590 
PROFESSOR GRAHAM ON THE MOTION OF GASES. 
Here the time of effusion of air of 2 atmospheres falls about 22 per cent, below 
that of air of 1 atmosphere, while in the upper part of the scale the difference was 
only 5 per cent. 
[8. Effusion of Air of different Temperatures, by Plate F. 
This plate was a portion of thin platinum foil, with an aperture of an irregular 
hatchet form, of which the two greatest cross diameters were and g-^th of an 
inch. The perforated plate was attached to the end of the little brass cylinder by 
means of soft solder. A two-pint jar, giving a cavity of 72 54 cubic inches, was used 
as the aspirator-jar, and the time of the fall of the gauge barometer was observed 
from 28*5 to 23*5 inches, with the admission of dry air at different temperatures. 
1. The temperature of the room being 41°Fahr., and the height of the barometer 
29*616 inches, dry air entered the aspirator-jar in three experiments in 533, 532 and 
530 seconds, of which the mean is 53T66 seconds. The room being afterwards 
heated up to 52°, the time of effusion of an equal volume of air was found to be, in 
three experiments, 525, 52/ and 526 seconds, of which the mean is 526 seconds ; or, 
a rise of 1 1° in temperature has shortened the time of effusion by 5*66 seconds. 
Taking the density of air at 32° as 1, at 41° it will be 0*9820, of which the square 
root is 0*9909 ; and at 52° it will be 0*9609, of which the square root is 0*9802. Now 
the relative times of effusion observed, namely, 53T66 and 526 seconds, are as 0*9909 
to 0*9803, numbers which all but coincide with the square roots of the densities, 
0*9909 and 0*9802, at the two different temperatures. 
2. With the barometer at 30*186 to 30*150 inches, experiments were again made 
on the effusion of the same volume of dry air at 38°, 48° and 58°, four hours elapsing 
between each set of experiments, which were required to bring up the room and 
apparatus to a uniform and steady temperature. In three experiments at each 
temperature, — 
The time of effusion at 38° was 526, 527 and 526 seconds: mean 526*33. 
The time of effusion at 48° was 520, 521 and 520 seconds : mean 520*33. 
The time of effusion at 58° was 515, 516 and 515 seconds: mean 515*33. 
Here the first rise of 10° shortens the time of effusion 6 seconds, and the second 
rise of 10° shortens the time 5 seconds more. The density of dry air being 1 at 32°, 
it is at 38°, 0 9879 ; at 48°, 0*9684 ; and at 58°, 0 9497, of which three last densities 
the square roots are 0*9939, 0*9841 and 0*9745 respectively. Now the three mean 
times of effusion observed are in the proportion of the numbers 0*9939, 0*9826 and 
0*9731, which correspond more closely with the preceding square roots than could 
be expected from the nature of the experiments. It appears then that the effusion 
time of air of different temperatures is proportional to the square root of its density at 
each temperature. The velocity of the effusion will be inversely as the square root 
of the air’s density. Hence two volumes of air which have not the same tempera- 
