588 
PROFESSOR GRAHAM ON THE MOTION OF GASES. 
When the large bell-jar over a was not exhausted, the gas in the latter was of the 
atmospheric tension. With the barometer at 29*28 inches, and thermometer at 54°, 
the air was withdrawn from the globe a, in 388 seconds in one experiment, and in 
389 seconds in another. 
The pressure upon the air in a was then reduced to three-fourths of an atmosphere, 
by exhausting so that the gauge barometer stood at 7’32 inches from the bottom 
of the scale, which is one-fourth of the whole pressure of 29*28 inches. The globe a 
was thus occupied by air of the tension of three-fourths of an atmosphere, or 
21*96 inches. It was in this state connected with the vacuous receiver v of the air- 
pump, and the time required for the effusion of the constant volume of 65 cubic 
inches of air, measured in its rarefied state, between the marks b and c, observed. 
The effusion of this volume of air of three-fourths density was effected in two experi- 
ments in 389 and 392 seconds. 
Again, the air in a being made of 14*64 inches tension, or half an atmosphere, the 
constant volume was effused into a vacuum in 41 1 and 408 seconds. 
Lastly, with the air in a of 7'32 inches tension, or one-fourth of an atmosphere, 
the time of effusion was 438 and 439 seconds. The results therefore of the effusion 
of a constant volume are as follows : — 
Time of effusion. 
Air of 1 atmosphere 388*5 seconds ... 3. 
Air of 0*75 atmosphere .... 390*5 seconds . . . 1*0051. 
Air of 0*5 atmosphere .... 409*5 seconds . . . 1*0541. 
Air of 0*25 atmosphere .... 438*5 seconds . . . 1*1287. 
It thus appears that the effusion of air into a vacuum is very little affected by a 
moderate change of density ; air of 1 atmosphere and of 0*75 atmosphere passing in 
nearly the same time. The effect therefore of the ordinary changes of the barometer 
on the effusion of air must be small, if at all sensible. A retardation occurs in the 
effusion of air of diminished density, which amounts to an excess of 2 -jjoth of the time, 
on air of 0*75 tension ; of ^ffth on air of 0*5 tension, and ^th on air of 0*25 tension. 
Experiments were also made on the effusion of air of higher density than 1 atmo- 
sphere. The air was drawn of any required tension from 1 to 2 atmospheres from a 
strong globular vessel A (fig. 3. Plate XXXI II.), provided with a gauge barometer 
and mercury, by which the tension of the compressed air within it was observed. 
Before its admission into this vessel the air was previously condensed in another 
vessel D by a syringe, to a higher degree of density than was required in A, and the 
supply of compressed air, regulated by the adjustment of an intermediate stopcock, 
so as to keep the gauge of A at a constant elevation, which could easily be done 
within -^th of an inch. 
In experiments with compressed air, the latter was allowed to flow into the two- 
pint jar exhausted on the plate of the air-pump, and the time observed which the 
gauge barometer required to fall through its range from 28 to 2 inches. During the 
