Heat as energy transfer

In brief

A block of ice sitting on a kitchen counter melts. Something moved from the warm room into the ice to cause that melting, and everyday language calls it "heat entering the ice." But if you later asked whether the water that used to be ice now "contains" that heat, the honest answer is no: it contains energy, stored as increased molecular motion and altered configuration, but that stored energy is no longer called heat. Heat is the name reserved for energy in the act of crossing a boundary because a temperature difference exists on either side of it. The instant the energy arrives and settles into the receiving object, it becomes internal energy, not heat. This distinction, transit versus storage, is the single most useful discipline in all of thermodynamics, because sloppiness about it produces impossible bookkeeping.

The full treatment

First look: a river, not a lake

Think of heat the way you think of a river current, not a lake. A lake is a store of water; a current is water moving past a point. If someone asks "how much river is under this bridge," the question is confused, because a river is not a fixed amount of water, it is a flow. Similarly, asking "how much heat is in this coffee" is confused in the technical sense, because heat only exists while energy is crossing a boundary. What the coffee has, stored and measurable at any instant, is internal energy. What can happen to it, while it sits in a cooler room, is that energy flows out of it as heat, exactly the way water flows past the bridge, without the coffee "running out of river."

Building the idea: what makes a transfer count as heat

Energy can cross the boundary of a system by more than one route. Push a piston into a gas and you transfer energy by organized, directional force acting through a distance, which is called work. Put a warm object next to a cold one with no motion of either boundary and energy still crosses between them, driven purely by the temperature difference and carried by molecular collisions or radiation. That second route, energy transfer caused specifically by a temperature difference and mediated by disordered microscopic collisions rather than organized mechanical motion, is what the word heat is reserved for.

Three conditions must hold for a transfer to be classified as heat rather than work. First, there must be a temperature difference between the two regions; without one, as established in the discussion of thermal equilibrium, there is no net drift of energy at all. Second, the transfer occurs through molecular collisions, or the emission and absorption of electromagnetic radiation, rather than through a moving mechanical boundary doing organized force times distance. Third, the direction is fixed by the temperature difference alone: energy flows spontaneously from the higher-temperature region to the lower-temperature region, never the reverse, absent some additional device doing work to force it backward, such as a refrigerator.

The formal quantity: heat as Q, a process variable

Heat is given the symbol Q and measured in joules, the same unit as any other form of energy, or historically in calories. A crucial feature of Q is that it is a process quantity, not a state quantity. A state quantity, like temperature or internal energy, has a definite value at a given instant regardless of history. A process quantity, like heat or work, only has a value describing something that happened during a particular process, over some interval, and depends on the path taken, how fast the transfer occurred, whether it happened at constant pressure or constant volume, and so on. Writing "the heat of this object is five hundred joules" is a category error for the same reason that "the distance traveled by this car" is meaningless without specifying over what time interval or route; distance traveled, like heat transferred, is a quantity that describes a process, not a snapshot.

By convention, Q is taken positive when heat flows into a system and negative when it flows out. If two objects at temperatures T-hot and T-cold are placed in contact, with T-hot greater than T-cold, then over some time interval an amount of heat Q flows from the hot object to the cold one, with Q positive for the cold object and negative, of equal magnitude, for the hot object, since energy is conserved: whatever leaves one enters the other, assuming no losses to the surroundings.

Mechanism at the molecular scale

At the level of individual particles, heat transfer by conduction is exactly the collision process described for reaching thermal equilibrium: faster molecules at the boundary strike slower ones and, on average, hand over kinetic energy. Radiative heat transfer works differently in mechanism, energy carried by electromagnetic waves emitted by any object above absolute zero and absorbed by another, but the outcome is the same, net energy drift from hotter to colder, with zero net radiative transfer once both objects share a temperature and thus emit and absorb at matching rates. In both mechanisms, the defining feature is that the transfer is driven by the temperature difference itself, and it ceases, not because energy runs out, but because the difference driving it disappears.

Lineage

Confusion between heat as a substance and heat as a process was resolved only gradually. Eighteenth-century caloric theory treated heat as a weightless, conserved fluid that flowed from hot to cold bodies, and this framework, while wrong about the mechanism, correctly captured some of the bookkeeping, which is part of why it persisted so long. Benjamin Rumford's cannon-boring experiments in the 1790s showed that friction could generate heat without limit, incompatible with heat as a fixed, conserved substance stored in an object. James Prescott Joule's mid-nineteenth-century experiments, measuring the mechanical equivalent of heat by churning water with paddle wheels, established that heat and work are interconvertible forms of energy transfer, not different substances, finally cementing heat as energy in transit. Fermi's and later Feynman's expositions carry forward this settled distinction, treating heat and work symmetrically as the two routes by which energy crosses a system boundary.

The strongest case for it

Treating heat strictly as a boundary-crossing process quantity, rather than a stored substance, is what makes energy accounting for engines, refrigerators, weather, and living organisms tractable and correct. It correctly predicts that friction and other dissipative processes can generate apparently unlimited heat from mechanical work, matching Rumford's and Joule's experiments and every engine ever built. It correctly forbids nonsensical statements like "this object has run out of heat to give," since what actually runs out is a temperature difference, not a heat reservoir. It also correctly explains why insulated systems that exchange no heat can still change temperature, through work alone, as in the compression heating of air in a bicycle pump, something that a substance theory of heat struggles to accommodate cleanly.

The strongest case against it

The strict process definition of heat has to be applied carefully or it collapses back into the older, looser language. In casual speech and even in engineering shorthand, phrases like "heat content" or "heat capacity" persist and can mislead a careful reader into thinking heat is stored; heat capacity, properly understood, measures how much a system's temperature rises per unit of energy added, whether that energy arrives as heat or as work, not a reservoir of heat itself. A second limitation is that the clean separation of heat and work depends on identifying a boundary and a mechanism, and in genuinely disordered or turbulent processes, such as fast irreversible compression, the line between organized work and disordered heating can blur, requiring more careful thermodynamic treatment than the simple two-route picture offers. A common misconception is believing an object at high temperature necessarily "has more heat to give" than one at lower temperature but greater mass; a spark can be far hotter than a swimming pool yet transfer far less total energy, because the total energy released also depends on how much internal energy is stored, not on temperature alone.

Where it stands now

The definition of heat as energy transferred because of a temperature difference, distinct in kind from work and from stored internal energy, is settled physics with broad consensus, unchanged since Joule's mechanical equivalent of heat closed the caloric-theory debate in the nineteenth century. It is the operational basis for the first law of thermodynamics and for every calculation of engine efficiency, calorimetry, and thermal design used today.

Test yourself

A sealed metal box sits in a room. You are told that over one hour, one thousand joules of energy entered the box, and during that hour a fan mounted rigidly to the box's outer wall was also spinning, driven by a small motor outside the box, pushing air against the wall. Explain how you would determine how much of the one thousand joules should be classified as heat and how much as work, what additional information you would need to make that determination, and why it would be wrong to say the box now "contains one thousand joules of heat."