137 
1888 - 89 .] Mr John Aitken on Dust Particles. 
is a general outline of the apparatus and the method of working it ; 
hut, before going further, it will he better to describe in detail the 
different parts. 
Test -Receiver. 
In the apparatus described in the previous communication, the 
test-receiver was made of an ordinary flat-bottomed flask. When 
working with this apparatus it was noticed that the raindrops did 
not always fall vertically on the counting stage, hut from time to 
time they were seen to fall obliquely. Through the magnifying 
glass, it looked as if a miniature storm was raging inside the 
receiver, and driving the raindrops before it. When this happened 
the reading had to he rejected, as no reliance could be placed 
on the number obtained under these conditions. To investigate 
the cause of these irregular currents, the interior of the receiver 
was illuminated by a strong light concentrated in it by means of 
the water lens, and the movements of the air at the different points, 
when expansion was made, was examined by means of a lens, the 
condensed particles indicating the direction of the currents. 
Examination showed that when expansion was made, a vertical 
circulation of the air in the flask took place — the air next the 
sides of the flask forming the ascending current, while the descend- 
ing one occupied the centre. The cause of these currents is evident. 
When expansion is made the air is cooled, hut the air in contact 
with the sides of the flask rapidly absorbs heat from the glass, and 
an upward rush of air takes place all round the sides. If the flask 
has the same temperature all round, then these ascending currents 
meet near the top, curve over, and descend in the centre of the flask, 
leaving the part over the stage almost unaffected ; hut if one side 
of the flask is hotter than the other, then the current from that side 
is stronger than the current opposed to it at the top, and over- 
powers it, driving the air across the stage in a horizontal direction. 
The cause of the disturbance having been found out, attempts 
were made to remedy it, by placing screens round the stage to pre- 
vent the currents passing across it. The result of this arrangement 
was however unsatisfactory, because these screens, though they were 
less than a centimetre in height, gave rise to currents, by heating 
the air in contact with them ; so that, while they checked the larger 
currents, they caused small ones, which, owing to their nearness to the 
