DUST, FOGS, AND CLOUDS. 347 
by the fog which forms in the receiver. The effect seemed so great for so 
small a cause, that I repeated the experiment a great number of times, some- 
times putting in the wire and getting the fog, and sometimes going through all 
the motions and changes necessary for, but not putting it in, and getting no 
fog, that I am compelled to come to the conclusion, that the fogging is really 
caused by the contamination due to the touch. 
- A great number of different substances were tested in this apparatus, and, 
as might have been expected, all were active fog-producers. Amongst other 
substances tried were different salts. One point noticed was that their 
activity did not depend on their power of evaporating or subliming. Camphor, 
though subliming and evaporating quickly, scarcely ever gave any fog, only a 
heavy coarse-grained fog which settled at once, while ammonic carbonate, 
sodic carbonate, and sodic chloride were very active, indeed the latter salt is 
one of the most active substances I have tried. If we place a crystal of sodic 
chloride 745 grain in weight in the tube, and apply heat, it will continue to 
give off nuclei sufficient to form a dense fog for a long time, without apparently 
losing in size. 
We see from these experiments that when testing the fog-producing power 
of a flame, it will not do to collect the products of combustion and draw them 
into our experimental receiver, as the heat would raise a dust from the surface 
of the collecting tube sufficient to cause a dense fog; another method of 
experiment was therefore devised. It was, however, necessary before pro- 
ceeding further, to test the effect of the gas to be burned, to see if it was 
active as a fog-producer. Gas from the gas pipes was accordingly passed into 
the experimental receiver, and tested with steam, and found to be perfectly 
inactive. No cloudiness appeared. Any effect then produced by the burning 
gas could not be due to dust carried in by the gas. 
The apparatus was now arranged in the following manner to test the fog-pro- 
ducing powers of the products of combustion from a gas flame :—Two receivers 
were arranged alongside each other, and connected by means of a pipe. 
Gas was led into the first receiver by a pipe terminating a short distance 
inside the receiver in a glass tube, the end of which was drawn to a fine jet 
at which the gas was burned. The receiver used for this purpose was so large 
that the flame could not heat the glass sufficiently to make it active as a fog- 
producer. After the gas was lighted, a current of filtered air was drawn through 
the receiver to supply oxygen for the flame. The products of combustion were 
drawn into the second receiver through the connecting pipe. In this second 
receiver the products of combustion were tested from time to time with steam. 
At first, of course, the air which came would be unfiltered dusty air ; 
but as nothing but filtered air entered, this dusty air ought gradually to give 
place to pure air. It was found, however, that after filtered air had been 
