The Photon from First Principles:
Structure, Propagation, and Interaction
in the \(\varepsilon_0\mu_0\) Medium

D. J. Hallman — SCG@azfn.com
2026 (v3) — DOI: 10.5281/zenodo.21634358
CC BY 4.0 © 2026 D. J. Hallman

1. The Medium

Space is a physical medium.

This is not a philosophical assertion. It is the direct reading of two quantities that appear in every electromagnetic equation ever written: the permittivity \(\varepsilon_0\) and the permeability \(\mu_0\). These are not abstract constants. They are measurable properties of space itself — properties whose values can be confirmed independently from any electromagnetic measurement and whose ratio yields the impedance of free space:

Impedance is not a property of nothing. It is the ratio of a field’s resistance to propagation — a dimensioned, stable, measurable quantity confirmed to six significant figures in every antenna calculation, every waveguide design, and every transmission line ever built. \(Z_0\) was not postulated. It was measured. The medium was always there.

The product \(\varepsilon_0\mu_0\) governs how fast electromagnetic disturbances propagate:

\(c\) is not a speed limit imposed from outside. It is the recovery rate of the medium — the rate at which a disturbed region of space returns to its equilibrium state.

1.1 The Physical Inventory

Energy is dispositioned space — a local departure from the equilibrium \(\varepsilon_0\mu_0\) medium.

Mass is confined dispositioned space — energy whose geometry is closed and self-reinforcing. \(E = mc^2\) is not a conversion formula between two different things. It is an identity between two configurations of one thing.

Spatial recovery is the medium’s one rule. It operates everywhere, at every scale, at rate \(c\). Gravity is its gradient. The propagation of light is its cascade.

Charge is not a primitive. It is the topological signature of certain closure geometries — specifically, closures whose rotation axis bears a preferred directional relationship to their field expulsion geometry. The photon has no sustained closure and therefore no charge.

Electromagnetics is not a fundamental category. It is the behavior of electron closures interacting with each other and with the \(\varepsilon_0\mu_0\) medium. Maxwell’s equations describe this behavior correctly and precisely. They are the rules of electron closure dynamics in the medium, not the rules of the medium itself.

1.2 What a Photon Is

A photon is a propagating disturbance in the \(\varepsilon_0\mu_0\) medium. It is not a point. It is not located at a position. It is a process — a transverse oscillation of the medium unfolding through space, sustained by the medium’s own recovery dynamics.

A photon emitted from a moving source does not inherit the source’s velocity. It enters the medium at the moment of emission and propagates at \(c\) regardless of how fast the emitting atom was moving. The emitting atom is a boat on a river; the medium is the river; the photon is a wave the boat generates on the surface. The wave’s speed is the river’s, not the boat’s.

1.3 A Note on the Word “Electromagnetic”

Maxwell identified light as electromagnetic in character. This is correct at the endpoints. In transit, the photon carries no charge, no ratio perturbation, no electromagnetic character. A photon carrying alternating charge sign would curve back on itself in the \(\varepsilon_0\mu_0\) medium. Light travels straight. Therefore no alternating charge. The conclusion is geometric, not inferred: the straightness of light is itself proof that the photon carries no charge in transit.

In transit, the photon is a pure product perturbation of \(\varepsilon_0\mu_0\) — gravitational in character. The electromagnetic spectrum correctly names the endpoints. It correctly labels the frequency ordering of radiation from radio to gamma. What is incorrect is the implication that the propagating entity between the endpoints is electromagnetic in character. The spectrum is photonic throughout.

2. The Photon Is a Sine Wave

2.1 Propagation Is Not a Feature. It Is a Geometric Necessity.

Any energy constrained to propagate at \(c\) through a medium with symmetric recovery cannot propagate longitudinally. Transverse oscillation is not a property of light. It is a geometric necessity of the medium.

The \(\varepsilon_0\mu_0\) medium recovers from any perturbation symmetrically — its recovery pressure acts in all directions equally, with no preferred axis. A purely longitudinal wave would require the recovery to act only along the propagation axis — breaking the medium’s symmetry by selecting one direction. The medium does not do this. The recovery pressure deflects the energy transversely. The deflected energy arcs away from the propagation axis, reaches a maximum transverse displacement, and the medium’s symmetric restoring force returns it to the axis. At the maximum — the apex — the transverse velocity is zero and the energy is at rest in the medium.

The oscillation is not imposed. It is the mandatory geometric consequence of symmetric recovery in a medium that conserves momentum. A purely longitudinal wave in this medium is geometrically impossible. The photon is what symmetric recovery looks like when it propagates.

2.2 The Causal Constraint

Every photon, regardless of frequency, traverses the same distance in the same time. This single fact, applied without exception, fixes the photon’s transverse geometry completely — and it can be reached from two independent directions.

The causal argument. Send an infrared photon and an X-ray photon from the same source toward the same detector a meter away. Both arrive together. Both propagated at \(c\). If both have the same transverse amplitude, the X-ray’s field traces a far longer geometric path. A longer path covered in the same time implies a faster internal coordination speed. But the medium’s recovery rate is \(c\) for all wavelengths; there is no reserve of speed to draw on. The only resolution: the transverse amplitude must shrink as frequency increases, keeping the total arc length — and therefore the causal budget — the same for every wavelength.

The least-action argument. Maupertuis’s principle of least action asks what closure geometry requires no external length scale — no arbitrary ruler imported from outside the oscillation itself. There is exactly one such geometry: the amplitude equal to the oscillation’s own radian length scale. This is the self-referential condition \(\beta = 1\).

2.3 The Arc-Length Invariant

Model the photon’s transverse oscillation as \(E(x) = A\sin(kx)\) with \(k = 2\pi/\lambda\). The geometric arc length over one wavelength is:

The dimensionless arc-to-wavelength ratio is:

where \(E(m)\) is the complete elliptic integral of the second kind. For this ratio to be frequency-independent, \(\beta\) must be a constant. Both arguments above demand the unique self-referential value \(\beta = 1\). Substituting:

The arc traced by one wavelength of the oscillation is 21.6% longer than the forward propagation distance. The exact path this oscillation traces, \(y = \bar{\lambda}\sin(kx)\) with \(\beta = 1\), is a type-2 ellipse — the unique closure curve whose arc-to-chord ratio is \(\gamma_{\mathrm{cause}}\).

2.4 The Reduced Wavelength Was Always the Photon’s Radius

The self-referential condition \(\beta = 1\) fixes the amplitude:

This quantity — the reduced wavelength \(\bar{\lambda} \equiv \lambda/(2\pi) = \hbar/p\) — appears in the de Broglie relation, in the definition of \(\hbar = h/2\pi\), in the near-field to far-field transition of a radiating dipole, and in the Bohr radius. It has been correct in every one of those contexts for a century. What was never identified is what \(\bar{\lambda}\) physically is. It is the photon’s transverse radius, fixed by the self-referential closure condition.

Sustaining this arc at forward speed \(c\) requires an internal coordination speed \(c_{\mathrm{coord}} = \gamma_{\mathrm{cause}} \cdot c \approx 1.216\,c\). This is not a signal speed through spacetime. It is the rate at which causal influence must propagate within the photon’s own field structure to maintain phase coherence along the type-2 ellipse.

2.5 Two Phases Per Cycle

A photon cycle has two physically distinct phases.

The apex. At each displacement peak, the transverse velocity is zero. The sine lobe is at maximum curvature. The apex mass is maximum. At that moment the energy is stationary in the medium. The medium closes around it momentarily, forming mass. The closure is unsustainable because no permanent confinement geometry exists to hold it open. The apex begins to dissolve immediately.

