vi Supplement to ‘ Nature,” April 14, 1923 

follows. If there were no dust particles present the 
drops of water would have to be built up from aggre- 
gates of water molecules. Now the radius of a water 
molecule is 2x 1078 cm., therefore if a few molecules 
met together by chance they would only form a very 
small drop, which would be so small that it could not 
exist unless there was large supersaturation. For 
example, according to Table II. a drop having a radius of 
7:2 x 10-7cm. would require a supersaturation of 20 per 
cent., yet nearly 20,000 water molecules would be re- 
quired to form such a drop. If, however, there were 
dust particles present the molecules of water would be 
deposited on them, and the radii of the initial drops 
would be so large that little supersaturation would be 
required to maintain them. 
This explanation appeared to satisfy every one for a 
long time, but in recent years considerable doubt has 
begun to be expressed. Already, in 1885, Assmann 
had searched for the dust nuclei when cloud particles 
evaporated under the microscope, and had come to the 
conclusion that if any were present they must have a 
smaller radius than 2-5 x r0~* cm. 
In 1912, Wigand made the reverse experiment. He 
first took careful counts of the number of nuclei in the 
air, then he created large quantities of dust by beating 
carpets, blowing up large clouds of coal, coke, and 
ordinary dust by means of bellows. Although he made 
such large clouds of dust that it was extremely un- 
pleasant to work in them, he could not find any increase 
in the number of nuclei in his condensation apparatus. 
In rgr0, A. Wegener directed attention to another 
difficulty. The distance one can see through the 
atmosphere depends on the number of-dust particles 
present and their size. From measurements made at 
Ben Nevis it had been found, froma comparison between 
the transparency of the atmosphere and the number of 
nuclei measured in Aitken’s instrument, that the damper 
the air the less the number of nuclei necessary in order 
to see the same distance. The observations gave the 
following result : 
TABLE III. 

Depression of the | Number of Nuclei present in x ¢.c. with a Constant 


Wet Bulb. Distance of Vision. 
T-1-2-:2" C, I'25 x 10% 
2*2-3°9 I-71 
3°9-5'5 2:26 

This can only be explained if the size of the dust 
particles increases as the humidity increases, even when 
the humidity is still far below its saturation value. 
But this is not an effect which one would expect if the 
only function of the dust particles is to act as nuclei, 
for there would be no condensation on them until the 



air has reached its saturation point. At all humidities 
less than roo per cent. the dust particles would remain 
dry and therefore of constant size. 
From these and other observations meteorologists 
have been led to the opinion thatscondensation does not 
commence on dust particles, if dust is to be understood 
in the ordinary way, but on hygroscopic substances, 
and that Aitken’s instrument does not measure the 
number of dust particles present but the number of 
hygroscopic particles. ‘ 
A great deal of work has recently been done on this 
question, especially by Kohler in Norway. Working 
on a mountain in the extreme north of Norway, Kéhler 
analysed the water obtained from the large deposits of 
rime which formed on the surroundings of his observa- 
tory. Rime is frozen cloud particles, and in this way 
he was able to determine the chemical contents of the 
actual cloud particles before they had had time to 
become contaminated. He found that calcium chloride 
was always present, and concluded that sea-salt obtained 
from the spray of the sea forms the true nuclei of cloud 
condensation. His results indicate that when the drops 
are extremely small there is sufficient salt present to 
reduce the vapour pressure to the same extent as the 
small radius of the drop increases it, thus allowing 
condensation to take place. Kohler is tempted to 
generalise his results and to contend that sea-salt is the 
main foreign substance on which condensation takes 
place. It is, however, not necessary to go so far as this, 
for there are many other sources of hygroscopic sub- 
stances. Lenard and Ramsaner have shown that sun- 
light—probably only the ultra-violet part—acts on the 
oxygen, nitrogen, and water vapour of the atmosphere, 
producing very hygroscopic substances. 
Large quantities of material capable of becoming 
condensation nuclei are produced by all processes of 
combustion. Thus the household fires and factory 
chimneys of centres of industry produce vast quantities 
of nucleus-forming material, chief of which is sulphur- 
ous oxide, SO,. This, when illuminated by sunlight in 
the atmosphere, is a very hygroscopic substance cap- 
able of causing condensation in unsaturated air. It is 
estimated that in England something like 5000 tons of 
sulphur are burnt each day in coal fires, giving enough 
sulphur products to pollute the atmosphere from Land’s 
End to John o’ Groat’s. Other products of combustion 
are also hygroscopic; thus it is not surprising that 
air from large industrial centres contains enormous 
quantities of nuclei. 
It is not necessary for hygroscopic particles to be 
large in order for water to be deposited upon them. 
Their chemical affinity for water is sufficiently large to 
counterbalance the surface tension forces which cause 
small pure-water particles to evaporate unless there is 
Salhi a 
—_— —_" 

