280 
On the Measure of Temperature , 
[Sept. 
obtain, the influence of the external air must be almost inappreciable. In the 
other method again, it is necessary to allow for the loss of heat occasioned by the 
difference of temperature between the mass and surrounding medium, which allow- 
ance will also depend on the duration of the experiment. This correction can be 
determined with sufficient exactness by a subsequent observation, to be made on 
the rate of cooling of the same mass of water that had been employed in the ex- 
periments. Besides, the bulk of the bodies on which we have experimented — the va- 
ried circumstances under which each result has been obtained — and the acknow- 
ledged accuracy of the thermometer we used — all appear to us to have contributed 
to the correctness of the results we are about to submit. 
The great capacity of iron, (compared with other metals,) and the convenience 
of plunging it into boiling mercury, have been our inducements for commencing the 
series of comparisons we proposed to make, with this substance. The following 
determinations are deduced from a great number of measurements, the differences 
among which are but small. 
Mean capacity of iron from 0° to 100° 4 = ,1098 
0 to 200 = ,1150 
0 to 300 = ,1218 
0 to 350 = ,1255 
The conclusion to which this increasing series points, will be found to be further 
confirmed by the following table of the other metals : the determinations made be- 
tween 0° and 100° and between 0° and 300° only have been inserted. 
* 
Mean capacity between 
0 and 100°. 
Mean capacity between I 
0° and 300°. 
Mercury, 
,03.30 
,0.350 
,1015 
Zinc, 
,0927 
Antimony, 
,0507 
,0549 
Silver, 
Copper, 
,0557 
,0611 
,0949 
,1013 
Platina, 
,0335 s. 
,0177 
,0355 
Glass, 
,0190 
It is then equally true of the capacities of solid bodies, as it is of the dilatations, 
that they increase with the temperature, as measured by a thermometer formed of 
air: they would even increase contrary to Crawford's opinion, when referred to the 
indications of the mercurial thermometer. If this result bad been deduced from 
experiments made with a body having a constant volume, there would be no doubt 
as to the consequences to which such a truth leads. But the gaseous state or con* 
dition is the only one which allows of this condition being secured; and in the case 
of a gas, the experiment would be attended with insurmountable difficulties. How- 
ever, if the dilatation of solids were uniform, we ought not to attribute the increase 
of capacities to the quantity of heat which goes to produce the augmentation of 
volume, because this quantity being in such a case proportional to the temperatures, 
the relation of the capacities would not be affected. The same consequence does 
not hold where the dilatations are increasing; it is certain that, in this case, the ca- 
pacities referred to different points of the thermometric scale, would partake of the 
irregularities of the dilatations. We cannot form any estimate of the effect due to 
such a cause ; but what renders it probable that it is not altogether inconsider- 
able, and that the increase of capacity which we have observed, depends, at least in 
some degree, upon it, is the fact that the metals, the dilatation of which agrees with 
the most rapidly increasing series, are precisely those of which the capacities 
increase fastest. But this question can only be decided by observations, which 
should comprehend an interval of temperature much greater than that in which 
our experiments have been made. We hope soon to be able to throw light on tiis 
subject. 
We have shown, when treating of the expansion of solid bodies, that thermo* 
meters, constructed of the least fusible of the metals, if adjusted, as iu the case of 
the ordinary thermometer, by the fixed points of boiling water and melting ice, 
would, under the same circumstances of temperature, indicate very different in- 
strumental values. The same discordance would evidently be found, it appears 
from what precedes, if we were to estimate tempeiatures, as several philosophers 
Cptysiciens) have proposed, by the proportional quantities of heat lost by any 
4 The capacity of water being unity. 
