194 
Proceedings of the Royal Society of Edinburgh. [Sess. 
thermometers were changed from the one vessel to the other at each test, so 
as to check any difference in them. 
Sulphur Oxides. 
The impurities in polluted air which first called for attention were the 
sulphur oxides, as previous investigations have shown them to be powerful 
producers of nuclei of both kinds. These gases are thrown in abundance 
into the air of polluted areas, being the products of the combustion of the 
sulphur in coal. The first of these oxides tested was sulphurous oxide, S0 2 , 
and sulphurous acid, H 2 S0 3 . In these experiments both the fumes of 
burning sulphur and the solution of it in water were used. As both these 
forms of the compound gave very much the same reactions, the latter was 
for convenience most generally used, and we shall here consider their action 
to be alike, unless when specially referred to, and we will for simplicity 
use the symbol S0 2 for both. 
One generally associates sulphurous oxide with fumes and fogs, yet I 
find that S0 2 , so long as it is in good company, — that is, in pure air — shows 
no tendency to produce any kind of nuclei. It can be kept in pure air and 
water vapour for a long time without producing a single nucleus of con- 
densation. But unfortunately this gas has a great tendency to form, from 
our point of view, undesirable associations. It seems to be ever on the out- 
look for something with which it may combine, or ready to take the 
smallest hint from outside influences to change and combine with its 
previously unattractive neighbour, when it falls from its high estate of a 
free-moving gaseous molecule to the condition of a solid or liquid particle 
confined to brownian movements, and probably ends its independent 
existence in a fog particle, or possibly in a raindrop, after which all 
independent existence ceases. 
The study of S0 2 as a nucleus-producer is an extremely interesting one. 
It is so unstable that there are many gases which make nuclei with it, and 
also many outside influences which change it from the gaseous condition 
to a solid or liquid particle. In experimenting with S0 2 a little of the 
ordinary commercial acid mixed with a great deal of water is put in the 
flask S (fig. 1) to keep up a supply of the gaseous acid. In the flask E 
was placed a solution of any gas it was desired to add to the S0 2 in S. 
When it was required to test the effect of outside influences on the S0 2 the 
flask E was removed and the filter F connected direct with S. As already 
stated, we can have S0 2 along with filtered air drawn into the test-flask T 
and expanded without the S0 2 giving any nuclei of condensation, not a 
