1888-89.] Mr John Aitken on Dust Particles. 
149 
The Air-Pump. 
Turning now to the air-pump used for producing an expansion in 
the test-receiver, and shown at P, Plate I. The one used has a 
capacity of 150 c.c. There is no necessity for the pump being 
exactly of that size, hut its capacity should bear a certain proportion 
to the air- contents of the receiver, and be such as to make the neces- 
sary calculations as simple as possible. It may be thought that the 
pump is rather large, and that we might with advantage reduce the 
size of both pump and receiver. But it must be remembered that if 
we reduce them, then the measuring apparatus must be reduced also ; 
and as the measure is reduced to as small dimensions as is thought 
advisable, there would require to be some good reason for change in 
this direction. No doubt, with a pump large compared with the 
receiver, the degree of expansion is more trying on the joints and stop- 
cocks than if a smaller pump were used ; and there is also a tendency 
to spontaneous condensation when the expansion produced by a large 
pump is employed. Yet, in spite of these disadvantages, it is 
thought desirable to use a large pump, as the high expansion gives 
a great' fall in temperature, and thus causes the raindrops which 
descend on the stage to be larger and more easily counted than if a 
smaller pump were used. 
With the pump of a capacity of 150 c.c., and the receiver a 
capacity of 350 c.c., the calculations are easily made. Whatever 
number is counted on the stage per square mm. requires to be 
multiplied by 100 to get the number per c.c. in the air of the 
receiver, and this multiplied by 500 — that is, by the contents of the 
receiver and pump — gives the total number in the measured quantity 
of air sent into the apparatus. Suppose, for instance, that 1 c.c. of 
dusty air was measured and sent into the receiver, and if we counted 
2 drops per square mm., then as there was 1 cm. of air above 
the stage, there would be 200 drops per c.c. of the air in the 
receiver ; but as the original 1 c.c. of air was mixed with 350 
c.c. of pure air, and expanded to 500 c.c., there will have been 
200 x 500 = 100,000 dust particles in the original 1 c.c. 
If we had mixed only T X o c.c. of the dusty air, and got the same 
number of drops, then the number of particles would have been ten 
times greater, or 1,000,000 ; but if we had used 10 c.c. and got the 
