244 Proceedings of Royal Society of Edinburgh. [sess. I 
passage of heat through it ; whereas, at night, an opposite action 
takes place. The chilling tends to deposit moisture on dust, and 
so impede terrestrial radiation. 
In connection with this subject, there is a point that has been 
observed for some years, with regard to the earth’s radiation, which, 
as is hears on our subject, we may refer to here. Tor some years 
I have taken the usual meteorological observations, and along with 
these there has been kept a record of the earth’s radiation at night. 
When taking the readings in the evening, it has been noticed that 
on the evenings when there were no clouds, that the radiation 
effect was strongest just about sunset. If it was observed shortly 
afterwards, the radiation was almost always weaker, even though 
the sky kept clear. This would seem to indicate that the air 
is more diathermanous to terrestrial radiation just about sunset 
than it is later. The explanation which I would now offer of this 
is, that at sunset the air is more diathermanous than it is later, on 
account of the dust in the air holding little moisture. After the 
sun has been set for some time the air chills, and the dust particles 
increase in size by the vapour deposited on them ; and this increase 
in size tends to check the free passage of the earth’s heat into 
space. It is evident we cannot here attribute the decreased diather- 
mancy, with the advance of the evening, to the action of the water 
in its gaseous form, as there is at that time no sudden increase in 
its amount in the air. It would rather appear to be due to change 
which then takes place in the condition of the moisture from vapour 
to liquid. I have said the air is more diathermanous just about 
sunset than it is later. Probably it is more diathermanous to the 
earth’s variations during the day than it is at sunset ; but observa- 
tions on this point are difficult to make, owing to the complications 
introduced by the sun’s heat radiated and reflected by our atmo- 
sphere. 
At first sight, we might think that the truth or falsehood of these 
conclusions might be easily determined. If dust increased the day 
temperature and decreased the fall at night, then the mean tempera- 
ture will be increased by a large amount of dust. Can it then be 
proved from general meteorology that this is the case? Do inhabited 
countries have a higher mean temperature than uninhabited ones, 
other things being equal ; or has the temperature of these islands 
