1819.] and onthe Laws of the Communication of Heat. 167 
body which we employed, the varied cireumstances in which 
each determination was made, the precision of the thermometer 
which we employed, seemed all to concur to ensure the accuracy 
of our results. 
The great capacity of iron (relatively to the other metals), and 
the possibility of plunging it in boiling mercury, mduced us to 
begin with this substance, the comparisons which we proposed 
to make. The following determinations are deduced from a 
ereat number of measurements, agreeing nearly with each other. 
Mean capacity of iron from 0° to 100°= 0-1098* 
0 to 200 = 0:1150 
0 to 300 = 01218 
0 to 350 = 0°1255 
The result, indicated by the way in which these numbers vary, 
is verified in the following table for the other metals. We have 
satisfied ourselves with inserting the measures taken at 100° and 
at 300°. 
Mean capacity between Mean capacity between 
0° and 100°. 0? and 300°. 
Merenty. 02. eh Ud ee a fe SOCK: 003506 
ae ere He cage! Cn ER AR ar ... 01015 
Antimony. ...... OO RD .- 0°0549 
Sercrs sss oe rR a a a a Core OOH 
Seer eter. a dae ps ee rete 0-1013 
Platinum....... SOE eae ees aia E 0°0355 
od aA aotight ss: a ii ce RARE 0-1900 
The capacity of solid bodies then follows the same law with 
their liquidity ; it increases with the temperatures, measured by 
an air thermometer. They would even be increasing, contrary 
to the opinion of Crawford, if we were to employ a mercurial 
thermometer. Ifthis observation had been made upon bodies of 
an invariable volume, there would remain no doubt respecting 
its consequences ; ‘but the gaseous is the only state which per- 
mits us to satisfy this condition ; and in that state the experiment 
presents insurmountable difficulties. If the dilatation of solids 
were uniform, we could not ascribe the increase of capacities to 
the quantity of heat which produces the increase of volume ; for 
this quantity remaining then proportional to the temperatures, 
could not affect the ratio of the capacities. The caseis not the 
same when the dilatabilities are increasing. It is evident that in 
this case the capacities taken at different heights of the thermo- 
metrical scale ought to be affected by the irregularity of the law 
of dilatation. We cannot form any conjecture of the intensity of 
the effects due to this accidental cause. But what would lead to 
the belief that they should not be neglected, and that the increase 
* The capacity of water is reckoned 1, 
