120 
In some of the experiments which I made the papers in 
the leeward box became sensibly damp in consequence of 
the box being placed too near the fountain, or from the 
acting of strong gusts and eddies of wind. In these cases 
the coloration of the papers was very much retarded or 
altogether prevented, and it therefore appears that spray 
has, like fog or haze, the power of absorbing or decomposing 
atmospheric ozone. 
Having now proved, expeiimentally, that spray produced 
by mechanical means, and ordinary fog or haze produced by 
the condensation of aqueous vapour in the air, are precisely 
similar in their relations to atmospheric ozone ; and having 
also shown that the quantity of ozone usually found in the 
air could not have been held in solution by the water from 
which the aqueous vapour in the atmosphere is derived it 
becomes evident that the production of atmospheric ozone 
is in some way dependent upon the minute state of division 
in which water exists in the air in the visible form of 
clouds, fogs, and haze, and often also probably in an 
invisible form. This consideration has led me to infer 
that water exposed to the air has tlie power of condensing 
Oxygen upon its surface into a thin film of ozone. When, 
therefore, complete evaporation of the vesicles or globules 
of moisture which constitute a cloud or fog takes place the 
ozone is left free to diffuse itself through the air ; but when 
evaporation takes place from the surface of a large and 
practically inexhaustible mass of water the ozone is not set 
free but remains adhering to the surface It will, however, 
be objected that if ozone is formed in this way test papers 
ought to be coloured very rapidly in a fog or dense haze, 
and that no bleaching action could take place upon papers 
which had already heen ozonised. The explanation which 
I will venture to offer is that ozone associated with moisture 
and in the presence of the oxydised potassium will combine 
with the freed iodine and form iodic acid which, uniting 
