3‘9 
Steam Engine. 
or 27| ^ above the arithmetical mean. In this experiment, 
the water, being cooled from 100^ to 972® has lost a quan» 
tity of caloric reducing its temperature only ; but this 
caloric, communicated to the pound of mercury, has pro- 
duced, in its temperature, a rise of no less than 5 a®. 
Therefore, a quantity of caloric, necessary to raise the tem» 
perature of a pound of water 2|°, is sufficient to raise that 
of a pound of mercury 57i® ; or, by the rule of propor- 
tion, the caloric, which raises the temperature of a pound 
of water 1°, will raise that of a pound of quicksilver a^- 
bout 23 Hence it is inferred, that the quantity of ca- 
loric contained in water, is to that contained in the same 
-weight of quicksilver as 23° to 1°. Or, stating the calo- 
ric of water at 1®, that of quicksilver will be ^ part of 1°, 
or 0,435^. 
When this comparison is extended to a great variety of 
bodies, they will be found to differ very considerably in their 
capacities for caloric. The results of numerous experi- 
ments of this kind are comprised in a table of specific ca- 
loric. 
The capacities of bodies for caloric, inSiience conside- 
rably, the rate at which they are heated and cooled. In 
general, those bodies are most slowly heated, and cool 
most slowly, which have the greatest capacities for heat.f 
Thus, if water and quicksilver be set, in similar quanti- 
ties, and at equal distances before the fire, the quicksilver 
will be much more rapidly heated than the water ; and, on 
removal from the hre, it will cool tvith proportionally 
greater quickness than the water. By ascertaining the 
comparative rates of cooling, we may even determine, 
With tolerable exactness, the specific caloric of bodies ; 
* The above numbers, which differ from those commonly sta- 
le cl, are g-iven on the authority of Mr. Dalton. 
T See Martino, on Heat, page 74. 