The zero crossing. Between each pair of apexes, the arc straightens. The apex mass releases. That released energy propagates forward as a field disturbance in the \(\varepsilon_0\mu_0\) medium: uncharged, no closure radius, no winding direction, no handedness. The zero-crossing disturbance seeds the next apex. The photon is self-threading by this alternation:

The photon does not propagate because something pushes it. It propagates because each apex dissolution is the push for the next apex. This is why photons traverse cosmological distances without dissipating. The energy is not being spent. It is being cycled.

The energy distribution. The apex mass and the propagation momentum distribute continuously across the cycle as:

where \(\theta\) advances from 0 at the zero crossing to \(\pi/2\) at the apex. At the apex: \(\cos^2(\pi/2) = 0\) — all momentum is zero, apex mass is maximum. At the zero crossing: \(\sin^2(0) = 0\) — all mass is zero, momentum is maximum. The same distribution governs the LC oscillator and the antenna. All three are the \(\varepsilon_0\mu_0\) medium cycling between a confined and a propagating form.

2.6 The Photon as a Particle

A particle is a permanent closure — one where the geometry has closed on itself completely, sustaining confinement indefinitely without requiring continuous propagation to maintain it. The photon is not a particle by this definition. It is a propagating sine wave whose apex momentarily satisfies the mass condition but cannot sustain closure because no permanent confinement geometry exists at that scale.

Wave-particle duality dissolves in this picture. The photon is not sometimes a wave and sometimes a particle. It is always a wave. It exhibits particle-like properties — discreteness, energy quantization, localized interaction — because the wave has a specific geometry with specific coupling conditions. The wave IS the particle IS the wave. There is no duality. There is one geometric object that orthodoxy described from two incomplete angles simultaneously.

3. The Apex, the Zero Crossing, and Self-Threading

3.1 \(E = mc^2\) Is the General Case

\(E = mc^2\) is conventionally presented as a statement about the conversion of mass into energy under special circumstances. This framing has it exactly backward. \(E = mc^2\) is the general case. It states that energy and mass are the same thing in two different geometric configurations. The photon apex is the most direct satisfaction of this equation available in physics.

At the apex, the transverse velocity is zero, the momentum is zero, and the energy is at rest in the medium. That is what \(E = mc^2\) means: energy at rest is mass. The apex satisfies it constitutively. The mathematics requires no interpretation: \(\cos^2(\pi/2) = 0\). All momentum is zero. All velocity is zero. The energy is not slowing down. It is stopped.

The apex does not approximately satisfy \(E = mc^2\). It satisfies it constitutively. A massive particle satisfies \(E = mc^2\) only at rest in a specific inertial frame, relative to some observer. The photon apex satisfies it in the medium itself — at rest relative to the \(\varepsilon_0\mu_0\) field that defines rest absolutely. The photon apex is the stillest mass in the universe.

3.2 The Battery Loses Mass

A radio transmitter running from a battery loses mass. This is not a thought experiment. Every photon emitted carries \(h\nu/c^2\) away from the source. The battery is lighter after transmission than before.

Orthodox physics accepts this numerically while having no physical story for it. The electron falls to a lower orbital and a photon “appears.” Where did the mass go? Into the photon — which is officially massless. That is the contradiction sitting in plain sight: a massless particle carrying away mass from a source that measurably loses it.

In the \(\varepsilon_0\mu_0\) picture the mechanism is immediate. Emission is field abandonment. The electron closure contracts to a lower radius. The geometry it was maintaining propagates outward as the photon. The mass that propagates is the mass of the vacated geometry: \(m = h\nu/c^2\). No contradiction. No massless particle carrying mass. The accounting closes mechanically.

The battery is the experiment. The mass loss is the measurement. The mechanism is field abandonment.

3.3 Why the Apex Forms Mass

A curved path through the \(\varepsilon_0\mu_0\) medium requires centripetal acceleration. Acceleration is locally equivalent to gravity by the equivalence principle. Gravity is a \(\varepsilon_0\mu_0\) density depression. Therefore a curved path through the medium creates a local \(\varepsilon_0\mu_0\) density depression at the point of maximum curvature. A local density depression in the medium is mass.

The photon’s \(\beta = 1\) arc reaches maximum curvature at the apex, where the transverse velocity is zero and the path curves most tightly through the medium. The acceleration at that point is maximum. The local density depression is therefore maximum at the apex. The apex is a mass event — not imposed, not postulated, but required by the geometry of a curved path through a physical medium and the equivalence principle connecting acceleration to gravity.

This is the same root mechanism as the Sagnac effect — which Sagnac himself identified in 1913 as evidence of acceleration through the medium — and as gravitational time dilation, confirmed by Pound and Rebka. The photon apex, the Sagnac phase shift, and gravitational time dilation are three manifestations of the same physical fact: acceleration through the \(\varepsilon_0\mu_0\) medium creates a local density depression. The scale and geometry differ. The mechanism does not.

3.4 How Much Mass

The quantity of mass at the apex is derived from three knowns.

The photon’s energy: \(E = h\nu\), established from the arc-length closure condition of Section 2.

The mass the emitting atom loses: \(\Delta m = h\nu/c^2\), established from energy conservation at emission. Weigh the atom before and after. The difference left with the photon. This follows from \(E = mc^2\) exactly.

The energy distribution across the cycle: \(\sin^2\theta\) for mass, \(\cos^2\theta\) for propagation momentum. At the apex \(\theta = \pi/2\): \(\sin^2(\pi/2) = 1\), \(\cos^2(\pi/2) = 0\). All energy is in mass form. No energy is in propagation form.

These three give the apex mass directly:

This is \(m_{\mathrm{total}}\). The same mass is confirmed by two independent routes: energy conservation at emission, and the \(\sin^2/\cos^2\) distribution evaluated at \(\theta = \pi/2\). Two routes, same answer, genuine cross-check. The orthodox photon mass-equivalent \(h\nu/c^2\) is not an approximation of something larger. It is the exact mass the photon carries, derived from first principles without postulate.

3.5 The Apex Is Mass at Rest in the Medium

The \(\sin^2/\cos^2\) distribution states the apex condition exactly. At \(\theta = \pi/2\):

Mass is maximum. Momentum is exactly zero. Not approximately zero. Not minimized. Zero. The forward propagation is a chain reaction, not a flow through the apex. The apex dissolves and the dissolution seeds the next event forward. The energy at the apex does not travel — it stops, becomes mass, and the effect propagates.

In the LC oscillator, energy cycles between the electric field of the capacitor (apex — stored potential, zero current) and the magnetic field of the inductor (zero crossing — maximum current, zero potential). In the antenna, energy cycles between the charge distribution at the tips (apex) and the current maximum at the feed point (zero crossing). In the photon, energy cycles between the apex mass and the propagation momentum at the zero crossing. Three systems, three scales, one law.

3.6 Between Apexes: the Coordination Speed

At the zero crossing, \(\sin^2(0) = 0\) — apex mass is zero, momentum is maximum. The energy equivalent between apexes propagates at \(c_{\mathrm{coord}} = \gamma_{\mathrm{cause}} \cdot c \approx 1.216\,c\).

This is not a violation of any limit. \(c\) is the cycle-averaged propagation rate, not an instantaneous speed limit within the cycle. \(c\) is not imposed from outside. It is the geometric average of absolute stillness and superluminal energy propagation, locked by the \(\beta = 1\) closure condition of the medium itself.

3.7 An Independent Refutation of Kinematic Time Dilation

Kinematic time dilation requires photons to be massless point particles on null worldlines. This is load-bearing for the entire KTD framework.

Conservation of energy at emission proves the photon has mass \(h\nu/c^2\). Weigh the emitting atom before and after emission. The difference left with the photon. No SCG geometry required for this proof. It follows from energy conservation alone.

The apex geometry adds a second, independent refutation. The photon’s mass is at absolute rest in the \(\varepsilon_0\mu_0\) medium — the medium that defines rest absolutely. KTD applies to moving masses. The photon’s mass is never moving. KTD called the stillest mass in the universe massless. Both conclusions fail.

