Temperature predicts thermal contact, but it does not count how much energy an object contains.
By the end: choose a boundary, identify equilibrium, and use absolute temperature without confusing it with energy.
This course assumes high-school science and basic algebra, not calculus or prior thermodynamics. It introduces logarithms, microscopic counting, and the required chemistry terms as they arise. The goal is to follow expert explanations and test their assumptions, not design pressure equipment, certify appliances, or reproduce every formal statistical-mechanics derivation. Displayed numerical values are rounded; calculations use unrounded model values.
Choose what you are describing
A system is the matter or region you choose to study. Everything outside its boundary is its surroundings. A closed system exchanges energy but not matter; an open system can exchange both. An isolated system exchanges neither. A sealed metal vessel can be closed without being insulated. Insulation restricts heat transfer, not necessarily work. A piston can move even if its walls are insulated.
Specify the boundary before counting a transfer. If two blocks inside an insulated box exchange energy, heat crosses each block’s boundary but not the boundary of the combined box. Internal transfers cancel in the combined accounting. Most models in this course use closed systems, neglect motion of the whole object and changes in its height, and describe uniform equilibrium states.
Follow a transfer across nested boundaries
Imagine two solid blocks, A and B, inside an ideal insulated, rigid enclosure. A starts hotter than B. Choose A alone as the system and draw its boundary around that block, not around the enclosure. Energy passing from A to B crosses this boundary outward. No matter needs to leave A for that transfer to occur. The block is therefore closed, yet it is not isolated. Closed describes the restriction on matter crossing a boundary, rather than a restriction on every possible energy transfer. A sealed block can still exchange energy through its surface without exchanging any of its constituent particles.
Now choose both blocks together. The same contact transfer crosses an internal dividing surface, but it never crosses the outer system boundary. If A loses 600 J while B gains 600 J, their combined internal-energy change is zero. This does not mean that neither block changes. Their temperatures can approach each other while the total stays fixed. An outer energy balance therefore cannot by itself reveal how energy is distributed among the parts. In the boundary drawing, follow the arrow once for each choice of system. Ask whether it actually crosses the selected outline before assigning it to an external energy account.
The same 600 J transfer crosses one boundary, not the other
Schematic boundaries and transfer arrow, not scaled block dimensions. A and B have negligible work. The insulated, rigid outer enclosure exchanges no energy or matter.
A boundary can also cut through a mechanical connection. Consider a sealed gas beneath a movable piston. Molecules remain inside, but a moving piston can transfer energy to a load outside the gas boundary. Insulating the cylinder would restrict heat transfer without stopping this work transfer. Rigid and insulated impose different restrictions. Rigid prevents displacement of that boundary, while insulated suppresses energy transfer driven by a temperature difference. Neither word alone says whether electrical energy or matter could cross some other part of the chosen boundary.
The location of an energy store matters too. Suppose a gas and an initially compressed spring are enclosed together. The spring can release energy and warm the gas through internal mechanical interactions. The gas alone gains energy, but the larger isolated assembly does not receive energy from outside. If the system includes the spring, its changing stored mechanical energy must appear in the account. Treating the whole assembly as gas internal energy would erase the source of the increase rather than explain it.
Use the second boundary sketch to compare two descriptions of a hypothetical motor driving a stirrer in a sealed fluid. With only the fluid selected, shaft work crosses inward. With motor and fluid selected, the shaft transfer is internal and electrical work can cross the larger boundary instead. These are alternative accounts of one arrangement, not competing explanations. For either boundary, identify the energy source and the final stores. Do not count both the electrical input and the resulting shaft input as separate external inputs to the larger system.
Enlarging the boundary changes which work input is external
Schematic motor and fluid; no operating protocol or efficiency value. The shaft transfer is internal to the larger selection. Electrical input can also change the motor's energy or be dissipated, so no numerical equality of the two arrows is assumed.
