Chemistry. XXiil 
that of the second is 1:366, for air; 1-301 for hydrogen gas ; 
1:431 for carbonic acid gas ; and 1-415 for olefiant gas.” 
It would appear from this that the cooling power of a gas is 
proportional to a certain power of its elasticity, which power is 
different for every gas. For hydrogen it is 0°38; for air 0°45 ; 
for carbonic acid gas 0:517; and for olefiant gas 0°501. Now 
as these last three numbers differ but little from 0-5, we may say 
that in the gases to which they belong, the cooling power is 
proportional to the square root of the elasticity. 
If we reckon the cooling power of air at a given elasticity to 
be = 1, then the cooling power of hydrogen gas in the same 
circumstances is = 3°45, and that of carbonic acid gas = 0-965, 
But these numbers will change with the elasticity of these three 
gases. 
~ T shall not attempt any further account of the results obtained 
by Dulong and Petit. The termination of their paper has been 
inserted so lately in the Annals of Philosophy that it must be 
fresh in the recollection of all my readers ; and I find it very 
difficult to make their conclusions intelligible without introduc- 
ing the algebraic formulas by which they have expressed them. 
This I have endeavoured to avoid for two reasons. First, 
because I wished to make the subject intelligible to those read- 
ers who are not acquainted with mathematics ; and, secondly, 
because | could not well have given the formulas without intro- 
ducing the tables on which these formulas were founded, which 
would have swelled this article greatly beyond the limits which I 
can spare for it. 
6. Production of Cold.—It is well known that cold is produced 
by mixing together two solid substances, which, by their mutual 
action, are converted into liquids. During the liquefaction, the 
heat of liquidity is rendered latent, which occasions the cold. 
Hence the cold produced is limited by the latent heat of the 
solid body to be converted into a liquid ; and this solid body is 
in all cases ice. Now the latent heat of water is not very consi- 
derable. The cold produced can never exceed it; and indeed 
from the obvious circumstances of the experiment, it never can 
even approach it.. The latent heat of elastic fluids is much more 
considerable. When air is suddenly condensed to the fifth part 
of its natural bulk, the heat evolved is sufficient to kindle tinder; 
a temperature which must be higher than 600° of Fahr, When 
this air is suddenly allowed to expand to its original bulk, it 
resumes and renders latent all the heat which it had lost by com- 
pression. Hence if the temperature could be measured by a 
thermometer, it would be found to sink at least 600°, What a 
fpedigions cold then would be generated by suddenly setting at 
iberty air condensed to the fiftieth or hundredth part of its 
original bulk? Gay-Lussac has proposed this as a method of 
producing cold without limit. There can be no doubt that the 
cold pro ie by this method may be increased without Limit ; 
