166 Proceedings of Royal Society of Edinburgh. [sess. 
5 p.m., had less dust in it than it had an hour previously. This 
decrease in the number was, however, caused by the two tests not 
being made at the same part of the room. The 4 o’clock test was 
made near the window, while the later one was made near the fire. 
The quantity of dust in different parts of the room near the floor 
varied greatly ; the apparatus showed this very clearly, and by its 
means it was possible to trace the circulation of the air in the room. 
For instance, at the beginning of the tests, while the apparatus was 
near the window, the quantity of dust in the air, at 4 feet from the 
floor, rose at once when the gas "was lighted, and became about as 
high as it was later on. This was due to the cold window producing 
a down current, and drawing the products from near the ceiling 
towards the apparatus. Again, near the fire, the number was 
frequently much below that entered in the table, owing to the 
supply of air to that part of the room being drawn from the lower, 
colder, and uncontaminated air of the apartment. 
One point which impressed me greatly in making these last tests 
was the short time required to change the air near the ceiling of 
this room, compared with the slow circulation which we have re- 
ferred to, near the ceiling of the Meeting-Room of the Royal Society. 
When the apparatus was tested before beginning the test last re- 
corded, the number of particles of the air in the room was counted, 
and the gas was lighted to see how quickly the products would diffuse 
through the air of the room. The gas was not allowed to burn long, 
but on beginning the test about an hour after it was extinguished, 
not a vestige of the products of combustion could be detected. This 
was different from what might have been expected, after the experi- 
ence in the Meeting-Room. It will be remembered that, on making 
the tests next day, the air near the ceiling was found still to contain 
a very large number of particles. Why this difference in the two 
rooms? A short time sufficed to clear all the products of com- 
bustion out of the air near the ceiling of the room here, while a 
whole night seemed to have reduced the number in the other room 
to only about one half. The windows in the room of the Royal 
Society do not go nearly up to the ceiling, whereas in the room 
here they rise to within a short distance of it. Again, some of the 
difference will probably be due to a difference in the porosity of the 
walls and ceilings of the two rooms. 
