Its row in the atlas →

Wheeler's geons and quantum foam (1955–1962)

Where the word "wormhole" came from: Wheeler's programme of building matter, charge and multiply connected space out of nothing but geometry and fields.

Kind: physics · Loophole: L3 · Standing: K3 · Bill: B-none · Last reviewed: 2026-09-12

The claim

Wheeler's geometrodynamics asked whether everything could be geometry. A geon, Phys. Rev. 97, 511 (1955), is a lump of electromagnetic radiation held together by its own gravity for a time long compared with its period: "a completely classical, divergence-free, self-consistent picture of the Newtonian concept of body" over masses from about $10^{39}$ g to $10^{57}$ g [HIGH] S1. It is mass without mass: the gravitational mass comes entirely from the energy of the field, with no material source. Misner and Wheeler, Ann. Phys. 2, 525 (1957), then added charge without charge (the rotating case, the Kerr interior passage, was treated by Wheeler's group as a wormhole of the same family): electric field lines threading a handle in a multiply connected space look, from outside, like a positive charge at one mouth and a negative charge at the other, with no charge anywhere. They gave the handle its name, "wormhole" [HIGH] S2. Wheeler's companion paper, Ann. Phys. 2, 604 (1957), argued that at the Planck scale the vacuum should be a quantum foam in which such handles form and dissolve continually, so that multiply connected topology is not exotic but the default texture of space [HIGH] S2.

As a proposal about faster-than-light travel the claim is deliberately modest and, in its strongest form, is a question rather than an assertion: if space is multiply connected, then between two points there are alternative routes, and "one may ask if a signal traveling at the speed of light along one route could be outpaced by a signal which has traveled a much shorter path through a handle" (Fuller and Wheeler 1962) [HIGH] S1. Wheeler's own group asked the question and answered it in the negative for the Schwarzschild handle. The programme's lasting contribution to L3 is the vocabulary, the foam as a hypothetical source of primordial wormholes, and the first clear statement that the causality question is the one a wormhole must answer.

Origin and lineage

The two-sheeted Einstein–Rosen bridge is the ancestor. Wheeler's Geons (1955) is a self-consistent solution programme, not a wormhole paper; the wormhole enters with Misner and Wheeler (1957), where the Reissner–Nordström throat carries electric flux. Misner's "Wormhole initial conditions", Phys. Rev. 118, 1110 (1960), gives exact multiply connected initial data for the vacuum equations [HIGH] S2. Graves and Brill, Phys. Rev. 120, 1507 (1960), follow the charged throat in time and find it pulsates rather than pinching off, cushioned by the electric field, but always behind horizons [HIGH] S2. Fuller and Wheeler (1962) settle the causality question for the Schwarzschild case. The programme ended for want of an observational handle, and Wheeler's students carried the wormhole forward in two directions: Hawking's Euclidean wormholes in quantum gravity, and Thorne's traversable class, Morris and Thorne 1988, whose proposed source for a wormhole is exactly Wheeler's foam, from which an advanced civilisation might pull a Planck-scale handle and enlarge it. Maldacena, Milekhin and Popov open their 2018 paper with a 1966 Wheeler drawing of a wormhole threaded by magnetic field, and their construction is a charge-without-charge wormhole made traversable; see Maldacena, Milekhin and Popov 2018.

The mechanism

Three separate pieces of mathematics live under this heading.

The geon. Wheeler takes the coupled Einstein–Maxwell system and looks for a spherically symmetric configuration of high-frequency electromagnetic waves confined by their own gravitational field. In the geometric-optics limit the waves are a thin shell of light rays orbiting near the radius where photons circle a mass $M$, and the self-consistent metric is found by solving the field equations with the averaged stress-energy of the radiation as source. The characteristic size is set by the gravitational radius, and the leakage rate is set by barrier penetration through the effective refractive-index barrier the geometry creates [HIGH] S1 (abstract). The result is an approximate, long-lived, unstable solution, not an exact one: the instability of the idealised spherical geon "relative to pairing of light rays" has a time scale long compared with the vibration period [HIGH] S1. Later work (Ernst 1957; Brill and Hartle 1964; Anderson and Brill 1997) refined the construction and its instability [MED] S2 (Lobo's survey).