Two independent lines of evidence converge on the same conclusion: the forensic examination of the kinematic term established the failure of KTD from frequency shift evidence; this paper establishes an independent conservation-based and geometric refutation from photon structure alone. Neither argument requires the other. Both point the same direction.

\(h\) Is Maxwell’s Constant

4.1 What Planck Measured

In 1900, Max Planck introduced a constant \(h\) to fit the blackbody spectrum. He described it as a mathematical device, a “fortunate guess,” and spent years afterward trying to derive it from classical theory. He never succeeded. The constant was accepted as a primitive of nature — a postulate without a mechanism.

It is not a primitive. It is a geometric consequence of the arc-length closure condition derived in Section 2, and it was already implicit in Maxwell’s equations of 1865. Maxwell contained \(h\). He did not know it. Planck measured a geometric cycling cost in 1900 without knowing that was what he was doing. \(h\) is Maxwell’s constant, identified thirty-five years late.

4.2 The Derivation

The arc-length closure condition fixes the photon’s transverse radius at \(r_{\mathrm{ph}} = \lambda/(2\pi)\). No \(h\) input. No quantum postulate. Pure geometry from Section 2.

A photon of wavelength \(\lambda\) carries momentum \(p = E/c\) and geometric radius \(r_{\mathrm{ph}} = \lambda/2\pi\). Their product:

Every step follows from the photon’s geometry:

\(h\) is constant because \(r_{\mathrm{ph}}/\lambda = 1/(2\pi)\) is constant — the same ratio, for every photon, at every frequency, everywhere in the \(\varepsilon_0\mu_0\) medium. The universality of \(h\) is the universality of the closure condition itself.

4.3 Quantization Is Geometry

Discrete energy levels in atomic spectra require no quantization axiom. The closure condition permits only specific transverse radii. Only specific radii produce stable geometric coupling between a photon and a target closure geometry. Only stable couplings produce observed spectral lines. The quantum ladder is a radius ladder. No axiom required. The geometry enforces it.

4.4 Antenna and Atom Are the Same Mechanism

The boundary between classical electromagnetic theory and quantum theory of light does not exist. In both atomic emission and antenna emission, an electron closure reconfigures and ejects a propagating perturbation whose sine wave geometry is set by the source oscillation scale. The atomic transition produces a photon whose apex curvature matches the electron closure geometry. The antenna produces a photon whose apex curvature matches the oscillation geometry of the current. Same sine wave. Same apex mass at each apex. Different scale. One mechanism across all frequencies.

4.5 Predictions

Transition duration. Because the photon’s transverse radius \(r_{\mathrm{ph}} = \lambda/(2\pi)\) is physically real, the atomic transition that produces a photon takes finite time: \(\Delta t = \Delta r/c\), where \(\Delta r = r_{\mathrm{ph}}^{\rm upper} - r_{\mathrm{ph}}^{\rm lower}\) is the difference between the photon radii corresponding to the upper and lower closure states. For hydrogen’s Lyman-alpha transition (\(\lambda = 121.6\) nm), \(\Delta t \approx 0.53\) attoseconds. The Standard Model assigns zero duration to all transitions. Attosecond spectroscopy is the test.

Density-dependent \(h_{\mathrm{eff}}\). In regions where \(\varepsilon_0\mu_0\) deviates from its equilibrium value, the local propagation speed shifts, modifying the relationship between frequency, wavelength, and radius. The effective local Planck constant becomes \(h_{\mathrm{eff}} = h_0\sqrt{(\varepsilon_0\mu_0)_0 / (\varepsilon_0\mu_0)(x)}\). This is confirmed indirectly by gravitational time dilation — itself a density-dependent \(h_{\mathrm{eff}}\) operating continuously in every GPS satellite.

5. The Photon Carries No Charge

5.1 The Question

Maxwell identified light as electromagnetic in character — oscillating electric and magnetic fields propagating through space. This identification is correct at the endpoints. It is not correct in transit.

Three independent proofs establish that the photon carries no charge in transit. Each stands alone. Together they form a closed case.

5.2 Proof 1: Straight-Line Propagation

A photon carrying alternating charge — positive at one apex, negative at the other — would produce alternating field perturbations in the \(\varepsilon_0\mu_0\) medium. Each half-cycle’s field contribution would be unwound by the opposite half-cycle. No net propagating wave could escape a source carrying alternating charge. Light escapes. Light propagates indefinitely. Light travels in a straight line at \(c\).

Therefore no alternating charge. The product-only conclusion is not merely consistent with observation. It is geometrically required by the propagation of light itself. In transit, the photon is a pure product perturbation of \(\varepsilon_0\mu_0\) — gravitational in character.

5.3 Proof 2: Malus’s Law

In 1809, Étienne-Louis Malus measured the transmission of reflected light through a calcite crystal as a function of angle. He found:

Transmission is zero at \(\theta = 90°\). No residual transmission. No additional coupling term. This has been confirmed in every polarizer experiment conducted in the two centuries since, including at single-photon counting rates.

If the photon carried alternating charge, a charge-independent coupling term would appear in every polarizer interaction. Transmission would never reach exactly zero. But Malus’s Law holds exactly. The exactness of \(\cos^2\theta\) to zero is not a numerical curiosity. It is the geometric signature of a product-only disturbance coupling through oscillation plane projection alone. Any ratio perturbation, however small, would leave a residual. There is none. Malus measured this in 1809 — accidentally proving the photon carries no charge 116 years before the photon was named.

5.4 Proof 3: Charge Is Closure Topology

Charge is not a primitive. It is the topological signature of stable \(\varepsilon_0\mu_0\) field closures whose rotation axis bears a preferred directional relationship to their field expulsion geometry. Charge does not propagate. What propagates between charged objects is always the field disturbance generated by their geometry — carrying no handedness, no closure radius, no winding direction.

The photon has no sustained closure. At the apex the medium forms an unsustainable transient closure that dissolves immediately. No sustained closure means no committed handedness. No committed handedness means no charge.

5.5 What This Means for Radio Electronics

Electrical circuits operate on the ratio face of \(\varepsilon_0\mu_0\) — charge separation, current flow, impedance, voltage. An antenna converts ratio perturbations into product perturbations and back again. The oscillating current in the transmitting antenna is ratio face. The radiated photon is product face. The receiving antenna converts the arriving product perturbation back into ratio face.

Marconi built ratio-to-product transducers in 1895 without knowing what he was transducing. The physics was always there. The label was wrong.

6. The Fine-Structure Constant from Arc Geometry

6.1 137 Is Not a Mystery

The fine-structure constant \(\alpha \approx 1/137\) has been called the greatest mystery in physics. Feynman described it as a magic number that no one understands. It is not a mystery. It is the geometric coupling efficiency between the photon’s complete arc geometry and the electron’s circular closure geometry. Once the photon’s three field components are correctly identified, the number 137 falls out of pure geometry with no empirical input and no adjustable parameters.

6.2 Three Components of the Photon’s Arc

Component 1: The primary transverse oscillation. The arc-length ratio \(\gamma_{\mathrm{cause}} = \frac{2}{\pi}E(-1) \approx 1.2160\). This is the dominant component.

Component 2: The forward hemisphere correction. The apex product depression is asymmetric, biased in the direction of propagation. The forward hemisphere correction is:

Component 3: The Sagnac cycling mass depth. The mass depression at the apex is a three-dimensional product perturbation of \(\varepsilon_0\mu_0\). Three-dimensional coupling carries a factor of \(3/2\) relative to two-dimensional coupling by the sphere-to-disk projection ratio. The Sagnac depth contribution is \((3/2)\,\delta\).

