chloroplastid the 
chlorophyll is held in some such manner as to render it 
resistant to the action of light, and that it is only when 
released from this vital combination that it can be oxidized 
and decomposed. 
Chlorophyllous cells, however, suddenly killed by chloroform, 
and exposed to intense light as soon as the chloroform 
has evaporated, take from 2 — 3 times as long to bleach 
as living ones do. In such dead cells the plasma is slightly 
more opaque, but on the other hand the slightly bluish 
green dead chloroplastids are distributed evenly through- 
out the cell, and are hence exposed to a slightly greater 
average intensity of light on this account. Living chloro- 
plastids may at first evolve oxygen, and hence help to bleach 
themselves, but this certainly ceases long before the bleaching 
is complete. If a closed-cell-preparation of a leaf of Elodea , 
which has been kept in darkness for some time to remove 
all free oxygen, is suddenly exposed to cooled and io 
times concentrated sunlight, the leaf-cells are killed without 
any perceptible bleaching of the chlorophyll having taken 
place, and hence it is evident that in such intense light 
C0 2 -assimilation ceases almost immediately, if any is ever 
possible. Similarly if a leaf of Elodea is passed through 
chloroform, a condition of anaesthesia may be induced which 
lasts for five or ten minutes, or even longer. Rotation ceases, 
the chloroplastids distribute themselves regularly over the 
cell, the power of response to external stimuli is temporarily 
in abeyance, and from the results given later it is also certain 
that the power of C0 2 -assimilation is temporarily lost. Such 
D d 2 