Use equilibrium to define temperature
Bring two objects into thermal contact without doing work on them. If no net energy transfer is driven by a temperature difference, they are in thermal equilibrium. The zeroth law says that objects each in thermal equilibrium with a third object are in thermal equilibrium with each other. This makes a thermometer useful: it compares temperatures rather than measuring a stored quantity called heat. A real thermometer needs time to equilibrate and may disturb a small sample.
Kelvin is an absolute temperature scale. Add 273.15 to the numerical Celsius reading to get kelvin. A difference of 10 °C is a difference of 10 K, but a ratio of Celsius readings has no general thermal meaning. At 300 K and 600 K, a dilute classical monatomic gas has a twofold ratio of mean translational kinetic energies per particle. At 20 °C and 40 °C it does not. Those temperatures are 293.15 K and 313.15 K.
Read equilibrium without mistaking a stationary reading for it
A thermometer reaches a temperature through interactions with its surroundings. Its reading refers directly to the thermometer itself. To infer a sample temperature, the thermometer and the relevant part of the sample must reach thermal equilibrium. The zeroth law then lets equal readings compare samples that never touch each other. If A and B separately equilibrate with a thermometer at the same reading, they share a temperature under those conditions. No claim about equal mass, chemical composition, or stored energy follows from that comparison. The shared reading identifies one thermal property, rather than a complete description of either sample's state.
Consider a thermometer with constant whole-sample capacity 10 J/K, initially at 300 K, and a small sample with capacity 30 J/K, initially at 340 K. In an isolated, rigid contact account, warming the thermometer by one kelvin requires 10 J while cooling the sample by one kelvin releases 30 J. Their initial temperature difference is 40 K. At equilibrium the thermometer rises 30 K and the sample falls 10 K. Both reach 330 K, so the final reading differs from the sample's original temperature.
The disturbance follows from finite capacities, not a defective thermometer scale. Check the transfers separately. The thermometer gains 10 times 30, or 300 J, and the sample loses 30 times 10, also 300 J. If the thermometer capacity were instead 1 J/K, the equilibrium temperature would be 10,500 divided by 31, about 338.71 K. Its smaller energy demand causes a smaller sample change. Reducing that disturbance does not remove the separate need for the thermometer to reach equilibrium. These are ideal contact examples, not instructions for obtaining an accurate measurement with a particular instrument or material.
A finite thermometer changes the sample it joins
The vertical scale is temperature from 300 to 340 K, not energy or time. Lines connect initial and final equilibrium values for each body; they do not specify a time history. Capacities stay constant and the pair is isolated with negligible work.
A constant reading can also occur without thermal equilibrium. Picture a layer with its left face held at 340 K and its right face at 300 K. Once its temperature profile is steady, a point halfway through a uniform layer can remain at 320 K while energy continually passes through. Steady means that the local state no longer changes with time. Equilibrium would require the absence of the temperature differences driving that conduction. A stationary profile alone therefore cannot establish that the entire layer has one temperature.
Trace the two profiles in the figure before reading their endpoint labels. For the steady layer, compare neighboring positions rather than comparing one position at two times. For an equilibrated isolated layer, compare both positions and times. The first comparison reveals a spatial gradient even when the second shows no change. During a transient, different parts may also warm at different rates. A single sample temperature is useful only when internal differences are negligible for the question, or when the reading is explicitly identified as a local value.
A steady gradient differs from uniform thermal equilibrium
One-dimensional schematic profiles on identical axes. Position is the fraction of layer thickness, not metres. The steady layer has constant conductivity and maintained face temperatures. The uniform profile illustrates a separate equilibrium state at 320 K.
Work through the mechanism
Convert a Celsius reading
t is the numerical Celsius reading. T is absolute temperature. Differences have the same size on these scales, but temperature ratios require kelvin.
Compare temperatures from the same absolute zero
Bar heights encode kelvin above zero. Celsius readings have an offset and cannot be used for these ratios. The bars show temperature, not total internal energy or an actual thermometer design.
Reading
Absolute temperature
What can be compared
20 °C
293.15 K
A temperature, not an energy inventory
40 °C
313.15 K
About 1.068 times 293.15 K
300 K to 600 K
Absolute ratio 2
Twice mean translational energy for the classical ideal gas
Temperature ratios require kelvin and a stated physical relationship.
