208 Proceedings of the Koyal Society of Edinburgh. [Sess. 
stituents of coal — namely, carbon and hydrogen — and see if they have any 
nucleus-producing power. For these tests we will use the products of 
combustion of ordinary household gas, and see whether they have any of 
the objectionable properties of the sulphur products. In testing the products 
from the gas the same methods were used as previously ; but before 
proceeding a word of caution may be useful here, as to the method of 
collecting these gases. It is necessary that all parts of the apparatus that 
will get heated by the gases should be previously highly heated to 
thoroughly cleanse them ; because, if this is not done, there is a risk of 
something being given off from the heated surfaces which may invalidate 
the results. In these tests a chimney 2 feet high and 3 inches diameter 
was placed over the flame, the products being drawn from the top of the 
chimney and passed through a coil of metal pipe to cool them, and then 
passed to the testing apparatus. The products from the perfect and the 
imperfect combustion of the gas were tested — that is, from non-luminous 
and luminous flames. 
The products, as before, were put to two different tests. First, to find 
if they gave any true fog nuclei — that is, nuclei with the power of 
condensing water in unsaturated air. The method of testing this was to 
draw the products direct and without being filtered into the silica sunning- 
flask, where they were exposed to the sunshine. The gases were then 
drawn into the test-flask, in which there was some water ; but no expansion 
was made in this test, the gases being simply left to see if there were any 
nuclei in them having an affinity for water vapour and capable of causing 
condensation without supersaturation. When this was done it was found 
that neither the sunned products from perfect combustion nor those from 
imperfect combustion gave any nuclei of spontaneous condensation. As 
oxidising agents other than those formed during combustion were found 
to act powerfully on S0 2 , these products were also tested in the same way. 
A little peroxide of hydrogen was put in the sunning-flask to act along 
with light on the gases ; but even with the aid of the peroxide the sun 
produced no nuclei of spontaneous condensation. In this test no salt was 
added to the water in the test-flask to reduce the vapour tension ; the air 
would therefore be fully saturated, and probably in some places a little 
supersaturated. 
In tests of this kind it is sometimes difficult to say whether or no there 
is any spontaneous condensation, as sometimes the particles are extremely 
minute and individually invisible, even with the aid of a lens, and owing 
to the unavoidable reflections of the light in the flask it is sometimes 
difficult to say whether any reflected light comes from the particles, or 
