PROFESSOR GRAHAM ON THE MOTION OF GASES. 
617 
approaching- closely to 04375 or 0-44, and the coefficient of carbonic acid to ap- 
proach closely to 0’75. The coefficient of nitrogen has fallen at the same time from 
0^3/3 (its coefficient of effusion, that of oxygen being 1) to nearly 0*875. 
With this length of the tube, the influence of effusion upon the transpiration rate 
of the gas has ceased to be sensible. 
The action of the tube attains a certain uniformity, at the same time, in another 
respect. When the tube is deficient in length, the coefficient of the gas varies greatly 
with the extent of the exhaustion of the vessel into which the gas is flowing; the rate 
of transpiration inclining most to the rate of effusion at a high degree of exhaustion 
of the aspirator jar. But the amount of this variation progressively diminishes as the 
tube becomes longer, till at last the coefficient of transpiration remains nearly if not 
perfectly constant, whether the aspirator jar be entirely vacuous, or contains already 
gas of the tension of half an atmosphere, three-fourths or even seven-eighths of an 
atmosphere. 
The tube is now of the length most favourable for experiments of transpiration. 
Farther addition to the tube appears to have no effect in altering the coefficient of 
transpiration of air, provided the aspirator jar into which the air passes is vacuous 
or nearly so; or the coefficient of transpiration into a vacuum appears to remain con- 
stant for all greater lengths of the tube. But the extent of the barometric range in 
the aspirator jar to which the vacuum coefficient is found to apply is gradually limited. 
Instead of extending over seven-eighths of an atmosphere it may be contracted to an 
eighth of an atmosphere or less, by greatly lengthening the tube. 
The coefficients of transpiration which I have endeavoured to ascertain are properly 
therefore the relative times of passage of the gases into a vacuum, at the mean atmo- 
spheric temperature, or near that temperature. 
But even with the most favourable length of the tube, when the aspiration is feeble, 
and does not exceed 2 or 3 inches of mercury, the rate of transpiration is modi- 
fied from an interference, which is connected with the excessive resistance of the tube 
to the passage of the gas. I have not been able to undertake an examination of the 
nature and extent of this interference, but suppose it to depend upon friction, while 
the rate of transpiration seems again to depend upon a constitutional difference in the 
gases themselves. 
The theory of the transpirability of gases, which at present appears to me most 
probable, is that it is a kind of elasticity depending upon the absolute quantity of 
heat, latent as well as sensible, which different gases contain under the same volume; 
and therefore that it will be connected more immediately with the specific heat than 
any other property of the gases. 
The only other gases besides those already experimented upon, which could be re- 
tained over water, and exposed to the metallic parts of the apparatus without injury, 
are olefiant gas, nitric oxide and sulphuretted hydrogen; and these were submitted 
to transpiration with several of the tubes already used. I subjoin the results, although 
less complete than is desirable, under the head of each gas. 
4 L 
MDCCCXLVI. 
