I 
2S6 
SPECIFIC HEAT OP GASES. 
therefore, necessary to ascertain the quantity of heat thus lost 
and that was almost always very difficult. 
It was to obviate this inconvenience, that Count Rumford^ 
adopted the happy idea of taking his point of departure ^ 
not at the temperature of the surrounding air, but somewhat^ 
lower, to which he took care to reduce his calorimeter, and tolH 
continue his experiment no longer than was necessary to ac- 
quire re-heating, by a term as much above the surrounding tem- 
perature as that was above the initial temperature. By this^, 
means the healing the calorimeter was rendered independent j| 
of the loss of heat, which might arise from the air 3 for if, | 
on the one hand, during the first part of the experiment, this air, 
being hotter than the calorimeter, affords it any quantity of heat 
w'hatever, in the latter part of the experiment, by an inverse 
reason it takes off a quantity nearly equ.al ; we say nearly 
equal, for to make it exactly so, it would be necessary that the 
division of the times in which the re-heating takes place should 
be equal, which cannot always be. 
This ingenious modification has supplied us w'ith the means 
of determining, with correctness, what quantity of each gas was 
necessary tooo^nmunicate to the calorimeter an elevation to a 
given temjjerature, supposing that every gas grows cold in the 
same degree. The results of these experiments will be seen 
in the table. 
N. B. — 1 ®. We ascertained, by two preliminary experiments, 
made with great care, and which gave results very nearly the 
same, that in the experiments made agreeably to this process, 
the tube employed to heat the gas coniuiunicated, in ten mi- 
nutes, to the calorimeter, a lieat sufficient to raise its tempera- 
ture to 0 ®,igO j and as the heat thus afforded to the calorime- 
ter was sensibly proportional to the time, nothing is more 
easy than to correct the results in that respect. 
2- The correction respecting the pressure has been made in 
a manner analogous to that which has been calculated in the 
preceding table. 
It results, from what we have advanced, that the numbers 
contained in the last column of this table, are in an inverse 
ratio 
1 
