Overview
Every version of this story has the same three acts. An inventor working outside the institutions builds a device that produces more energy than it consumes, or draws it from a source nobody else has noticed. There is a demonstration, usually witnessed, usually poorly documented. Then the machine is destroyed, the notes vanish, the patent is refused, the funding evaporates, or the inventor dies at an inconvenient moment.
The cast changes but the structure does not, and the structure is the point. Individually the cases are unremarkable — inventors fail all the time. Assembled into a series, they look less like a run of disappointments than a policy.
The Canon
The roster is reasonably fixed. Nikola Tesla and the Wardenclyffe tower stand at the head of it, with J. P. Morgan's withdrawal of funding recast from a financier declining to pay for something unmarketable into the deliberate strangling of wireless power. Thomas Henry Moray demonstrated a radiant-energy receiver in Salt Lake City in the 1920s that supporters said powered lamps from no visible source, and his apparatus was reportedly damaged and his recognition withheld. Stanley Meyer claimed a cell that ran an engine on water and died suddenly in a restaurant car park in 1998, which his supporters do not treat as a coincidence. Tom Ogle, the hundred-mile-per-gallon carburettor, Eugene Mallove and cold fusion — the list extends in both directions and continues to acquire members.
The modern additions tend to be corporate rather than individual: the electric car recalled and crushed, the battery patent bought and shelved, the efficiency improvement acquired by an oil company and never produced.
The Physics
What unites the canon is that the strong version of the claim requires something no laboratory has ever observed. A device producing net energy from nothing violates the first law of thermodynamics; one extracting useful work from ambient heat without a temperature difference violates the second. These are not provisional findings awaiting a better instrument. They are the most heavily tested generalisations in physics, and every technology that works is consistent with them.
Supporters generally do not dispute the laws directly. They argue that the device draws on something the accounting has not included — zero-point energy, the quantum vacuum, an ambient radiant field — so that the system is open rather than closed and the ledger balances after all. That move keeps the argument alive indefinitely, because it relocates the claim from a measurement anyone could check to a mechanism nobody can specify.
The Patent Office
One detail is regularly presented as evidence of institutional obstruction, and it is worth stating accurately because the truth is more interesting than the version in circulation.
The US Patent and Trademark Office does treat perpetual motion claims differently from everything else. Its examination guidance provides that where operability is questioned, an examiner may require a working model — a demand made almost nowhere else in modern patent practice, since applications are otherwise assessed on paper. Applications of this kind are routinely rejected for lack of utility.
To the suppression reading, this is the mechanism itself: a rule that singles out precisely the inventions in question and imposes a standard no one else has to meet. To the patent office it is the accumulated response to more than a century of applications that did not work, and a practical alternative to examining each new one from first principles.
Both descriptions are accurate. They are the same rule.
What the Pattern Actually Contains
The individual episodes tend to dissolve on inspection in ways that do not require anyone to have been silenced. Morgan stopped funding Wardenclyffe because Tesla had changed the project from transatlantic telegraphy — which Morgan wanted — into broadcast power, which had no billing mechanism. Moray's demonstrations were never reproduced under controlled conditions, and the patent he did secure was for an electrotherapeutic device rather than an energy receiver. Meyer was found by an Ohio court in 1996 to have committed fraud against investors, two years before his death, which was recorded as a cerebral aneurysm.
What survives is the sociological observation, and it is real: the energy industry is enormously capitalised, patents are acquired defensively across every sector, and firms have historically suppressed unfavourable research about their own products. A public that has watched that happen with tobacco and with emissions data is not being irrational in suspecting it might happen with an engine.
The suppression theory takes that reasonable suspicion and attaches it to a class of devices that would have to work first. Its durability comes from the fact that the first half of the argument is sound, and the second half is where it stops being checkable.