Separate an absolute scale from a temperature difference
The absolute-temperature bars start at zero kelvin. Their lengths can therefore represent kelvin ratios without hiding an offset. Celsius uses the same interval size but places zero elsewhere. To see what this changes, compare 25 °C with 75 °C. The difference is 50 °C, equivalently 50 K. Their absolute temperatures are 298.15 K and 348.15 K, so the ratio is about 1.168, not three. Adding a fixed offset preserves subtraction but changes division. A Celsius ratio can even change sign across zero Celsius while both absolute temperatures remain positive. That algebra explains why temperature differences and temperature ratios require different treatment.
In a classical dilute monatomic ideal gas, mean translational kinetic energy per particle is proportional to absolute temperature. Each particle moves in three spatial directions, and the classical average contributes one half of the Boltzmann constant times temperature for each direction. Together these give three halves of that constant times temperature. Consequently, the 25 °C to 75 °C change raises the mean translational energy by about 16.8%. This proportionality concerns an average over molecular motion, not the energy of every particle at every instant. Collisions continually redistribute energy among particles even when the equilibrium temperature and the average stay fixed.
The same interval survives a change of temperature zero
Aligned positions refer to the same temperatures. Horizontal distances encode a linear temperature scale. Celsius zero is 273.15 K; kelvin zero lies outside this cropped 250–375 K interval.
Amount still matters for the total. Take two samples of the same classical monatomic ideal gas at 300 K, with twice as many particles in the second. Each has the same mean translational energy per particle, but the second has twice the total translational energy. Contrast that with one fixed sample warmed from 300 K to 600 K. Both its per-particle mean and its total translational energy double. The particle diagram separates these operations by changing the number of representative groups in one case and the mean energy in the other.
Use that diagram as a prediction task. Cover the total-energy labels and compare the sample at 600 K containing one group with the sample at 300 K containing two groups. Multiplying amount by mean energy gives the same total in this idealized example. Yet placing them in thermal contact still drives energy from the 600 K sample toward the 300 K sample. Equal initial totals do not create thermal equilibrium. The direction of contact transfer depends on temperature, while the eventual changes also depend on their capacities.
Particle amount and mean energy multiply to give a total
Each circle represents an equal group containing N particles, not one molecule. E is the mean translational energy per particle at 300 K. Counts and labels encode amount and energy; circle sizes are schematic. Same classical monatomic ideal-gas species throughout.
The microscopic statement needs its model limits. Internal energy in a real material can include interactions and molecular rotation or vibration as well as translation. Quantum energy spacings can prevent classical modes from taking the predicted average energy at low temperature. Thus a temperature ratio is not automatically a ratio of every material's total internal energy. Zero kelvin is the absolute scale's endpoint, not permission to extend a classical constant-capacity formula indefinitely or to assert that every kind of microscopic motion has disappeared there.
Separate temperature from energy and uniformity
A large warm object can contain more internal energy than a small hot one, relative to a shared reference. Temperature is intensive: doubling an otherwise identical sample does not double its temperature. Internal energy is extensive: combining two identical noninteracting samples doubles their total internal energy. Internal energy includes microscopic motion and interactions; it depends on amount, material, phase, and state. Even at the same temperature, two materials need not have the same internal energy per kilogram. Heat is energy in transit because of a temperature difference, not a substance stored inside an object.
Thermal equilibrium is only one part of full equilibrium. Pressure differences can drive motion, and composition differences can drive diffusion or reaction even when temperatures match. During a rapid expansion, a single pressure or temperature may not describe the whole gas. Quantum effects also matter near absolute zero; do not extrapolate the simple gas model to claim that all microscopic motion ceases there.
Compare energy changes using an explicit reference
Saying that one body contains more internal energy needs a reference as well as a material model. A useful comparison is how much its internal energy changes when it warms from an agreed initial state. Suppose a hypothetical solid has constant specific capacity 500 J per kilogram per kelvin and negligible expansion work. For 2 kg warmed from 300 K to 330 K, multiply 2 by 500 by 30 to obtain a 30,000 J increase. This calculation gives a change, not an absolute inventory of every microscopic energy. Choosing a different energy zero would change the quoted inventory but would leave this warming difference unchanged.
