DUST, FOGS, AND CLOUDS. 349 
the platinum wire being previously highly heated to cleanse it as much as 
possible. The receivers being closed, and the gas not lit, air was drawn 
through the apparatus till the air in the receivers was purified ; and no cloudy 
condensation took place on admitting steam. Contact with the battery was now 
made, and the gas lit. At once a densely fogging atmosphere was produced. 
No doubt part of this fogging was due to nuclei driven off the heated 
platinum wire, but as the wire was previously cleansed, and only heated for a 
short time, and quickly removed from the flame, there would be but little due 
to this cause, and what dust it did give off would be so fine that the heat of 
the flame would not be likely to break it up any further, and it would be 
gradually removed by the circulation, and its place filled with filtered air. It 
was, however, found that though the supply of air was kept up, and the flame 
kept burning for some time, the fogging showed no signs of decreasing. On 
shutting off the gas, the fogging at once began to diminish, and soon cleared 
away, showing that the fogging was due to the products of combustion. 
These experiments seem to indicate that the combustion of dustless gas 
and dustless air do of themselves give rise to condensation nuclei, and do not 
act by simply breaking up larger dust motes into smaller ones. These nuclei 
produced by the combustion of gas must be extremely small, as a very small 
flame so loads a considerable current of dustless air as to cause it to become 
full of a very fine and closely packed form of fog when mixed with steam. 
The question may here be put, Is it really dust which is driven off by the 
heat from the surface of glass, from the brass and iron wires, and from the 
other substances? It is extremely difficult to get a direct answer to this 
question, but I think that, reasoning from the known conditions necessary for 
the condensation of vapour, it is extremely probable that it really is an ex- 
tremely fine form of solid matter which is produced under these circumstances. 
Further, they have all been put to the test of the cotton-wool filter, and all of 
them have been filtered out and the air made non-cloud-producing. If it was 
some gas or vapour which was produced by the heat, we see no reason why the 
cotton-wool should have kept them so completely back. 
Another set of experiments was now made to test the fog-producing power 
of air and gases from different sources. The air to be tested was introduced 
into the experimental receiver, and steam blown in and mixed with it. Its fog- 
producing power was tested by the density of the cloudiness produced, and also 
by the time the fogging took to settle. It was always found that the air of the 
laboratory when gas was burning gave a denser fog than the air outside, some- 
times two or three times as dense. The products of combustion from a BuNSsEN 
flame and from a smoky flame were compared. They were found to be about 
equally bad, and both much worse than the air in which they were burned. 
These products were collected by holding the open end of the receiver over the 
