Electromagnetic Closure and the Fine-Structure Constant: A Geometric Derivation 10.5281/zenodo.19157339 Kelly B. Heaton
& Coherence Research Collaboration
Version 1.2 issued April 12, 2026 with minor corrections
thecoherencecode.com & lucernaveritas.ai March 2026 ------------------------------------------------------------------------ Abstract The fine-structure constant \(\alpha = Z_0/2R_K \approx 1/137.036\) is derived from the non-terminal reciprocal singularity \(1/z\) at \(z = 0\), completed on the Riemann sphere. The derivation proceeds without free parameters, postulates, or imported physical structure. The singularity forces three consequences on the completed space: an irremovable source defect at \(p\), a unique antipodal sink \(p^*\) required by the Poincaré–Hopf theorem (\(\chi(S^2) = 2\)), and a logarithmic potential field between them. The self-similar scaling structure of this field has a unique attractor, the golden ratio \(\varphi = [1;\,1,1,1,\ldots]\), forced by the Markov spectrum theorem and the uniform resolvent condition (Hurwitz, 1891). Rational winding is excluded by the existence of matter. The irremovable defect forces a departure \(\delta > 0\) from \(\varphi\). The electromagnetic constants \(\varepsilon_0\) and \(\mu_0\) are not introduced as physical inputs. Their own definitions coincide with the mathematical primitives already present in the structure: \(\varepsilon_0\) is the capacity for containment (the primitive \(1\)) and \(\mu_0\) is the openness to penetration (the primitive \(0\)). Their ratio \(Z_0 = \sqrt{\mu_0/\varepsilon_0}\) is therefore the \(0/1\) geometric ratio of the vacuum, definitionally. No additional structure is required. Within this framework the constancy of \(c = 1/\sqrt{\varepsilon_0\mu_0}\) is a mathematical consequence, not a physical postulate. The departure \(\delta\) is assembled from four established facts: the residue of \(1/z\) at \(0\) equals \(1\); \(Z_0\) is the \(0/1\) vacuum ratio; \(R_K = h/e^2\), where \(h\) is the invariant of the self-similar scaling process and \(e\) is the minimum charge of the defect; and the minimum winding number is \(2\). The result is: \[\alpha \;=\; \frac{Z_0}{2R_K} \;=\; \frac{e^2}{2\varepsilon_0 hc} \;\approx\; \frac{1}{137.036},\] with zero free parameters. Neither \(h\), \(e\), nor \(c\) is a primitive of the structure; all are derived consequences. The derivation is not an extraction of physics from mathematics. It is the recognition that the physical vocabulary of electromagnetism names the mathematical structure that the non-terminal reciprocal singularity \(1/0\) necessarily generates. Keywords: fine-structure constant, Riemann sphere, reciprocal singularity, punctured sphere, Poincaré–Hopf theorem, logarithmic potential, golden ratio, Markov spectrum, Hurwitz theorem, self-similar scaling, vacuum impedance, vacuum permittivity, vacuum permeability, von Klitzing constant, Josephson constant, Diophantine impossibility, impedance matching, orbital impedance, electromagnetic primitives, first-principles derivation. Introduction The fine-structure constant $$\alpha \;=\; \frac{e^2}{2\varepsilon_0 hc} \;=\; \frac{Z_0}{2R_K} \;\approx\; \frac{1}{137.036} \label{eq:alpha_def}$$ has resisted first-principles derivation since its introduction by Sommerfeld in 1916 (Sommerfeld 1916; Feynman 1985). In the second form of [eq:alpha_def], $Z_0 = \sqrt{\mu_0/\varepsilon_0}$ is the impedance of free space and R_(K) = h/e² is the von Klitzing constant. This expression identifies α as the ratio of two independent electromagnetic invariants of the vacuum: the impedance Z₀, which governs the partition of field energy between the electric and magnetic modes, and the resistance quantum R_(K), which governs the action cost of closing one electromagnetic loop (K. B. Heaton and Coherence Research Collaboration 2026b). Any derivation of α must reproduce this two-part structure: a numerator quantity of Z₀-character divided by a denominator quantity of R_(K)-character. The problem has attracted serious attention since Eddington’s conjecture that 1/α could be obtained “by pure deduction” (Eddington 1946). A comprehensive survey of subsequent attempts is given by Jentschura and Nandori (Jentschura and Nandori 2014), who conclude that no physically grounded derivation has succeeded. Wyler’s formula (Wyler 1969), $$\alpha_W \;=\; \frac{9}{8\pi^4} \left(\frac{\pi^5}{2^4\cdot 5!