242 Proceedings of Royal Society of Edinburgh. [sess. 
these two curves, or the difference between the minimum air tem- 
perature and the minimum ground temperature, shows the intensity 
of the earth’s radiation on the different nights. As the air is 
partly cooled by its passage over surfaces cooled by radiation, 
we might expect the fall of temperature from the previous day’s 
maximum would be proportional to the difference between the 
minimum temperature and the minimum temperature on the grass. 
This, however, will only he the case when all other things are 
equal, such as amount and duration of cloud, diathermancy of upper 
air, radiating power of lower air, and velocity of wind. 
Selecting the nights on which there was little wind, let us see if 
we can trace any effect due to dust. On the 30 th October there 
was very little dust, and it will he noticed there was a great fall in 
the temperature on the following night, though the wind does not 
seem to have fallen very low. On this date the air was the purest 
observed at this station, and on the following night the radiation 
was the greatest observed. The fall in temperature of the air, 
though considerable, was not the greatest recorded ; but we are not 
entitled to expect it would be, as the fall in temperature of the air 
depends on the duration as well as on the intensity of the radiation. 
No very satisfactory conclusion can be drawn from the other 
observations, as on the occasions when the amount of dust was low 
there seems to have generally been some wind, and this would 
check the fall of the thermometer on the grass. Further the dust 
observations were not taken at sufficiently short intervals for 
ascertaining whether the air at night had the same amount of dust 
as on the previous evening, or on the following morning. It may 
be as well to notice here, that on the 12th November there was a 
fair amount of dust, and yet the temperature fell considerably at 
night. The reason for this would seem to be, that though there 
was a good deal of dust, it was probably confined to the lower 
stratum of air, so that while the lower air was a good radiator 
from the amount of dust in it, it was unprotected above by dusty 
air. The reason for stating this is that on the following morning, 
tne 13th, when the amount of dust was great, there was a dense fog, 
but this fog was only low down, while immediately overhead the 
sky looked clear. 
The relation between the dust and the night temperature appeared 
