1888 - 89 .] Mr John Aitken on Dust Particles. 
151 
wise more than 100 c.c. of air will he taken in, and an unknown 
quantity will pass out into the pump. 
But suppose that the particles are still too wide apart, or that we 
wish to check our results, and we desire to get a greater quantity of 
dusty air into the receiver, then all that is necessary, after making the 
expansion and counting the drops, is to open the receiver stopcock, 
as before, to allow the air to enter, and then work the pump for some 
time, say making fifteen or twenty strokes ; in this way we can clear 
the purified air out, and fill the receiver entirely with which we wish to 
test. When working in this way, the numbers counted on the stage re- 
quire to be multiplied by 1 *43, to allow for the effect of the expansion. 
We have seen, then, how we can, by means of the small cylinder, 
measure any quantity of the air to he tested from c.c. to 10 c.c.; 
and also how, by means of the air-pump, we can measure 100 c.c., or 
cause the receiver to he filled entirely with the dusty air. There is 
evidently, however, too great a step between the largest quantity 
measured by the small cylinder measure and the contents of the 
large pump — that is, from 10 c.c. up to 100 c.c. To bridge over 
this interval, the air-pump has a scale attached to the piston-rod, 
and sliding with it. On this scale is marked, not the quantity 
displaced by the piston, hut the quantity taken out of the receiver 
by the pump, when the receiver stopcock is closed. When provided 
with this scale, we need not always take in 100 c.c. ; hut, by making 
a partial stroke, we may take in, say 50 c.c. of dusty air, or any other 
quantity we may desire. In this way a perfectly graduated scale is 
obtained, and we can send into the receiver any measured quantities 
of dusty air we may desire from ^ c.c. to 100 c.c., or even fill the 
receiver full of the air to be tested. 
It may be asked, Why not graduate the pump to actual displace- 
ment, and open the receiver stopcock before making the stroke? 
The objection to this plan is, that some of the dusty air would 
escape with the pure air while the stroke was being made, and we 
should not have any means of knowing how much was lost. 
Another plan for measuring large quantities of air with the 
apparatus described has occasionally been used for check experi- 
ments, but it is more troublesome to work. In working by this 
plan, the pipe T, for supplying water to the receiver, is connected 
with a graduated vessel full of water. By the use of the air-pump 