Charge without charge. Take the Reissner–Nordström metric,

$$ds^2 = -\left(1 - \frac{2m}{r} + \frac{q^2}{r^2}\right)dt^2 + \left(1 - \frac{2m}{r} + \frac{q^2}{r^2}\right)^{-1}dr^2 + r^2\,d\Omega^2 ,$$

and read the two exterior regions as two mouths in one space. The Maxwell field is source-free everywhere, $\nabla_\mu F^{\mu\nu} = 0$, yet the flux through a sphere around one mouth is $4\pi q$ and through a sphere around the other is $-4\pi q$, so Gauss's law reports charges at the mouths that are nowhere in the field equations [HIGH] S2. This is the electromagnetic reading of the throat, and it is the seed of every later "wormhole threaded by flux" construction. Graves and Brill showed that the throat of the charged wormhole contracts to a minimum and re-expands after a finite proper time rather than pinching off, but the inner horizon at $r_- = m - \sqrt{m^2 - q^2}$ separates the two exteriors causally, and the throat is not traversable in the sense of admitting a round trip [HIGH] S2.

Foam. Wheeler's argument is dimensional. Quantum fluctuations of the metric over a region of size $L$ are of order $\Delta g \sim \ell_P / L$ with $\ell_P = \sqrt{\hbar G / c^3} \approx 1.6 \times 10^{-35}$ m, so at $L \sim \ell_P$ the fluctuations are of order unity and the topology itself should fluctuate [HIGH] S2. No equation predicts the density or lifetime of foam wormholes; the picture has no consensus mathematics and is graded accordingly. Morris and Thorne's proposed engineering step, enlarging a foam wormhole to macroscopic size, has never been given a mechanism [HIGH] S2 (Lemos, Lobo and Oliveira §1.3: "The engineering work was left to an absurdly advanced civilization").

The causality check. Fuller and Wheeler's calculation is on the Einstein–Rosen bridge page: the Schwarzschild throat pinches off in finite time, no signal can be sent between the two non-catastrophic regions, and causality is preserved [HIGH] S1. For the charged wormhole the same conclusion follows from the horizons; the pulsation of Graves and Brill happens inside them.

The headline number is that the wormholes of this programme are either Planck-scale ($10^{-35}$ m, foam) or hidden behind horizons (geons and charged throats). Neither offers a route.

What it costs

Nothing exotic, and no shortcut results, which is why the bill is B-none. Geons are ordinary electromagnetic radiation plus gravity; charge without charge is a vacuum Einstein–Maxwell solution; the foam is a conjecture about the vacuum. None of them asks for negative energy, and none of them delivers a traversable throat. Wheeler's wormholes obey the energy conditions and are, as STA-2 says they must be, non-traversable.

The unpaid costs belong to the descendants. To make a foam wormhole into a route you need to stop it collapsing (the exotic matter of Morris and Thorne 1988), enlarge it by some thirty orders of magnitude with no known mechanism, and then transport a mouth at sublight speed to where you want to go.

