Why Rubbing Your Palms Together Proves Heat Isn't a Fluid At All

For a long time, heat was pictured as caloric, an invisible fluid filling the pores of a substance, flowing from a hot body to a cold one exactly as water flows through a pipe connecting two tanks at different heights, until the levels equalise. In 1798, Benjamin Thomson (Count Rumford) made an observation the caloric theory could not explain: boring a brass cannon generated enough heat to boil water, and crucially, the amount of heat produced depended only on the work done by the horses turning the drill, not on how sharp the drill was. If heat were truly a fluid squeezed from the metal's pores, a sharper drill should have scooped out less caloric fluid, producing less heat -- but this was never observed. The natural explanation was that heat is a form of energy, and Rumford's cannon was really demonstrating the conversion of energy from one form, work, into another, heat. This insight underlies something as simple as rubbing your palms together on a cold day: no fluid flows in from anywhere, yet your hands genuinely warm up, because the work of rubbing is being converted directly into heat.

Two gas systems, A and B, separated only by an adiabatic wall (an insulating wall that blocks any flow of heat), can have completely independent pressures and volumes -- any state of A is compatible with any state of B. Replace that wall with a diathermic wall (a conducting wall that allows heat to flow), and the two systems' pressures and volumes spontaneously change until they settle into new values and stop changing entirely; the two systems are then in thermal equilibrium with each other. A system itself is said to be in a state of thermodynamic equilibrium when its macroscopic variables, pressure, volume, temperature, mass, and composition, do not change with time. Note carefully that thermodynamic equilibrium is a different idea from mechanical equilibrium (zero net force and torque): a rigid, sealed, fully insulated container of gas with fixed pressure, volume, and temperature is in thermodynamic equilibrium, regardless of whether any forces act on it.

Imagine systems A and B, separated from each other by an adiabatic wall, but each in contact with a third system C through a conducting wall. A and B will each individually reach thermal equilibrium with C. Now replace the adiabatic wall between A and B with a conducting wall, while insulating C from both -- remarkably, the states of A and B change no further; they are already in thermal equilibrium with each other. This observation is the Zeroth Law of Thermodynamics: two systems in thermal equilibrium with a third system separately are in thermal equilibrium with each other. (R.H. Fowler formulated this law in 1931, well after the First and Second Laws had already been named -- hence 'zeroth'.) The Zeroth Law shows there must exist a physical quantity with the same value for any two systems in thermal equilibrium -- this quantity is called temperature (T). If A and B are each separately in equilibrium with C, then TA=TC and TB=TC, which forces TA=TB, meaning A and B are also in equilibrium with each other, exactly as observed.

Internal energy (U) is simply the sum of the kinetic and potential energies of a system's molecules, measured in the frame where the system's own centre of mass is at rest -- it deliberately excludes the bulk kinetic energy of the system moving as a whole. A bullet fired from a gun has its mechanical kinetic energy changed, not its temperature; only when it strikes wood and stops does that kinetic energy convert into heat, raising the temperature of the bullet and surrounding wood, because temperature relates to the internal, disordered motion of molecules, not the bullet's overall motion. Crucially, internal energy is a state variable -- it depends only on a system's current state (its pressure, volume, temperature), never on the path taken to reach that state. Heat and work, by contrast, are NOT state variables; they are two distinct modes of transferring energy TO a system, each changing its internal energy. Heat is energy transfer due to a temperature difference between a system and its surroundings; work is energy transfer brought about by other means entirely, such as pushing a piston, that involve no such temperature difference. Saying 'a gas has a certain amount of heat' is as meaningless as saying 'a gas has a certain amount of work' -- but saying 'a certain amount of heat was supplied' or 'a certain amount of work was done' is perfectly meaningful.

If ΔQ is the heat supplied to a system, ΔW the work done BY the system on its surroundings, and ΔU the resulting change in internal energy, the general principle of conservation of energy gives ΔQ=ΔU+ΔW -- the First Law of Thermodynamics. For a gas in a cylinder with a movable piston, expanding against a constant external pressure P, the work done by the gas is simply ΔW=PΔV. This lets us calculate real changes: the measured latent heat of water is 2256 J/g, so converting 1 g of water to vapour needs ΔQ=2256 J; since 1 g of water occupies about 1 cm³ as liquid but 1671 cm³ as vapour at atmospheric pressure (1.013x10⁵ Pa), the work done by the expanding vapour is ΔW=P(Vg-Vl)=1.013x10⁵x1671x10⁻⁶≈169.2 J. The First Law then gives ΔU=2256-169.2=2086.8 J -- showing that most of the heat supplied during boiling goes into increasing the water's internal energy (breaking molecular bonds as it becomes vapour), with only a small fraction spent pushing back the atmosphere. Since U is a state variable, ΔU depends only on a system's initial and final states, never on the path taken between them -- but ΔQ and ΔW individually DO depend on the path, even though their difference, ΔQ-ΔW, does not.