Compare a 1 kg sample of that same solid warmed from the same 300 K reference to 350 K. Its increase is 1 times 500 times 50, or 25,000 J. The cooler 2 kg sample therefore has the larger energy increase relative to the stated reference. This is a precise version of the large-warm versus small-hot distinction. It does not imply that equal kilograms of another substance would follow the same numbers. Its capacity, phase, and interaction energy could differ even at the same temperature. The comparison works because the two stated samples share a material, phase, and reference temperature.
Capacity can be stated for a whole sample, per kilogram, or per mole. A mole specifies an amount containing a fixed number of particles, rather than a universal mass. Multiplying specific capacity by mass gives whole-sample capacity. Multiplying molar capacity by amount in moles does the same. For example, a hypothetical 2 mol sample with molar capacity 25 J per mole per kelvin has total capacity 50 J/K. A 10 K rise then corresponds to 500 J under the stated negligible-work, constant-capacity assumptions.
A cooler larger sample can have the larger warming energy
Bars share an energy-change scale beginning at zero. Both hypothetical samples use c = 500 J/(kg K), reference temperature 300 K, one solid phase, and negligible work. These are changes relative to that reference, not absolute energy inventories.
Test the distinction on the capacity strips below. Both strips cover a 10 K rise, but one represents one copy of a sample and the other represents two identical noninteracting copies. Each copy needs the same energy increase, so the combined energy change doubles while the temperature change stays fixed. Now imagine splitting the doubled sample back into its two copies without changing their states. Each retains the common temperature. Temperature is intensive because it does not scale with the number of identical copies; capacity and internal energy do scale.
Finally, decide what extra information each comparison requires. Equal temperatures establish thermal equilibrium for ordinary contact, but cannot establish equal energy changes without amounts and capacities. Equal energy changes cannot establish equal temperatures without initial states and a material relation. A body at one uniform temperature may still have mechanical or chemical reasons to change. For the next lesson, retain the boundary and the endpoints, then ask which transfers supplied the internal-energy change. That question requires an energy balance, not another conversion between Celsius and kelvin.
Two copies double capacity without doubling temperature rise
Each equal rectangle represents one identical hypothetical sample of capacity 50 J/K warmed by 10 K. Rectangles encode equal energy increments, not physical volume. The second row contains two independent copies with additive capacities.
Exercises
Audit a temperature ratio
A claim says that gas at 40 °C has twice the mean translational kinetic energy per molecule of the same ideal gas at 20 °C. Is that right?
Hint
Convert both readings before taking the ratio.
Worked answer
Step
Result
Convert the cooler gas
20 + 273.15 = 293.15 K
Convert the hotter gas
40 + 273.15 = 313.15 K
Take the absolute ratio
313.15 / 293.15 ≈ 1.068
Changing a temperature scale’s zero changes numerical ratios.
No. The temperatures are 313.15 K and 293.15 K. Their ratio is about 1.068, so the mean translational energy rises by about 6.8%, not 100%, within the classical ideal-gas model.
Move the boundary
Two blocks exchange 500 J inside a sealed, insulated, rigid box. Does the combined system receive 500 J of heat from outside?
Hint
Track the same transfer across each possible boundary.
Worked answer
System
Heat across its boundary
Receiving block
+500 J
Giving block
−500 J
Both blocks together
0 J from outside
The system boundary determines which transfers are external.
No. One block receives 500 J and the other loses 500 J. For the combined contents, those are internal transfers. No heat enters across the insulated outer boundary and the rigid boundary does no expansion work. The combined internal energy stays constant even while its internal temperatures change.
Sources and further reading
These optional references support the stated laws and model assumptions. The numerical examples and diagrams are authored teaching models, not experimental measurements or engineering specifications. Selected relevant passages were reviewed, not entire textbooks or a comprehensive research literature.