DUST, FOGS, AND CLOUDS. 357 
Tf all nuclei are absent, water may be cooled below the “freezing-point” or 
heated above the “ boiling-point ” without any change taking place ; but there 
seems to be a limit to the amount it may be cooled or heated under these con- 
ditions without the water freezing or boiling. However carefully we may make 
the experiments after the water has been cooled to a certain amount, it always 
freezes without the presence of a free surface, and it also boils without the 
presence of a free surface when heated much above its “boiling-point.” In 
these cases there always, however, appears to be some want of continuity or 
uniformity produced by the presence of some substance which exercises an in- 
fluence on the water, and determines a weak point at which the change begins, 
and when once begun progress is of course rapid. In water we can easily under- 
stand how the sides of the vessel and the surfaces of foreign matter, &c., will 
form weak points, from which “free surfaces” are developed, extending into 
the mass of the liquid ; but it is much more difficult to understand how weak 
points can be formed in gases, and even when started they have no power of 
propagating themselves. These considerations would seem to suggest that the 
rainy condensation in filtered air may be produced by some form of nuclei 
which passes the cotton-wool filter, and which are perhaps very small, and do 
not become active as nuclei till a considerable degree of supersaturation is 
attained. 
There are, however, certain considerations which show that if the degree of 
supersaturation is sufficiently great, then condensation will probably take place 
without nuclei. Professor JaAmMEs THomson* has shown that the isothermal 
curves obtained by Dr Anprews from his experiments on carbonic acid at 
temperatures below the critical temperature of that substance may not be 
really so discontinuous as they appear, and that there may be a condition of 
that substance which would be represented by a continuation of the vapour 
part of the curve beyond the “boiling” or “condensing line.” To test this 
point Professor THomson suggested an experiment in which saturated steam, 
surrounded by a heated vessel, was to be expanded till it was cooled below 
its condensing point for its pressure, and the effect on the volume and pressure 
noted. This experiment, I believe, has never been made. We, however, see 
from the experiments described, that the theoretical extension of the curve 
discovered by Professor THomson has a real existence. This curve of Professor 
THomson’s shows that the degree of supersaturation possible has a perfectly 
definite limit, beyond which supersaturation is impossible. Further, if we 
examine these curves of Dr ANDREWS, which we may extend to water, they 
show us that it is only for temperatures below the critical temperature of the 
substance that supersaturation is possible. At temperatures above the critical 
* “ Proceedings of the Royal Society,” No. 130, 1871. 
