160 
Proceedings of Royal Society of Edinburgh. [sess. 
of the previous day had greatly decreased, and the number now 
counted was 1650. The day continued fine, with transparent air 
and bright sun ; and when tested at 1 p.m. the number was only 500 
per c.c. After this hour the tests were unreliable, owing to change 
of wind bringing the smoke towards the place of observation. 
The night of the 10th was frosty ; but the next morning opened 
dull, and the air was thick. The number of particles was found to 
be 4600 per c.c. The day got stormy and duller, the thickness in 
the atmosphere increased a good deal, and when tested at 3 p.m. it 
was found to contain 9250 particles per c.c. The air got clearer 
towards night, and at 11 p.m. the number was reduced to less than 
one-half. This clearing continued, and next morning the air was 
quite clear, and the number of particles was down to 625 per c.c. 
From an examination of these Colmonell observations, it will 
be noticed that on these occasions there was a direct connection 
between the transparency of the air and the number of dust particles 
in it. When the air was clear and transparent, the number fell to 
500 or 600 per c.c., when it was thick the number rose to 5000 per 
c.c.; and when it was very thick and hazy, the number was nearly 
double that. It will also be seen from the table that rain had little 
or no effect on the occasions these tests were made. The 8th was a 
wet, stormy day, with thick air ; and the number was as high as on 
the 11th, when there was no rain, and the air not quite saturated. 
We cannot draw any very satisfactory conclusions from these 
Colmonell observations, as they are far too few for anything like a 
stable foundation. It was, however, distinctly observed while the 
tests were being made, that there was a direct relation between the 
numbers counted and the transparency of the air. Before making 
each succeeding test, it was found possible to say whether .the number 
would be greater or less than the preceding, by simply observing 
the clearness of the distant hills A It would, however, be rash to 
* Though there was a close relation between the transparency of the air and 
the number of dust particles in it, during these Colmonell observations, yet it 
is evident that we will not be entitled to expect this relation to hold good 
under all conditions. The amount of vapour in the atmosphere, or rather the 
degree of saturation of the air, will have an effect on the size of the particles, 
more especially when the air is nearly saturated. We are, therefore, only 
entitled to expect this relation to hold good while the degree of saturation 
remains constant, which it probably did during these observations, the air being 
nearly saturated during most of the time. 
