1830.] 
and the Communication of Heat. 
359 
Hence, we may infer, that the cooling power of a gas is, all else being equal, 
proportional to a certain power of its elasticity ; but the exponent of this power 
varies with the gas. It is 0,38 for hydrogen, 0,45 for air, 0,517 for carbonic acid, 
and 0,501 for olefiant gas. These three last numbers differ so little from 0,5, that 
we may say their cooling power varies nearly as the square root of their elasti- 
ff we now compare the law which we have announced, to the approximations 
proposed on the same subject, by Messrs. Leslie and Dalton, we may judge of the 
error into which they hare been led, by the erroneous suppositions on which all 
their calculations have been founded, and by the want of the necessary precision 
attributable to the several methods they have employed. In fact, the former of 
these gentlemen 1 finds, by photometrical experiments, calculated after the law of 
Newton, that the cooling power of the air is as the 5th root of the densi > , 
(tension?) and Mr, Dalton* finds it proportional to the cube-root, assuming a 
certain invariable law as representing the total cooling effect foi every body, 
^Kuowin? tbehifluence exerted on the cooling process, by the temperature and 
density of^ the gas in which it is going on, it remains to discover in what manner, 
for a given condition of the fluid, that process may be found to depend on the 
dl We C have^ already* remarked, that the law which expresses this dependence is the 
same for each gas, though its elasticity vary. Let us see what happens when we 
TaS from one^aL to another, and fir this purpose let us refer in the preceding 
fables to the rates of cooling due to the sole contact of air, of ^ hydrogen, of 
carbonic acid, and of olefiant gas, these four gases having a tension of 0 ,72. 
Excess of temp, 
of the therm, 
over that ot the 
gas. _ 
Rates of cooling due solely to the contact of the gas. 
Common air 
pressure, 0m,72. 
Hydrogen gas 
pressure, 0 m ,72. 
Carbonic acid 
pressure, U m ,72. 
Olefiaut gas 
pressure 0m,72 
200 
180 
160 
140 
120 
100 
80 
5°, 48 
4 ,75 
4 ,17 
3 ,51 
2 ,90 
1 2 ,27 
1 ,77 
• • • • 
16°, 59 
14 ,26 
12 ,11 
10 ,10 
7 ,98 
6 ,06 
5°, 25 
4 ,57 
4 ,04 
3 ,39 
2 ,82 
2 ,22 
1 ,69 
7°, 41 
6 ,45 
5 ,41 
4 ,70 
3 ,84 
3 ,12 
2 ,34 
the numbers in the ttnra column uy luubc ^ v 5 
ratio of tlie loss of heat, occasioned by common air and by hydrogen gas, 
3,49 3,42 3,45 3,48 3,51 3,43. 
And as to ™ far within the limits of those 
TatCS m which everv observation is subject; we may safely conclude that the law 
errors to wfntn y for hydrogen as for common air. 
we are in search of, 13 der ivable, 5 with regard to the two other gases, by tak- 
The same c0 ®s®[j t f jj as given in the second column, with the rates 
Ifagthe ratios K T hus ”f or carbonic add, we shall have 
as found m each of ttor columns^ ^ ^ ^ 9M . 
and for olefiant gas^_ J 3g , 30 1)33 1>32 1,37 1,32. 
, thL as far as that depends on the contact of a gas, is in- 
The law of cooling, then, f Qf £ c as . And it will be obvious, on 
The law of cooling, And it will be obvious, on 
dependent of thenatnreaud of ^ ^ analog S us 0 ne, of the rate of cooling 
comparing any of the a^ov seeking is different from that of radiation, 
in a vacuum, that the law it would be useless here giving any account, 
After a great many tr.als, of wh “ w ® ul ^ le contact of a gas, varies with the 
we found that the rate of c COr din* to a law which has some analogy with 
S“olS ra ^ e c»U»g power of . gas with its elasticity ; that is to any, 
. An Experimental Enquiry Mo the Nature and Propagation of Heat, p. 
> A New System of Chemical Phdosopb, part 1. p. 