Just as heat capacity S=ΔQ/ΔT and specific heat capacity s=(1/m)(ΔQ/ΔT) were defined for a given mass, the molar specific heat capacity, C=(1/µ)(ΔQ/ΔT), is defined per mole. For a solid of N atoms, each vibrating with average energy kBT per dimension (3kBT in three dimensions), the law of equipartition of energy predicts a molar specific heat of C=3R for one mole -- a prediction that agrees well with experiment at ordinary temperatures for most solids (carbon is a notable exception, and the agreement breaks down at low temperatures). The old unit of heat, the calorie, was originally defined as the heat needed to raise 1 g of water by 1°C; since specific heat varies slightly with temperature, it was later redefined precisely as the heat needed to raise 1 g of water from 14.5°C to 15.5°C, giving 1 cal=4.186 J -- though since heat is simply a form of energy, the SI unit joule is now preferred and 'mechanical equivalent of heat' is a superfluous, outdated idea. For gases specifically, specific heat depends on the conditions under which heat is supplied, giving two distinct molar specific heats: Cp (at constant pressure) and Cv (at constant volume). Using the First Law, ΔQ=ΔU+PΔV, along with the fact that an ideal gas's internal energy depends only on temperature, a short derivation shows Cp-Cv=R exactly, for any ideal gas -- a clean, universal relation known as Mayer's relation.

State variables like pressure, volume, temperature, and mass describe equilibrium states; the relation connecting them, such as the ideal gas relation PV=µRT, is called an equation of state. State variables come in two kinds: extensive variables (like internal energy U, volume V, and total mass M) that halve if the system is divided into two equal parts, and intensive variables (like pressure P, temperature T, and density ρ) that stay unchanged in each half. A quasi-static process is an idealised, infinitely slow process in which the system remains in equilibrium with its surroundings at every single stage -- real slow processes approximate this well. Four special quasi-static processes matter most: in an isothermal process (constant T), PV=constant (Boyle's Law) and the work done by an ideal gas expanding from V1 to V2 is W=µRT ln(V2/V1); in an adiabatic process (no heat exchange at all), PVᵞ=constant, where γ=Cp/Cv, and the work done is W=µR(T1-T2)/(γ-1); in an isochoric process (constant V), no work is done at all, so all supplied heat changes internal energy directly; and in an isobaric process (constant P), the work done is W=P(V2-V1)=µR(T2-T1). A cyclic process returns the system to its exact starting state, so ΔU=0 over a full cycle, meaning total heat absorbed exactly equals total work done by the system.

The First Law alone permits many things that are never observed -- a book on a table could, in principle, cool spontaneously, converting its own internal energy into mechanical energy and hopping into the air, without violating conservation of energy at all. Yet this never happens. Some additional principle of nature forbids it: the Second Law of Thermodynamics. Two equivalent statements capture it precisely. The Kelvin-Planck statement: no process is possible whose sole result is absorbing heat from a reservoir and converting all of it into work. The Clausius statement: no process is possible whose sole result is transferring heat from a colder object to a hotter one. In simple terms, the Second Law says a heat engine's efficiency can never reach 100%, and a refrigerator's coefficient of performance can never be infinite. A process is reversible only if the system and surroundings can both be returned exactly to their original states, with no change anywhere else in the universe -- this requires the process to be quasi-static AND free of dissipative effects like friction or viscosity. Since dissipation is present everywhere in nature and can only be minimised, never fully eliminated, irreversibility is the rule, not the exception: a hot vessel's base cools by spreading heat to its cooler sides, but the reverse (a cool part spontaneously warming the hot base) is never seen; free expansion of a gas into a vacuum, an explosive chemical reaction, and gas leaking from a cylinder to fill a room are all genuinely irreversible.

A reversible heat engine operating between just two temperatures is called a Carnot engine, first analysed by French engineer Sadi Carnot in 1824. Its cycle has exactly four reversible steps: an isothermal expansion at the hot temperature T1, absorbing heat Q1; an adiabatic expansion cooling the gas to T2; an isothermal compression at T2, releasing heat Q2; and an adiabatic compression returning the gas to its original state at T1. Carnot's theorem proves two remarkable, universal results: no engine operating between two given temperatures can ever be more efficient than a Carnot engine, and the Carnot engine's efficiency does not depend at all on what working substance it uses. Its efficiency works out to a strikingly simple formula: η=1-T2/T1 -- depending only on the two absolute temperatures, nothing else. A Carnot engine operating between a 500 K furnace and a 300 K environment can therefore never exceed 40% efficiency, no matter how it is built or what gas fills it. This single result, along with the companion relation Q1/T1=Q2/T2, is so universal that it can be used to define a truly fundamental thermodynamic temperature scale, independent of any particular substance's properties -- a fitting capstone to a chapter that began by showing heat was never really a substance at all.