6.3 The Fine-Structure Constant

The three components add in quadrature:

The fine-structure constant is the three-dimensional coupling efficiency between the photon’s complete arc geometry and the electron’s circular closure geometry:

The CODATA value is \(1/\alpha = 137.035999084\). The gap of 0.0015% is identified as contamination from kinematic time dilation assumptions in the empirical extraction procedure — not a missing geometric term. The geometry is exact. The measurement carries the residual. This derivation contains no empirical input and no adjustable parameters.

6.4 A Dimensional Cross-Check

An independent approximate cross-check uses \(\delta_{\rm cross} \equiv \gamma_{\mathrm{total}}/2\pi \approx 0.1948\) — distinct from the Component 2 correction \(\delta\) defined above — the photon’s total arc-length ratio expressed as a fraction of a full turn. Treating each spatial dimension as contributing one factor gives \(\alpha_{\mathrm{3D}} \approx \delta_{\rm cross}^3 \approx 1/136.9\). This is a coarser approximation presented only as an order-of-magnitude cross-check.

6.5 A Falsifiable Prediction

In a system constrained to two spatial dimensions — a 2D electron gas, a topological insulator surface — the Sagnac depth oscillation has no third plane to project into. The effective fine-structure constant should approach:

No additional parameters. No fitting. Change the dimension. The coupling efficiency changes with it. Topological insulator surfaces are the test.

7. Every Boundary Is Snell and Fresnel

7.1 One Law, Every Boundary

A photon encountering any material boundary undergoes two simultaneous physical interactions. Refraction (Snell’s law): \(n_1\sin\theta_1 = n_2\sin\theta_2\). Reflection (Fresnel’s equations): \(r = (n_2 - n_1)/(n_2 + n_1)\). Nothing is special about any particular boundary. A slit wall is a boundary. A prism face is a boundary. A gravitational gradient is a boundary. A lens is a boundary. The equations are the same.

7.2 The Slit Wall Is a Lens

A slit wall is a material boundary. The gap between slit walls is vacuum — it does nothing to the photon. The slit walls are the complete optical actors. Each slit wall imposes a phase delay \(\Delta\phi = (2\pi/\lambda)(n - 1)d\) on each photon that passes it. The interference pattern on the screen is the spatial distribution of those individual outcomes. It is a noise map — a direct readout of the variation in photon conditions at the slit wall.

7.3 The Which-Way Result

When a polarizer is placed at one slit wall, it forces every surviving photon to exit with identical polarity. Identical polarity at the slit wall means identical slit wall interaction. Identical trajectory means a dot. The interference pattern vanishes. No collapse. No measurement effect. No consciousness required. The boundary changed and the pattern responded.

7.4 Huygens Is the Shadow of Snell

Huygens’ principle attributes diffraction to secondary wavelets emanating from every point on a wavefront. It is not the physical explanation. It is a mathematical reconstruction of boundary refraction that works because it is reproducing the same geometry from a different starting point. Three problems confirm this: it requires sub-photon amplitudes with no physical carrier; its obliquity factor has no physical derivation; and the slit wall material changes the interference pattern, which Huygens has no mechanism for — Snell and Fresnel contain it explicitly.

There is a further inconsistency in how Huygens’ principle is applied in practice. For the single slit, it is invoked as the physical explanation of diffraction. For the double slit, it is quietly set aside and replaced by path length geometry. A principle that is invoked selectively and replaced when a more direct calculation is available is a mathematical convenience, not a physical mechanism.

8. The Universal Coupling Law

8.1 One Equation, Five Centuries

The same equation has been discovered independently by six different scientific communities across five centuries:

where \(\theta\) is the angle between the incoming photon’s oscillation geometry and the constrained geometry of the receiver. This is not a coincidence. It is the same geometric event — the projection of a propagating \(\beta = 1\) sine wave onto a constrained coupling geometry — appearing at every scale and frequency at which photon-matter interaction occurs.

8.2 Birefringence — The Purest Form

Birefringence is the coupling law in its purest form. A birefringent crystal partitions the incoming oscillation into two components along its fast and slow axes: \(I_{\rm fast} = I_0\cos^2\theta\), \(I_{\rm slow} = I_0\sin^2\theta\), \(I_{\rm fast} + I_{\rm slow} = I_0\). Both components survive. No absorption. The partition is pure geometry. Every other interaction in this section is birefringence plus a material decision about what happens to the \(\sin^2\theta\) component.

8.3 Malus’s Law (1809)

\(I = I_0\cos^2\theta\). Malus measured this from classical beam intensities. His result was never a statement about probability. It was always a statement about the geometric projection of a continuous oscillation onto a constrained conducting axis. Malus’s Law is the oldest experimental proof that the photon carries no charge in transit — published 116 years before the photon was named.

8.4 Einstein’s Photoelectric Effect (1905)

\(E_{\rm kinetic} = h\nu - \phi\). The work function \(\phi\) is the minimum energy for the projected \(\cos^2\theta\) component to satisfy the electron’s binding geometry. Below threshold: heat. Above threshold: freed electron. The photoelectric effect is Malus’s Law applied to a threshold geometry.

8.5 Einstein’s \(B\) Coefficients (1917)

\(B_{12} = B_{21}\). Einstein derived the equality from thermodynamic detailed balance. In the \(\varepsilon_0\mu_0\) picture the equality is geometric necessity. Absorption and emission are the same event traversed in opposite directions. The coupling geometry is identical in both directions. \(B_{12} = B_{21}\) is this framework’s deepest fingerprint in Einstein’s 1917 paper.

8.6 Friis Antenna Equation (1946)

\(P_{\rm received} \propto \cos^2\theta_{\rm mismatch}\). The antenna electron is the atomic electron at radio frequencies. Reception is absorption. Transmission is emission. Same equation. Different notation. Different community. Different century. Marconi, Hertz, and Friis were doing photon physics. They called it electrical engineering.

8.7 Bell Correlation Function

\(E(a, b) = \cos^2(a - b)\). Two entangled photons carry correlated oscillation planes set at emission by the geometry of the source. This is a shared causal history, not a nonlocal bond. The correlation function is the product of two local geometric projections from a common source. Nonlocality is not required.

The Bell inequality is violated because its foundational assumption fails. Bell assumed polarizers are binary hidden-variable samplers. Polarizers are \(\cos^2\theta\) projectors. The three-polarizer experiment proves this: a fixed binary state cannot be unblocked by inserting a third filter between two crossed ones, but inserting a polarizer at 45° restores \(I = I_0/4\). The middle filter coerced the photon to a new oscillation plane. A binary measurement device cannot do this.

The Bell inequality violation proves the detector model is wrong. Not that locality is wrong.

8.8 Raman Scattering — Photon Superheterodyne

Raman scattering is the superheterodyne of a photon with a molecular vibrational mode. The molecule is the local oscillator. The incoming photon mixes with the molecular vibration at frequency \(\nu_m\). The scattered photon exits at the Stokes frequency \(\nu - \nu_m\) (difference) or the anti-Stokes frequency \(\nu + \nu_m\) (sum). This is photon frequency conversion through constrained geometry. Engineer the molecular geometry and you engineer the conversion frequency and efficiency.

8.9 The Unification

One geometric event. One equation. Six communities. Three centuries. No quantum mystery at any entry in this table. The geometry was always the same. Only the notation changed.

9. Emission and Absorption

9.1 Emission Is Field Abandonment

A photon is not ejected from an atom. The electron falls from closure radius \(r_{n_2}\) to \(r_{n_1}\). The larger closure geometry it was maintaining propagates outward as the photon. The photon is what the medium does with the geometry the electron left behind.

The photon’s frequency is set by the closure transition. \(E = h\nu\) is not a postulate relating two independent quantities. It is a geometric identity of a single physical object measured from two directions simultaneously.

The excited atom is heavier. An atom in an excited state is heavier than the same atom in its ground state by exactly \(h\nu/c^2\). When the atom emits, it loses exactly that mass. Something leaves carrying it. That something is the photon.

