In the eighteenth century heat was investigated to determine how it worked. But from the view of the time they wre trying to determine 'the nature of heat'. From experience, if you but a hot body next to a cold body (hot iron on a block of ice for example) the hot body grows colder and the cold body hotter, until both are the same hotness-coldness.

It is reasonable to say 'something' is being transferred between the two bodies. It is not unreasonable to say two 'somethings' are being transfered between them (a coldness from the cold body to the hot; a hotness from the hotbody to the cold). Perhaps for psychological reasons people tend to say something goes from hot to cold, as the sensation of hotness is perceived more readily than coldness.

All the substances which moved around, and were not visible, like electricity were called fluids, as liquids move more easily than olids, by analogy with macroscopic fluids. Whether these fluids were composed of atomss of a specific type (the caloric atom) or were someting else was not certain but proposals for each idea and others were made at the time. A theory of a different nature entirely was to say that heat was a result of something else, specifically if heat was propogating motino of atoms (a vague sort of idea itself)a and was not itself a distinct object seperable from other things of the same nature.

However the idea that heat was a substance by analogy with other substances suggested applying Lavoisoire's idea of the conservatin of mass to the conservation of heat [mass]. And this concept accounted for a large number of experimental results.

Another useful hypothesis was to assume heat is a substance (and a fluid one at that) somewhat like air (another analogy formed] in a car tire and a bicycle tire. The temperature of heat would then be compared to the pressure in the tires. The pressure can be felt, and in a limited way temperatures can be felt. A pressure gauge can measure one more precisely and a thermometer can measure the other more precisely. And still thinking of heat as a substance leads one to think of the amount of heat that is available, and this just like the air in the tires. For example the air pressure in the two tires may be the same, but the amount of air is clearly much different. Likewise the temperature of a heated nail and that of a anvil may be the same, but the amount is much different as would be shon by dropping the hot nail ina trough of water, which would have little effect on the temperature of the water, and then dropping the heated anvil in a trough with a similiar amount of water and finding the temperature change much more.

Now the question: " Is the warmed body warmed by the same temperature change as the cooled body is colled? The answer is no if the two materials are made of different materials. If a kilo of water at 4o C is mixed with a kilo of water at 100 ^o C then the final temperature of both is 48^o C ((100-4)/2).
The same experiment using hot mercury and cold water showed both substances ended up at the same temperature, but as the same weight was used, the water was only slightly warmer, the mercury much cooler. Specifically, if a kilo of water at 0^{oC was mixed with a kilo of mercury at 100 oC then the final temperature of both would be 3oC (not 50oC). As mercury is very dense the volume of mercury at the same weight as water is much smaller in volume, so equal volumes were tried but the resulting temperature was still not 50oC!}

Notice that the data is now forcing hypothesis formation. One elegant formulatin was the suggestion (by Joseph Black) that each substance can only hold so much caloric fluid or heat, and using the tacit assumption that a rise in temperature of a substance implies a quantity of heat then it is possible to find out the relative capacities for holding heat [heat capacities] by defining a 1^oC rise in a kilo of water is an amount of heat called 1 Calorie. [England used 1 lb of water rising 1^o Farenheit calling that a British Thermal Unit (BTU)]. Now since it has been assumed by analogy that heat is conserved (like mass) not lost nor gained, then wh4en a kilo of water is warmed a certain number of degrees by (say) a killo of mercury the amount owarmed is equal to the amount of heat lost by the mercury. Since the heat lost and gained is the same then by [again] defining the heat capacity of water as 1 (kilo/^oC] we can compute a value of the heat capacity for mercury with the temperature change data.
Specifically:

Temperature Change and Heat Capacity Calculation

	Temperature
	Start	End
Capacity
H2O	0oC	3oC
Hg	100oC	3oC

This table implies that the heat capacity of Hg(Mercury) compared with the heat capacity of H₂O (water) is the same as a change in the temperature of mercury compared to the change in temperature of water which gives the capacity of mercury as 0.03 compared to H₂O value of one. Now every substances capacity can be found in the same way and with tath you can predict the final temperatures in a variety of mixtures. A reasonable consequence of this heat capacity knowledge is to look on it as a measure of how much that substance resists a change in temperature. Substances of high heat capacity require moree heat to warm them and conversely give out more heat on coolig than do substances of lower heat capacity. And water with its high heat capacity becomes a moderator of local climate as anyone living next to a large body of water knows.

Again measuring the rise in temperature of mixed fluids starting at two different temperatures with respect to time givres the (non - intuitive) result that the rise in temperature from the freezing point to the equilibrium temperature does not occur gradulally but suddenly after a time interval has past. Until that time interval passes the temperature stays constant at the freezing point until all the substance is unfrozen then the rise in curvilinear but approximately continuous thereafter.
Now suddenly this phenomena (called laten heat) explains the [usual and therefore expected] behavior of snow not suddenly all melting when it reaches 0^oC which would cause truly disasterous rushes of water in warm weather but instead slowly the snow melts away. As the book quotes Black:

pp 89. "It is evident the melting ice receives heat very fast, but the only effect of this heat is to change it into water, which is not the least sensibly warmer that the ice was before....A quantity, therefore of the heat, or of the matter of the heat, which enteres into the melting ice, produces no other effect but to give it fluidity, without altering its [temperature]..."

The same phenomena was found when the substance was boiling (heat of vaporization). It was found a definite amount of heat is needed in order to melt a definite weight of ice to the same weight of water. And when the water was refrozen the heat lost was also containing the latent heat gained.

Now if you believed that heat was a substance and further believed that it was composed of atoms of caloric then there is a clear explanation of latent heat as a chemical reaction betwen atoms of heat and those of water so that (on melting) caloric atoms (say) were more fluid and bound themselves around the water atoms and made it more fluid in turn than water surrounded by water as it was in the state of ice.

Thinkig of heat as a fluid would give rise to the question "how fast does it flow?", and "how long does it take hot and cold bodies to reach an equilibrium temperature" which is a similiar question. Fourier was able to crate a fluid model of hat that would make accurate estimates of the time interval from initial mixing to equilibrium temperature.

Thus, as opposed to the atomic motion theory, the caloric theory (a fluid or a fluid of atoms) was very successful in explaining a large body of experimental evidence and of predicting it too.
Now Count Rumford in the late eighteenth centure did experiments on heat. He was trying to show that it did not exists as a fluid, or a seperate substance at all, but rather that it arose from the motions of atoms of any substance, and therefore a property of material and physics.

Here is the results of his experiments:

page 92:

1. [Rumford] used...latent heat to perform...a [good] experiment to see if 'caloric' had any weight [as a substance ought to]. He found that it did not.
2. He showed that [friction] could produce an unlimited quantity of heat [strongly suggesting that heat] cannot be composed of an indestructible substance.
3. He showed that a solutin of salt in water is carefully overlaid with a less dense layer of pure water and left quiet, without stirring, the salt spreads upward into the pure water thus suggesting that the atoms of amtter are in constant motion, a fact consistent with his idea that heat is themotin of atoms.