IN THE PRESENCE OF DUST-FREE AIR AND OTHER GASES. 
307 
absolute size of the water drops which would actually be in equilibrium in vapour of 
the various degrees of supersaturatioii. 
They furnish, however, a convenient means of expressing the relative efficiency of 
the nuclei in helping condensation. Thus, the nuclei producing rain-like condensation 
are equivalent in their effect on the condensation, to water drops of radii between 
6'4 X 10“® and 8'7 X 10~® centim. There are, as we have seen, certainly not more 
than 100 of these in each cubic centimetre of moist air, and they are absent in 
hydrogen. The nuclei equivalent to water drops whose radii lie between the narrow 
limits 5‘9 X 10“® and 6‘4 X 10centim. amount to many millions per cubic 
centimetre. 
It is difficult to account for the immense number of these latter nuclei, otherwise 
than on the view that they actually are simply small aggregates of water molecules, 
such as may come into existence momentarily through encounters of the molecules. 
On this view the dimensions of the molecules cannot be small compared with 
6 X 10~® centim. Bards* states that if it were possible to measure the super¬ 
saturation required to make steam condense in the absence of dust, the dimensions 
of the molecules could be calculated with the aid of Lord Kelvin’s formula. 
Probably he takes some such view as that here suggested. 
The nuclei which bring about the rain-like condensation, and the greater number 
of which appear to be equivalent in their power of causing condensation to water 
drops of not much less than 8‘7 X 10centim. are probably of a different character. 
As, however, I am continuing these experiments, it would be premature at the 
present stage to discuss the various views that might be held as to their nature. 
[Note added Jidy 22, 1897. 
Further Experiments on the Action of the X-rays .—I have lately repeated the 
experiments on air, using an expansion apparatus provided with a window of very 
thin aluminium, so arranged that the whole of the contents of the tube were exposed 
to the rays of a suitably placed Rbntgen lamp. 
This gave results identical with those already obtained, no condensation resulting 
when the air was expanded while exposed to the rays unless exceeded 1’25, 
while with expansions even slightly exceeding this, a comparatively dense fog 
resulted ; only a few scattered drops appearing with similar expansions in the 
absence of the rays. 
As was to be expected, much denser fogs were obtained under the action of the 
rays with this apparatus than with that formerly used. 
A glass plate of 7 millims. thick, placed over the window, appeared to cut off the 
effect of the rays completely.] 
* ‘Phil. Mag.,’ vol. 38, p. 34 (1894); also “Report on the Condensation of Atraospheinc Moisture.” 
U. S. Department of AgTiculture, Weather Bureau, 1895. 
2 R 2 