9.2 Absorption Is Emission Run in Reverse

Absorption is the geometric inverse of emission. This is a statement about field geometry, not about time direction. Nothing moves backward in time. The positron is not an electron running backward in time — that picture belongs to a perturbation-expansion bookkeeping convention in quantum field theory, not to a physical description of what the positron is. The positron is a closure geometry of opposite handedness to the electron, propagating forward in time like everything else.

The incoming photon’s arc geometry couples to the receiving closure geometry. If the photon’s \(r_{\mathrm{ph}} = \lambda/2\pi\) matches the inter-shell geometry of the target closure, the field re-establishes the abandoned configuration and the electron rises. The excited atom is heavier by exactly \(h\nu/c^2\). The photon ends. If the photon’s \(r_{\mathrm{ph}}\) does not match, the coupling fails and the photon continues. No partial absorption. No probability cloud. Either the geometry matches or it does not.

9.3 The Coupling Is Geometric Curvature Matching

The photoelectric threshold is the minimum curvature condition for the projected \(\cos^2\theta\) component to satisfy \(\beta = 1\) at the receiving closure. Below threshold: geometric mismatch, energy deposits as heat. Above threshold: complete coupling, electron freed.

Photosynthesis. Chlorophyll’s closure geometry has been tuned over four billion years of evolution to match solar photon curvature at specific visible frequencies. The 95%+ quantum efficiency attributed to “quantum biology” is a high Einstein \(B\) coefficient — a geometrically optimized coupling constant.

Microwave heating. Water’s molecular closure geometry matches the microwave photon’s apex curvature. Dry paper and glass do not. Same mechanism as the photoelectric effect at a different scale.

Photovoltaic cells. The semiconductor bandgap is the minimum closure reconfiguration energy for the target material. Below bandgap: mismatch, heat. Above bandgap: complete coupling, freed carrier. The photoelectric effect and the photovoltaic effect are the same geometric event in different materials.

Three names. One mechanism. Geometric curvature matching at the receiving closure.

9.4 A Lens Is Heavier When Light Passes Through It

A direct consequence of the photon having real mass: a lens in the path of a beam is heavier during transmission than before or after. The photon’s mass \(m = h\nu/c^2\) is physically present in the lens medium for the duration of the transit time \(t = d/v_{\rm medium}\). For visible light through a 1 cm glass lens, \(m \approx 10^{-36}\) kg per photon and the transit time is \(\sim 50\) ps. The effect is currently below measurement threshold. It is nonetheless a necessary consequence of the photon having real mass. Orthodox physics has no equivalent prediction — the massless photon passing through a lens adds no mass to the system.

9.5 Emission Is Deterministic

There is no spontaneous emission as a random quantum event. Emission is deterministic. Its cause is always the field conditions at the orbital boundary — the local \(\varepsilon_0\mu_0\) density at the moment the closure geometry becomes energetically favorable to contract. Randomness in emission statistics is epistemic. It is what ignorance of local field conditions looks like in aggregate.

9.6 The Near Field and the Far Field

The near field of a radiating antenna or atom is not a photon. It is the ratio-perturbation wake of the oscillating closure geometry — tethered to the source, following the current. The far field is what escapes when the oscillating ratio perturbation can no longer be re-absorbed by the source geometry. At that boundary — the transition distance \(\sim\lambda/2\pi = r_{\mathrm{ph}}\) — the ratio perturbation converts to product perturbation and escapes as a propagating sine wave. The photon does not exist before that threshold. It comes into being at the moment the abandoned field geometry propagates beyond recapture.

10. Polarization: What It Is and What It Is Not

10.1 The Polarity Axis Is Physical

The photon’s oscillation plane is a fixed geometric property set at emission by the geometry of the electron closure contraction that produced the photon. It is as physically real and definite as the direction a rope oscillates when you shake one end of it. It persists in free propagation because nothing in an isotropic medium torques it. Polarization measured from light emitted billions of light-years away remains well-defined when measured on Earth. The orientation carries because nothing in free propagation changes it.

10.2 The Polarizer: Irreversible Coercion

A polarizer contains long molecular conducting chains oriented along one axis. When a photon arrives, its oscillation drives electrons along the chains. The \(\cos^2\theta\) component is absorbed and re-emitted with its oscillation plane reset to the chain geometry. The \(\sin^2\theta\) component deposits as a phonon — heat. The interaction is absorption followed by re-emission. The exiting photon is a new photon whose oscillation plane has been set by the chain axis. This is why the effect is permanent and irreversible: the original photon ended inside the polarizer. A new photon was emitted along the chain axis. There is nothing to undo.

10.3 The Crystal: Reversible Geometry

A birefringent crystal partitions the oscillation into two components propagating at different speeds. Both components survive. No absorption. No phonon. The total energy is conserved exactly. The differential propagation speed creates a phase offset at the exit face. A second crystal oriented oppositely undoes the net rotation exactly because the rotation was reversible elastic geometry — no energy was dissipated.

10.4 The Beth Torque: Mechanism Correctly Identified

In 1936, R. A. Beth placed a birefringent crystal on a torsion fiber, illuminated it with circularly prepared light, and observed a sustained deflection of the fiber. The torque is real. The mechanism is differential coupling over dwell time. The photon’s oscillation plane enters the crystal at an angle to the fast and slow axes. The asymmetric mechanical engagement exerts a torque on the crystal lattice continuously for the entire duration the photon is inside the crystal. The lattice gains angular momentum not as a discrete quantum handed off per photon, but as a continuous torque acting over each photon’s transit time. The angular momentum balance closes through time.

10.5 Circular Polarization Is Not a Photon Property

For the photon’s oscillation plane to rotate continuously during free propagation, a torque must act on it at every point along its path. The medium is isotropic. There is no physical mechanism in the \(\varepsilon_0\mu_0\) medium that continuously rotates a propagating sine wave’s oscillation plane. A torque with no source cannot sustain a rotation. A continuously rotating oscillation plane in a propagating photon is geometrically impossible.

“Circular polarization” is a description of a measurement context, not a property of the propagating photon. It describes a specific phase relationship between two orthogonal components set up by a birefringent crystal and read by a subsequent optical element. What the measurement actually shows is orthogonal linear polarization planes. That is the observation. The circular polarization interpretation is a theoretical overlay, not a reading of the detector.

10.6 \(E\) and \(B\) Are Simultaneous Readings, Not a Sequence

\(E\) and \(B\) are not two events in time. They are two simultaneous material responses of the \(\varepsilon_0\mu_0\) medium to one disturbance event. \(E\) is the permittance reading — how much the medium’s ratio displaces under the disturbance, governed by \(\varepsilon_0\). \(B\) is the reluctance reading — how much the medium resists that displacement, manifesting as a curl, governed by \(\mu_0\). These are two properties read off one event at the instant it occurs. The polarity axis is therefore a single, well-defined direction that persists unchanged in free propagation.

11. Annihilation and Pair Production

11.1 Annihilation Is Curl Cancellation

An electron and a positron are closure geometries of opposite handedness. When they meet, their curl geometries cancel. The combined geometry has zero net departure from \(Z_0\). The field energy stored in both departures has no closure to confine it. It propagates outward as the medium recovers. The two photons are not created in the event. They are the medium recovering.

Annihilation is two conjugate departures from \(Z_0\) meeting, summing to zero, and the medium recovering at \(c\) from both directions simultaneously. The energy accounting is exact: \(2mc^2\) in, \(2h\nu\) out, with \(h\nu = mc^2\) for each photon.

11.2 What Actually Differs Between the Two Photons

The two annihilation photons travel in opposite directions and carry orthogonal oscillation planes — both directly observed and confirmed. These are the complete physically distinguishing properties. Each photon carries a fixed oscillation plane, a definite frequency, and a transverse radius \(r_{\mathrm{ph}} = \lambda/(2\pi)\). Neither carries charge. Neither carries a winding direction. Neither carries handedness.

