as connected with the Theory of Substitutions. 351 
If a bulb which has been exposed to the sun be raised by a 
spirit lamp to such a temperature that its gaseous constituents 
are rapidly evolved, its extremity dipping beneath some of the 
solution in the bottle, after allowing a sufficient space of time for 
the disengaged chlorine to be re-dissolved, and the oxygen be 
turned out of the bulb, it will be found on keeping the arrange- 
ment in the dark, that oxygen will slowly disengage as before. 
Now, there is every reason to believe that any small amount 
of oxygen dissolved in the liquid would be expelled with the 
chlorine at a high temperature. We therefore have to infer that 
the chlorine after this treatment still retains the quality of causing 
_ the decomposition steadily to go forward. 
The oxygen which thus accumulates in the course of time in 
the dark, after an exposure to the sun, does not arise from any 
portion of that gas held in a state of temporary solution, nor from 
peroxide of hydrogen, nor from chlorous acid in the liquid, un- 
dergoing partial decomposition. From any of these states a high 
temperature would disengage it. 
VI. The evolution of gas is not of the nature of a fermentation ; 
for when it once sets in, the molecular motion is not propagated 
from particle to particle, but affects only those which were 
originally exposed to the rays. 
Let a bulb be filled with chlorine water which has been ex- 
posed to the sun, and in a second bulb place a quantity of the 
same liquid equal to about one third of its capacity. Fill up the 
remaining two thirds with chlorine water which has been made 
and kept in the dark; and after keeping both bulbs in obscurity 
for some days, measure the volumes of gas they contain. If the 
qualities of chlorine, which has been changed by exposure, were 
communicable, by contact or close proximity, from atom to atom, 
we might expect that both the bulbs would yield the same quan- 
tity of gas; but this is far from being the case, and in such an 
experiment I found that the bulb containing the mixture gave 
only one fourteenth of the gas which was found in the other. 
ViL The quantity of gas which thus collects in the dark, de- 
pends on the intensity of the original disturbance ; which in its 
turn depends on the time of exposure to the rays, to their intensity, 
and other such conditions. In other words, the rays are perfectly 
definite in their action,—a long exposure giving a larger amount 
of subsequent decomposition, and a short exposure a less amount. 
