190 
Proceedings of the Royal Society of Edinburgh. [Sess. 
Apparatus for Testing Air. 
The apparatus used in this investigation was the same as that described 
in a previous communication to this Society,* and is here reproduced (fig. 1). 
It consists of the test-flask T, in which the air is tested for fine dust or 
nuclei of cloudy condensation. This flask is connected by a tube with the 
pump P. It is also connected by another tube with the flask S, which in 
turn is connected with E as shown, while E is connected with the filter F. 
In the test-flask T is a little water to moisten the air. The interior of 
this flask is illuminated by the incandescent-gas light L, placed in a dark 
lantern which has a vertical adjustable slit on the side next T. The light 
is concentrated on the test-flask T by means of the condenser C, which 
consists of a globular glass flask filled with water. By working the pump 
P air is drawn through the filter F into the flasks, and the pump is worked 
till all dusty air is taken out. When testing the air for nuclei the stopcock 
K is closed and a short stroke of the pump is made to expand the air in 
T, the inside of T being examined while the expansion is being made to 
see if any cloudy condensation is taking place. This examination can be 
most easily done if the test-flask be surrounded by dark surfaces and the 
room be as dark as convenient for attending to work. A magnifying lens 
at the level of the flask, and fixed pointing in a direction at about an angle 
of 140 degrees with the direction of the light, and focussed on the 
centre of the flask, will be found to be of assistance in some cases where 
the nuclei are few and small. If the filter has been packed with sufficient 
cotton- wool, or whatever is used in its place, and all the joints are tight, 
there will be no condensation in the test-flask when the stopcock is closed 
and a short stroke of the pump made to expand the air. If condensation 
persists after sufficient pumping has been done, then either the filter is 
defective and more cotton- wool must be added, or some of the joints are 
faulty and must be put right before a test can be made. 
As it is necessary in this investigation to test not only whether nuclei 
are produced or not, but also whether the nuclei have any affinity for water 
vapour — that is, whether they are only haze producers or are true fog 
nuclei — the usual method by expansion was used for testing the presence 
of all kinds of nuclei ; and the method of testing whether they have an 
affinity for water or not was to draw the air with the nuclei into the test- 
flask T as usual, but now the stopcock K is not closed nor is any expansion 
made with the pump. The nuclei are simply left in the damp air and 
allowed to remain a few minutes, and are then examined to see if they 
* Proc. Roy. Soc. Edin ., vol. xxx., Part IV., No. 3. 