The literature frequently attributes opposite circular polarization to the two annihilation photons. This attribution imports the E/B oscillation picture of a photon — incorrect for propagating photons as established in Section 10. The photon is a mass/energy oscillation between apex and zero crossing. It has no charge topology and no persistent rotation of its oscillation plane. The “opposite circular polarization” of annihilation photons is a bookkeeping label applied within a framework this paper has shown to be incorrect. What the measurement actually shows is orthogonal linear polarization planes. That is the observation.

11.3 Pair Production Is Geometric Nucleation

A gamma ray above 1.022 MeV, passing near a nucleus, produces an electron-positron pair. The threshold energy matches \(2mc^2\) exactly. The orthodox account calls this creation from vacuum — something from nothing. The \(\varepsilon_0\mu_0\) picture is more precise: the medium nucleating two stable closure geometries when sufficient energy is supplied to lock them against the medium’s recovery drive. The medium was always there. The photon supplied the energy needed to form and stabilize two conjugate closure geometries.

The “from nothing” framing is a narrative layered on top of mathematics that says nothing of the sort. The mathematical description says the field reorganizes. The \(\varepsilon_0\mu_0\) medium is not nothing. It has a measured impedance of 376.730 Ω. It has structure available for nucleation. Pair production is the medium responding to field energy above nucleation threshold.

12. What the Measurement Tool Implies

The photon is the measurement tool for every result in this section. Each experiment listed below comes off a photon detector. The physical picture established in the preceding sections — the photon as a \(\beta = 1\) sine wave propagating through the \(\varepsilon_0\mu_0\) medium at recovery rate \(c\), carrying frequency set at emission, unchanged in transit — has direct implications for what those detectors are reading.

12.1 Pound-Rebka: The Medium Varies with Height

In 1959, Pound and Rebka measured a frequency shift between the ground floor and the top of a 22.5-metre tower at Harvard. The fractional shift was:

In the \(\varepsilon_0\mu_0\) picture this is a density reading. The medium is less dense at the top of the tower than at the bottom. The photon’s frequency at reception reflects the \(\varepsilon_0\mu_0\) density at the reception environment, not a property the photon acquired or lost in transit. The photon does not change. The medium it is read against does. Pound-Rebka is a direct measurement of \(\varepsilon_0\mu_0\) varying with gravitational potential across 22.5 metres.

12.2 GPS: The Medium Varies with Altitude

GPS satellites operate at \(\sim 20{,}200\) km altitude where \(\varepsilon_0\mu_0\) is lower than at Earth’s surface, and clocks there run fast by the measured \(+38.2\,\mu\text{s/day}\). The photon is the physical mechanism by which every clock keeps time. An atomic clock counts oscillation cycles of electromagnetic radiation — photon apex events at a specific frequency. The tick rate is the local photon frequency, set by the local \(\varepsilon_0\mu_0\) density. The clock runs faster because the medium is thinner and the photon cycles faster in it. The photon is not merely the measurement tool. It is the clock mechanism.

12.3 LIGO: The Medium Changes, Not the Ruler

The LIGO interferometer detects gravitational waves by measuring phase shifts in laser light reflected between mirrors separated by 4 kilometres. The orthodox description says the arm lengths change — space itself stretches and compresses as the wave passes.

The \(\varepsilon_0\mu_0\) picture is more precise. The mirrors do not move relative to any local reference. What changes is the local \(\varepsilon_0\mu_0\) product as the wave passes through the measurement environment. A change in \(\varepsilon_0\mu_0\) is a change in \(c\). A change in \(c\) is a change in propagation time across a fixed physical separation. A change in propagation time is a phase shift in the returning light. LIGO measures that phase shift.

The mirrors are not getting closer and further apart. The light is redshifting and blueshifting. A frequency shift without a distance change is a medium density change — exactly what Pound-Rebka measured across 22.5 metres in 1959.

LIGO reports a strain \(h = \Delta L / L \sim 10^{-21}\) for a gravitational wave event. In the \(\varepsilon_0\mu_0\) picture, that strain is a fractional change in propagation speed across the arm:

The phase shift the interferometer reads is real. The arm length interpretation is not — no mechanical force has acted on the mirrors, and no force is proposed. The medium changed. The propagation time changed. The phase changed. The length-contraction framing promotes a coordinate description of the same event to physical status without a mechanism.

The Fabry-Perot cavities in each LIGO arm are designed to amplify the gravitational wave signal by accumulating phase across approximately 300 bounces between mirrors. This design is correct for mechanical mirror displacement — each bounce traverses a persistently displaced mirror and accumulates an independent path length increment. Three hundred bounces genuinely amplify mechanical displacement by a factor of 300. A gravitational wave is not mechanical mirror displacement. It is a \(\varepsilon_0\mu_0\) density change — uniform and simultaneous across the entire arm. Every photon in the cavity at any given moment traverses the same changed medium together. The gravitational wave signal is present completely in a single pass. The cavity amplifies it by a factor of one.

The mirror the cavity was built to serve is the instrument’s dominant noise source. Every physical process that displaces it — seismic coupling, thermal expansion, acoustic disturbance, radiation pressure fluctuation — is amplified 300-fold by the same mechanism intended to amplify the signal. The cavity adds noise. It does not add signal.

LISA, adopted by ESA in January 2024, implements the minimal architecture by engineering necessity. Three spacecraft separated by 2.5 million kilometre baselines, exchanging single laser beams with no return mirror and no cavity. The gravitational wave signal is a differential frequency shift between baselines, read directly. The dominant noise sources of the LIGO design — the mirrors — do not exist in an instrument without mirrors to vibrate. LISA is the first instrument positioned to read the \(\varepsilon_0\mu_0\) medium velocity directly from the beam-pointing correction history, because it is the first instrument without a 300-fold noise amplification mechanism obscuring the signal.

Both LIGO and LISA settle a question that has persisted since 1909. The light clock thought experiment requires the photon to follow the mirrors laterally as the mirror system moves. The diagonal path is the entire derivation. But the photon belongs to the medium, not to the apparatus. It does not follow the mirrors. LIGO demonstrates this on every data run — the photon is released into the \(\varepsilon_0\mu_0\) medium and travels straight, while the mirrors move and LIGO’s entire engineering infrastructure exists to track and correct for the displacement between where the photon went and where the mirror is. LISA confirms it at cosmological scale: the Point-Ahead Angle Mechanism physically tilts the outgoing beam to lead the target — because the photon will not follow the spacecraft. The spacecraft moves through the medium. The photon does not. The thought experiment that appeared in physics textbooks for over a century as a derivation of time dilation was always showing us this instrument and this measurement. The diagonal path was never time dilation. It was the medium.

The measurement tool is the photon. The reading is the medium.

This parallel has a further consequence for the cosmological redshift debate. If LIGO’s apparent distance change is actually a measurement environment change — local \(\varepsilon_0\mu_0\) density varying as the wave passes — then the same logic applies cosmologically. The Snell inversion settles the point numerically: given the observed redshift, invert the \(\varepsilon_0\mu_0\) density gradient along the photon’s path that would produce it. The result is a density map of the medium along that path — not a recession velocity, not an expanding coordinate system, but a physical property of the medium that can be compared with independent mass-distribution observations.

12.4 Cosmological Redshift: Path-Integrated Medium, Not Recession

A photon from a distant galaxy arrives with a lower frequency than it was emitted with. The conventional interpretation is that the galaxy is receding and space is expanding between us and it. The \(\varepsilon_0\mu_0\) picture offers a different reading. The photon does not change in transit. The medium it propagates through is not uniform across cosmological distances. A photon accumulating path-integrated \(\varepsilon_0\mu_0\) variation across billions of light-years arrives with a frequency that reflects the integrated density history of its path, not a recession velocity.

