24 
28 per cent, oxygen, and 72 nitrogen. This difference 
no doubt arises, not so much from the air examined, as 
from the uncertain methods of trial made use of ; but it 
seems very probable, from the more accurate experiments 
of Berthollot, Davy, &c., that the atmosphere is com- 
pounded of oxygen and nitrogen gases in some constant 
proportion between those of 21 to 79 and 22 78; or 21 
to 22 per cent, oxygen. 
It may be objected, that a mixture of these two gases in 
any of the above proportions, forms a compound of a specific 
gravity superior to that of the atmosphere, and a small 
addition of carbonic acid gas, generally found therein, will 
add to the objection ; something therefore must exist in the 
atmosphere of a considerably less specific gravity than any of 
these mentioned gases. 
M. de Luc long ago stated, that the evaporation of water 
was not effected by chemical solution, but by union with 
caloric ; forming steam or vapour, which he found to exist 
in and augment the elasticity of the atmosphere. Mr. 
Dalton has lately verified these observations of de Luc ; and 
by numerous experiments determined, not only the existence, 
but also the precise expansive force of vapour in the atmos- 
phere, in all pressures and temperatures. From the force 
he gives to vapour at 55°, and the expansion it occasions in 
dry air, under the pressure of 30 inches of mercury, we find 
that aqueous vapour occupies a space equal to 1^ per cent, 
of the atmosphere at that temperature. This affords the 
bulk of vapour, but its specific gravity seems not to have 
been determined by any direct experiment. Mr. Watt 
never found steam more than 1800 times rarer than water 
" at the temperature of 212°, and under the pressure of the 
" air." Taking the expansion therefore at 1800, the specific 
gravity will be .55 ; that of water being 1000 : and this 
result very nearly corresponds with that of M. Schmidt, 
who by saturating dry air with moisture or vapour, calcu- 