Constraint scoring

ConstraintVerdictNote
CAU-1SATISFIESFuller and Wheeler asked exactly this question for the Schwarzschild handle and showed no signal is outpaced; the charged handle hides behind horizons.
CAU-2SATISFIESNone of the geon, charged-throat or foam constructions produces closed timelike curves; nothing appears for chronology protection to forbid.
CAU-3SATISFIESThe conversion to a time machine needs a wormhole that can be maintained and crossed; these cannot be, so the assembly step is never reached.
CAU-4N/ANo faster-than-light sector exists, so no preferred frame needs to be chosen.
CAU-5N/ANo entanglement enters; the programme is classical geometry plus a conjecture about vacuum fluctuations.
ENE-1SATISFIESElectromagnetic radiation and vacuum obey the null and weak energy conditions everywhere in these solutions.
ENE-2SATISFIESWith the energy conditions obeyed, the Olum and Gao–Wald theorems forbid a time advance, and the pinch-off and horizons deliver exactly that.
ENE-3N/ANo negative energy is invoked, so the quantum inequalities do not bound anything here.
ENE-4N/AThe Van Den Broeck pocket is a warp-bubble device with no counterpart.
ENE-5N/ANot a Natário-class warp metric.
ENE-6N/ANo negative energy is required, so the Casimir floor is not a comparison.
ENE-7SATISFIESMultiply connected vacuum and electromagnetic solutions obey the ANEC everywhere, so the theorem holds and delivers what Wheeler found: the handles pinch off and no signal beats the outside route. Charge without charge is topology that censorship permits because it is not a shortcut.
CON-1N/AWarp-wall horizons are a different object; the black-hole horizons here are scored under causality.
CON-2N/ANothing is laid along a route; foam wormholes, if they exist, are wherever they happen to be.
CON-3N/ANo Krasnikov-type pre-laid metric is involved.
STA-1N/AThe instability concerns superluminal bubble walls; the geon's own instability (light-ray pairing) is a different effect and is noted in the text.
STA-2SATISFIESThese throats contain no exotic matter, and as STA-2 predicts they are not traversable: Schwarzschild pinches off, Reissner–Nordström pulsates behind horizons.
STA-3N/ANo chronology horizon forms because no time machine can be built from a non-traversable handle.
HAZ-1N/AParticle sweep-up is a superluminal bubble effect with no analogue.
HAZ-2N/AThe interior Hawking bath is a warp-bubble effect with no analogue.
HAZ-3N/AThe tidal bound prices a human-traversable throat; a Planck-scale foam handle is $10^{-35}$ m across and the macroscopic ones are not traversable, so there is nothing to price.
LOR-1SATISFIESNothing massive is pushed to $c$; the geon's contents are light, moving at $c$, held by gravity.
LOR-2N/ANo tachyons appear.
LOR-3N/ANo group- or phase-velocity claim is made.
LOR-4N/ANo Scharnhorst-type effect is involved.
WRP-1N/ANot a warp drive.
WRP-2N/AThe shell classification of warp spacetimes does not apply.
WRP-3N/AThe positive-energy warp argument does not bear on a wormhole.
MAN-1N/AMultiply connected topology in four dimensions is not an extra dimension; MAN-1 concerns bulk shortcuts, which are not invoked.

Status of the argument

1955–1957: the geon programme and the wormhole are introduced; the "charge without charge" reading of the Reissner–Nordström throat is exact and uncontested [HIGH] S2.

1960: Misner gives exact multiply connected vacuum initial data; Graves and Brill find the pulsating charged throat [HIGH] S2.

1962: Fuller and Wheeler close the causality question for the Schwarzschild handle; no peer-reviewed challenge as of 2026-09-12 [HIGH] S1.

1960s–1990s: geon studies continue (Brill and Hartle 1964; Komar 1965; Anderson and Brill 1997) and confirm that geons are approximate and unstable; the programme of deriving all of physics from geometry is abandoned, Wheeler included [MED] S2 (Lobo 2007, §I: "due to the extremely ambitious program and the lack of experimental evidence soon died out").

1988: Morris and Thorne adopt the foam as the proposed origin of a traversable wormhole and leave the enlargement to an "absurdly advanced civilization"; no mechanism for enlargement has been proposed in the peer-reviewed literature as of 2026-09-12 [HIGH] S2.

The foam itself remains a picture rather than a theory. Loop quantum gravity and string theory each have their own account of Planck-scale geometry, and neither reproduces Wheeler's handles as a population of usable wormholes.

Sources