This is the same mechanism that operates in Pound-Rebka across 22.5 metres, in GPS across 20,200 km, and in LIGO across 4 kilometres. The scale changes. The mechanism does not. Calling the mechanism a density effect at 22.5 metres and an expansion effect at cosmological distances requires two different physical explanations for the same measurement result at different scales.

The expansion interpretation generates a difficulty it cannot easily resolve. At the Hubble horizon, galaxies must be receding at \(c\). Beyond it, faster than \(c\). The \(\varepsilon_0\mu_0\) picture has no such problem. Nothing moves faster than \(c\). The medium thins. The frequency drops. The apparent recession is a medium density reading, not a velocity.

12.5 Pound-Rebka and the Tower

Pound-Rebka measured a redshift between the ground and the top of a 22.5-metre tower. By the logic of the expansion interpretation — redshift equals recession velocity equals space expanding between the two points — the top of the tower is moving away from the ground. At measurable speed. Continuously. The tower is getting taller. It is not. The tower is stationary. The redshift is a \(\varepsilon_0\mu_0\) density difference across 22.5 metres.

12.6 The CMB: A Horizon Phenomenon

The cosmic microwave background is observed at \(T \approx 2.725\) K. The orthodox interpretation says it is relic radiation from a primordial fireball 13.8 billion years ago, redshifted by the expansion of space. A prediction of this interpretation is that the CMB should be cooling as the universe continues to expand. The CMB temperature is stable.

The \(\varepsilon_0\mu_0\) picture: the photon geometry makes no distinction between a CMB photon and an optical photon. Both are \(\beta = 1\) sine waves with \(r_{\mathrm{ph}} = \lambda/(2\pi)\). A photon from a sufficiently distant source accumulates sufficient path-integrated \(\varepsilon_0\mu_0\) variation to arrive at centimetre wavelengths regardless of the frequency at which it was emitted. The CMB becomes a horizon phenomenon — the frequency floor set by the path-integrated medium integral at the limit of the observable volume. The CMB tells us where our photon path integral saturates. It does not tell us where the universe began.

13. What This Displaces

The following is not a list of things this framework objects to. It is a list of things that become unnecessary once the photon is correctly described. Each item was introduced to solve a problem. Each problem dissolves when the medium is restored and the photon geometry is correctly stated. The solutions were never wrong as mathematics. They were wrong as ontology.

The massless photon. The photon has rest mass \(h\nu/c^2\). This follows from conservation of energy at emission alone. The massless photon was introduced to preserve the null worldline required by kinematic time dilation. KTD required a massless photon. The photon requires mass. One of these is wrong. The photon is heavier.

Wave-particle duality. The photon is a wave. Wave-particle duality was introduced when the medium was removed and the photon lost its substrate. Without the medium the photon had no physical description. Duality was the admission of that failure, elevated to a principle.

The null worldline. The photon has mass \(h\nu/c^2\) distributed across the cycle. It has physical extent \(r_{\mathrm{ph}} = \lambda/(2\pi)\). It is at absolute rest in the medium at each apex. The null worldline was introduced to give the massless point-photon a worldline consistent with special relativity. The photon never needed it. Special relativity needed it.

The affine parameter. Introduced to replace proper time on a null geodesic. It has no physical content. The physical quantity is the photon’s cycle — two apexes, two zero crossings, advancing at \(c\). That is the parameter. It always existed.

Virtual photons as force mediators. The force between two electrons is the gradient interaction between their \(\varepsilon_0\mu_0\) closure geometries — a real, continuous, local field interaction in the medium. No virtual particle required. Virtual photons were introduced when the medium was declared absent. The medium does the work.

The measurement problem. For photons, the problem dissolves. The photon is a classical wave in the \(\varepsilon_0\mu_0\) medium with a definite oscillation plane, a definite frequency, and a definite transverse radius. When it meets a polarizer, the polarizer coerces it through conducted energy exchange — a real physical interaction. There is no superposition to collapse. The measurement problem for photons is the ontological residue of removing the medium in 1905.

Quantum indeterminism as fundamental. The photon does not choose randomly which way to go at a beam splitter. It couples to whichever geometry it encounters, deterministically, according to the \(\cos^2\theta\) projection of its oscillation plane. Indeterminism was promoted from epistemic to ontological when the photon lost its medium.

Nonlocality. Bell inequality violations do not require nonlocality. They require that polarizers are \(\cos^2\theta\) projectors, not binary hidden-variable samplers. The three-polarizer experiment confirms the projector mechanism independently. Nonlocality was required by Bell’s hidden-variable assumption. That assumption fails because it models the wrong detector physics.

The electromagnetic spectrum as a description of transit. The name correctly describes the endpoint apparatus and correctly orders the frequencies from radio to gamma. What is incorrect is the implication that the propagating entity between the endpoints is electromagnetic in character. In transit, at every frequency from radio to gamma, the propagating entity is a \(\beta = 1\) product perturbation of the \(\varepsilon_0\mu_0\) medium. The spectrum is photonic throughout. Maxwell correctly identified the endpoints and correctly ordered the frequencies. He misidentified the character of what travels between them.

Huygens’ principle as a physical explanation. Huygens’ principle is a mathematical reconstruction of boundary refraction. It has no physical actor. Its sub-photon amplitude contributions have no physical carrier. Its obliquity factor has no physical derivation. Snell’s law and Fresnel’s equations are the physics. Huygens is the shadow. Furthermore, Huygens is invoked as mechanism for the single slit and quietly replaced by path-length geometry for the double slit — a selective application that exposes its status as a mathematical convenience rather than a physical explanation.

The photoelectric effect as a quantum mystery. The photoelectric threshold is a minimum curvature condition. No particle required. No quantum mystery. A wave with a specific geometric coupling condition meeting a target closure that either matches or does not.

Planck’s constant as a primitive. \(h\) is the arc-length closure condition of a \(c\)-constrained oscillation in the \(\varepsilon_0\mu_0\) medium, expressed in SI units. Maxwell’s equations contained it in 1865. The cycle was always the unit. \(h\) was always its SI measurement.

Quantization as an axiom. Discrete energy levels require no quantization axiom. The closure condition permits only specific transverse radii. The quantum ladder is a radius ladder.

Spin angular momentum of a single propagating photon. A propagating photon cannot carry spin angular momentum. There is no torque source in an isotropic medium to sustain a rotation. The Beth torque is real and correctly measured. Its source is differential mechanical coupling over dwell time in an anisotropic crystal, not a spin quantum carried by each photon. Spin angular momentum of a single propagating photon is geometrically impossible.

Circular polarization as a photon property. A propagating photon cannot have a rotating oscillation plane. “Circular polarization” is an apparatus property. It is not a photon property.

The cosmological constant problem. The vacuum energy catastrophe arises from quantizing a field in a vacuum declared empty. The \(\varepsilon_0\mu_0\) medium is not empty. It has a measured impedance. Quantizing an empty vacuum and discovering it has infinite energy is a category error. The medium is not empty. The catastrophe is a consequence of the wrong ontology applied to the right equations.

13.1 What Remains

Maxwell’s equations remain. They correctly describe electron closure dynamics in the \(\varepsilon_0\mu_0\) medium.

The Sagnac effect remains. It correctly describes phase accumulation in a rotating medium.

Snell’s law and Fresnel’s equations remain. They correctly describe every material boundary.

The Doppler effect remains. It correctly describes frequency shift between source, medium, and receiver.

\(E = h\nu\) remains. It correctly states the geometric cycling cost of one \(\beta = 1\) apex event in SI units.

\(E = mc^2\) remains. It correctly states the identity between energy at rest and mass. The photon apex satisfies it most directly.

The fine-structure constant remains. It correctly states the coupling efficiency between photon arc geometry and electron closure geometry. It is now derived rather than measured.

The experimental record remains entire. Every measurement stands. The ontology changes. The instruments were always reading the medium. They still are.

