896 Proceedings of Royal Society of Edinburgh. [sess. 
would have no value. There are, however, some reasons for 
supposing that solids are never reduced to such minute dimensions. 
The simplest method of showing this is by a study of the phe- 
nomena of cloudy condensation in dusty air. If we take the 
usual apparatus for showing cloudy condensation, a few experi- 
ments will show that the solid particles seem to have a lower 
limit to their size, and that they are never so small as to be 
capable of diffusing or not being separated hut by the action of 
gravitation. The apparatus required for illustrating these points 
is shown in the figure. F is a glass flask provided with an india- 
rubber stopper, the stopper having two apertures in it, in which 
are fitted two tubes. One of the tubes, C, is connected with an 
air-pump P, and the other, DD, shown in dotted lines, with a 
cotton- wool filter W ; a stop-cock, S, being introduced in the pipe 
DD between the flask and the filter, as shown. A little water 
is put in the flask F to moisten the air. Suppose, now, there is 
ordinary unfiltered air in the flask F. If we close the stop-cock 
S, and then make the smallest degree of expansion by pulling out 
the pump-handle a very short length, it will he noticed that cloudy 
condensation at once takes place, the very smallest expansion 
being sufficient to cause the dust nuclei to become centres of 
condensation. That is to say, there are particles in the air 
