A ceiling set by temperature alone
In 1824, the French engineer Sadi Carnot published a slim book, Reflections on the Motive Power of Fire, arguing that every heat engine, regardless of its design or the substance it runs on, has a maximum possible efficiency set purely by the temperatures of the hot source it draws heat from and the cold sink it dumps waste heat into. That maximum, later written as one minus the ratio of the cold temperature to the hot temperature, both measured on an absolute scale, means no steam engine, no matter how cleverly built, can turn all the heat it consumes into useful work; some fraction is always lost to the cold reservoir as a matter of principle, not engineering shortfall. Carnot reached this conclusion using a reversible, idealised cycle rather than any specific machine, which is why the result applies to every heat engine ever built or conceivable, including engines that did not yet exist in 1824.
A reversible cycle, and a proof by contradiction
Carnot’s argument worked by imagining a perfectly reversible engine, one that could be run forward to produce work or backward as a refrigerator without any loss, cycling a gas through four stages: expansion at a constant hot temperature while absorbing heat, further expansion with no heat exchange at all, compression at a constant cold temperature while releasing waste heat, and a final compression back to the starting point. He showed that no engine operating between the same two temperatures could beat this reversible cycle’s efficiency, using a proof by contradiction: an engine that did beat it could be paired with a reversed Carnot engine to move heat from cold to hot with no external input, which nothing in nature allows. Carnot reasoned this out using the caloric theory, then current, which treated heat as an indestructible fluid rather than convertible energy, a flaw that did not undermine his conclusion about efficiency limits.
Rescued from a wrong theory of heat
Carnot’s efficiency limit and the impossibility of exceeding it have never been overturned; they were, instead, generalised. Rudolf Clausius and William Thomson, later Lord Kelvin, rebuilt Carnot’s reasoning in the 1850s on top of the newer understanding that heat and work are interchangeable forms of energy, and in doing so produced the second law of thermodynamics in something close to its modern form: Clausius’s statement that heat cannot pass unaided from a colder body to a hotter one, and Kelvin’s equivalent statement that no cyclic device can turn heat entirely into work from a single reservoir. Both are now understood to be mathematically equivalent, and both descend directly from Carnot’s original argument about heat engines. The concept of entropy, a quantity that never decreases in an isolated system undergoing real processes, emerged from formalising exactly why Carnot’s limit could not be beaten.
Why real engines fall further short
The theoretical limit is one thing; reaching anywhere near it in a real machine is another. A true Carnot cycle requires each stage to happen infinitely slowly to remain reversible, which means an engine actually built to the Carnot ideal would produce zero net power, since it never finishes a cycle in any finite time. Real engines running at a practical speed fall further below the theoretical maximum, with efficiency dropping toward roughly the square root of the Carnot value once realistic speeds are accounted for, and material limits compound the gap further: early steam locomotives without condensers managed only a few per cent efficiency, while nuclear plants have been constrained by how much heat their fuel cladding can tolerate before failing, pushing efficiency up only as better alloys allowed higher operating temperatures.
Ruling out free energy from the environment
The second law’s reach extends well past engine design. Because it forbids any device from converting heat entirely into work without a temperature difference, it rules out an entire category of proposed inventions known as perpetual motion machines of the second kind, devices that would draw on the enormous ambient internal energy of the environment as free power, and it does so on principle rather than by noting that no one has yet built one. The same law, through the concept of entropy always increasing in real processes, is also usually cited as the physical basis for the arrow of time: broken cups do not reassemble and mixed gases do not separate on their own, not because some other law forbids it outright but because doing so would require entropy to decrease, which the second law does not permit.
A law that may explain time’s direction
This is dense material but foundational, and worth the time for how directly it connects a nineteenth-century engineering question, how much better can a steam engine get, to some of the largest ideas in physics, including why time seems to run in only one direction. The route from Carnot’s flawed but structurally sound 1824 argument to Clausius and Kelvin’s later, corrected formulations is a genuinely instructive example of a wrong theory producing a right answer, and it rewards readers willing to sit with the logic of the reversible cycle rather than skipping to the formula. It is less rewarding for anyone hoping for a story of dramatic discovery; the second law’s development was gradual, built by several hands over decades, and its significance lies in the reasoning more than in any single moment.