Conclusion

This paper has derived the photon from first principles.

Not described it. Not modeled it. Derived it — from the single requirement that a propagating disturbance in the \(\varepsilon_0\mu_0\) medium remain causally consistent for all wavelengths. Every property of the photon follows from that requirement without postulate, without free parameter, and without appeal to quantum mechanics as a foundation.

The photon is a transverse oscillation of the medium. Its geometry is fixed. Its mass is real. Its charge is absent. Its coupling law is \(\cos^2\theta\), operating identically across six communities and three centuries without any of them knowing they were measuring the same geometric event. The medium was always there. The photon was always this.

The tools used in this derivation were available in 1865. Maxwell’s equations contained Planck’s constant. Weber and Kohlrausch measured the medium’s impedance from a capacitor discharge before Maxwell had a wave equation. Malus proved the photon carries no charge in 1809, 116 years before the photon was named. Bradley measured the medium in 1727 with a telescope. Nothing new was required. What was required was the willingness to follow the physics where it led rather than where it was expected to go.

The paper makes five falsifiable predictions that distinguish it from the standard model:

Atomic transition duration is nonzero. For hydrogen Lyman-alpha, \(\Delta t \approx 0.53\) attoseconds. The standard model predicts zero. Attosecond spectroscopy is the test.

The fine-structure constant in a two-dimensional electron system approaches \(1/\alpha_{\rm 2D} \approx 26\)–\(27\). The standard model makes no such prediction. Topological insulator surfaces are the test.

Slit wall material changes diffraction fringe spacing systematically as \((n-1)d/\lambda\). Standard diffraction theory attributes fringe patterns to gap geometry alone. Thin-film slit walls of varying refractive index are the test.

LISA, the space-based gravitational wave observatory with 2.5 million kilometre baselines and no return mirror, is the first instrument positioned to read the \(\varepsilon_0\mu_0\) medium velocity directly from its beam-pointing correction history. LIGO’s Fabry-Perot cavities amplify mechanical mirror noise 300-fold while adding no gravitational wave signal amplification — the cavity amplifies displacement, and a gravitational wave is not displacement. LISA eliminates the mirror and the cavity. The pointing correction history is a clean, continuous, accumulating record of the solar system’s translational velocity through the medium. That record will begin accumulating from the moment of first lock.

The effective local Planck constant varies with \(\varepsilon_0\mu_0\) density in extreme gravitational environments. GPS already confirms this indirectly at the \(38.2\,\mu\text{s/day}\) level. Direct confirmation requires precision spectroscopy at varying gravitational potentials.

These predictions are specific, numerical, and experimentally accessible. If they hold, the framework is confirmed. If any fails, it must be revised. That is the correct relationship between a physical theory and the world it describes.

The medium was never absent. It was always in the equations. It was always in the measurements. \(Z_0 = 376.730\,\Omega\) is not a property of nothing. The battery loses mass when the transmitter fires. The tower is not getting taller.

The photon has been here the whole time. We were just not reading it correctly.

Appendix: Optical Relations in Radius Form

Replacing \(\lambda\) with \(2\pi r_{\mathrm{ph}}\) in any wavelength-based formula leaves the numerical prediction unchanged while making the causal geometry explicit.

Key: \(r_{\mathrm{ph}} = \lambda/(2\pi) = \bar{\lambda}\). The ratio \(r_{\mathrm{ph}}/\lambda = 1/(2\pi)\) is a geometric constant from the arc-length closure condition. It is not the fine-structure constant \(\alpha \approx 1/137\).

Note on the top three rows. The photon mass, momentum, and energy are not traditionally listed in optical tables because they follow from substituting \(r_{\mathrm{ph}} = \hbar/p\) — a relation physics has used since de Broglie without identifying \(r_{\mathrm{ph}}\) as the photon’s physical transverse radius. Once \(r_{\mathrm{ph}}\) is identified as a geometric dimension of the photon, these relations express the same physical quantities from a different angle, and the photon’s mass is no longer paradoxical.

Note on transition duration. For hydrogen Lyman-alpha (\(\lambda = 121.6\) nm): \(\Delta t \approx 0.53\) attoseconds. This is a falsifiable prediction distinguishing the two frameworks. Attosecond spectroscopy is the test.

References

Beth, R. A. (1936). Mechanical detection and measurement of the angular momentum of light. Physical Review, 50(2), 115–125.

Malus, É. L. (1809). Sur une propriété de la lumière réfléchie par les corps diaphanes. Bulletin de la Société Philomathique de Paris, 1, 266–269.

Maxwell, J. C. (1865). A dynamical theory of the electromagnetic field. Philosophical Transactions of the Royal Society of London, 155, 459–512.

Einstein, A. (1905). Über einen die Erzeugung und Verwandlung des Lichtes betreffenden heuristischen Gesichtspunkt. Annalen der Physik, 17, 132–148.

Einstein, A. (1917). Zur Quantentheorie der Strahlung. Physikalische Zeitschrift, 18, 121–128.

Planck, M. (1900). Zur Theorie des Gesetzes der Energieverteilung im Normalspektrum. Verhandlungen der Deutschen Physikalischen Gesellschaft, 2, 237–245.

Pound, R. V., & Rebka, G. A. (1959). Gravitational red-shift in nuclear resonance. Physical Review Letters, 3(9), 439–441.

Friis, H. T. (1946). A note on a simple transmission formula. Proceedings of the IRE, 34(5), 254–256.

Bell, J. S. (1964). On the Einstein-Podolsky-Rosen paradox. Physics, 1(3), 195–200.

Scully, M. O., Englert, B.-G., & Walther, H. (1991). Quantum optical tests of complementarity. Nature, 351, 111–116.

Sagnac, G. (1913). L’éther lumineux démontré par l’effet du vent relatif d’éther dans un interféromètre en rotation uniforme. C. R. Acad. Sci., 157, 708–710.

Schlamminger, S. et al. (2008). Test of the equivalence principle using a rotating torsion balance. Physical Review Letters, 100, 041101.

Ashby, N. (2003). Relativity in the Global Positioning System. Living Rev. Relativ., 6, 1.

Abbott, B. P. et al. (LIGO Scientific Collaboration and Virgo Collaboration). (2016). Observation of gravitational waves from a binary black hole merger. Physical Review Letters, 116, 061102.

European Space Agency. (2024). LISA Mission Adopted. ESA Science Programme Committee, 25 January 2024. esa.int/LISA

Maupertuis, P.-L. M. de (1744). Accord de différentes lois de la nature. Mém. Acad. R. Sci. Paris, 417–426.

de Broglie, L. (1924). Recherches sur la théorie des quanta. Ph.D. thesis, University of Paris.

Particle Data Group. (2024). Review of Particle Physics. Prog. Theor. Exp. Phys., 2024, 083C01.

Born, M., & Wolf, E. (1999). Principles of Optics (7th ed.). Cambridge University Press.

CODATA. (2018). Fine-structure constant \(1/\alpha = 137.035999084\). Reviews of Modern Physics, 93, 025010 (2021).

Hallman, D. J. (2026). \(\gamma_{\mathrm{cause}}\): A geometric closure invariant governing transverse electromagnetic oscillations. Zenodo. DOI: 10.5281/zenodo.20132405

Hallman, D. J. (2026). Forensic examination of the kinematic term in special and general relativity. Zenodo. DOI: 10.5281/zenodo.20132769

Hallman, D. J. (2026). The Sagnac formula inverted reveals mass, gravity, and particle structure. Zenodo. DOI: 10.5281/zenodo.20225842

Hallman, D. J. (2026). Bell versus Malus: A first-principles clash of models in quantum measurement. Zenodo. DOI: 10.5281/zenodo.19017274

All works are published open-access through Zenodo under a CC BY 4.0 license. Full publication list: Zenodo · ORCID 0009-0000-1710-3549 · Contact: SCG@azfn.com