OF GASES AND VAPOURS. 199 
c F ‘ 
= the specific heats c and ¢, for the various gases, and has 
1 
shown, @ posteriori, the admissibility of that assumption from the 
coincidence of the values of ¢ calculated by him with those ob- 
served by De la Roche and Bérard. 
13. The specific heats of the gases are given above by the 
equations (IV.) No. 7. For atmospheric air we find 
c = 0:267 — 0180. log /, 
Po 
é,= 0189 —0'180 log 2, Los SA 6.) 
) 
where ¢ and ¢, are the specific heats, the pressure and yolume 
being constant. 
Only ¢ can be measured directly ; Suermann has published some 
experiments on the subject, which however by no means agree 
with this formula. We obtain, for instance, for p, = 760, and 
According to the formula. According to Suermann. 
p = 69110 ¢ = 0°2744 0°2866 
p = 50886 ¢ = 0°2983 0-2964 
p = 319°63 ¢ = 0'3347 0°3149 
The values calculated from the formula follow, it is true, in 
general the same course as those observed, but they evidently 
increase at a greater rate. However, the observer himself has 
raised great doubt with respect to the admissibility of the ob- 
served values, to which indeed others might be added. This 
non-coincidence therefore must certainly not be regarded as a 
proof against the correctness of the advanced formula, 
There are two experiments of De la Roche and Bérard which 
may be mentioned here; according to these, the specific heat 
of 1t- air measured at 0° and under the pressure 1™-0058, and the 
specific heat of 14 measured at O° and under the pressure 
0™-7405, being in the ratio of 1°2044:1, or as 1°2396:1; the ob- 
servers regard the latter numberas far more accurate than the first. 
According to the above formula, the specific heats of the 
units of weight with the pressures 10058 and 0™-7405 equal 
0°245 and 0°269: consequently the relation of these heats re- 
ferred to quantities whose weights are in the proportion Re 
equal 10058 x 0°245 
0°7405 x 0°269 
which agrees more accurately with 1:2396 than could have been 
expected in such observations, 
—g Ove 
