THE THEEMAL EEEECTS OE FLUIDS IN MOTION. 
585 
given in Table III. are not so numerous as might be desired, we may infer from them, 
and the results in Table II., that common air and all other mixtures of oxygen and 
nitrogen behave more like a perfect gas, e. give less coohng effect than either one or 
the other gas alone. We might expect the mixture to be something intermediate 
between the two. But this does not appear to be the case. The two are very nearly 
equal in their deviations from the condition of a perfect gas. Nitrogen deviates less 
than oxygen, but oxygen mixed with nitrogen differs less than nitrogen ! 
In the case of carbonic acid, which at low temperatures (7°) deviates five times as much 
as atmospheric air, we might expect that a mixture of CO 2 and air would deviate more 
than ah’ and less than CO 2 . This is the case (see Table IV.). Further, we might 
expect the two to contribute each its proportion of cooling effect according to its own 
amount, and its specific heat volume for volume. But do the mixtures exhibit such a 
result % No ! See column 10, Table IV., in which also note, under experiments 8 and 9, 
the great diminution produced hy the admixture of hydrogen. 
If, instead of attributing to air and carbonic acid moments in proportion to their 
specific heats, or 1 : 1’39, as we have done in column 10, we use 1 : *7, we obtain more 
consistent results. 
Let h denote the cooling effect experienced by air per 100 inches of mercury, h' that 
by carbonic acid, and A that by a mixture of volume V of air, and V' of carbonic acid ; 
then we may take 
. _ CTVS + m'V'8' 
mV + m'V 
to represent the cooling effect for the mixture, where m and ml are numbers which we 
may call the moments (or importances) of the two in determining the cooling effect for 
the mixture. The ratio of m to ml is the proper result of each experiment on a mixture, 
if we knew with perfect accuracy the cooling effect for each gas with none of the other 
mixed. Now for common air we have direct experiments (Table I.), and know the 
cooling effect for it better than from any inferences from mixtures. But for pure CO 2 
we know the effect, for the most part, only inferentially. Hence, having tried making 
: m' : : 1 : 1’39 without obtaining consistent results, we tried other proportions; and, 
after various attempts, found that m : m':: 1 : *7, for all temperatures and pressures 
within the limits of our experiments, gives results as consistent with one another as the 
probable errors of the experiments justify us in expecting. Thus, using the formula 
^ V8 + Wx’7 
v+V'x-7 ’ 
we have, for calculating the effect for CO 2 from any experiment on a mixture, the follow- 
ing formula, 
(V + V'x’7)A-V8 
Hence, using the numbers in columns 3 and 9 of Table IV. which relate to mixtures of 
air and carbonic acid alone, we find 