}\right)^{1/4} \;\approx\; \frac{1}{137.036}, \label{eq:wyler}$$ derived from the conformal group of Maxwell’s equations, remains the closest prior result numerically, but Robertson (Robertson 1971) showed that the formula agrees with experiment only if the radius of Wyler’s bounded spaces is arbitrarily set to unity, a choice for which no physical justification has been given. More recently, Singh (Singh 2022) has derived an asymptotic value close to 1/137 from the eigenvalues of the exceptional Jordan algebra J₃(𝕆), but the derivation requires an assumed functional form for α(q) and a fitted electroweak symmetry-breaking scale, working top-down from high-energy algebraic structure with external inputs at each step. The present derivation proceeds differently from all prior attempts in three respects. First, it proceeds bottom-up from a purely mathematical starting point: the behaviour of the non-terminal reciprocal singularity 1/z as z → 0, completed on the Riemann sphere (Riemann 1851). No group-volume ansatz, no conformal symmetry assumption, no algebraic structure, and no fitted parameter enters. Second, $\varphi = (1+\sqrt{5})/2$ is not postulated. It is derived as the unique fixed point of the self-similar scaling recursion forced by the logarithmic field on the punctured sphere, and shown to be uniquely forced by the Markov spectrum theorem (Markov 1880) together with the uniform regularity condition on the resolvent. Third, and most importantly, the derivation does not import the electromagnetic constants ε₀, μ₀, Z₀ into the argument from physics. It identifies them with the mathematical structure: the vacuum permittivity ε₀ is the capacity for containment (the mathematical primitive 1) and the vacuum permeability μ₀ is the openness to penetration (the mathematical primitive 0), by their own physical definitions. The impedance $Z_0 = \sqrt{\mu_0/\varepsilon_0}$ is therefore the 0/1 ratio of the mathematical structure expressed in physical vocabulary. No additional structure is introduced: the identification is definitional, not axiomatic. A direct consequence, within this framework, is that the constancy of $c = 1/\sqrt{\varepsilon_0\mu_0}$ follows as a mathematical necessity rather than a physical postulate (Section 5, Corollary 35). Structure of the paper. Section 2 establishes the topological consequences of the non-terminal reciprocal singularity: the punctured Riemann sphere, the irremovable defect at p, the antipodal sink p^(*) forced by the Poincaré–Hopf theorem, and the logarithmic potential field. Section 3 derives the φ-attractor of the self-similar scaling, excludes rational winding by the existence of matter, forces φ uniquely by the Markov spectrum theorem and the uniform regularity condition, and establishes the departure δ > 0 from the ideal ratio. Section 4 identifies the electromagnetic realisation of the primitive structure, derives the orbital impedance Z_(orb)(n) = Z₀n/α from Coulomb and Biot–Savart, establishes the universal gap at IE/4 by a number-theoretic argument, and shows that neither h nor e is a primitive of the structure. Section 5 closes all remaining gaps: the continued fraction argument for the recursion x = 1 + 1/x, the definitional proof that no quotient is introduced, the assembly of δ = Z₀/2R_(K) from four established facts, and the derivation of the Josephson constant from the minimum winding condition. Section 6 identifies where hydrogen and atomic spectroscopy live inside the derived structure, without using them as premises. Section 7 states the result and its honest characterisation. The proof rests on zero free parameters, zero postulates, one mathematical regularity premise (P10), and one empirical observation: matter exists. Primitive reciprocal singularity and forced asymmetry We begin from a classical problem in the foundations of analysis: the status of division by zero, and in particular the behaviour of the reciprocal map z ↦ 1/z as z → 0. Historical context and geometric completion Brahmagupta (Brahmagupta 628AD) gave the first systematic arithmetic rules for zero but did not resolve the reciprocal case in a stable way, assigning 0/0 = 0. Bhaskara II (Bhāskara II 1150) treated a nonzero quantity divided by zero as ananta (without end), explicitly recognising the non-terminal character of the reciprocal singularity. The modern rigorous completion is geometric. Riemann (Riemann 1851) introduced the extended complex plane $\hat{\mathbb{C}} = \mathbb{C} \cup \{\infty\}$, now called the Riemann sphere, by adjoining a single point at infinity and equipping $\hat{\mathbb{C}}$ with the topology of the two-sphere S². In this completed space the map f(z) = 1/z extends continuously to all of $\hat{\mathbb{C}}$ by setting f(0) = ∞ and f(∞) = 0, and the “singularity” at z = 0 becomes a well-defined conformal automorphism (Ahlfors 1979; Conway 1978). Crucially, f is not merely large near z = 0: it is directionally structured. Writing z = re^(iθ), $$f(z) \;=\; \frac{1}{r}\,e^{-i\theta}, \label{eq:reciprocal_argument_reversal}$$ so the argument is reversed under f. The magnitude diverges while the orientation is reflected. This orientation reversal is encoded algebraically in the residue: $$\operatorname{Res}\!\left(\frac{1}{z},\, 0\right) = 1 \neq 0. \label{eq:residue}$$ A nonzero residue is the defining property of a simple pole, as opposed to an essential singularity (Ahlfors 1979). Near an essential singularity, Picard’s theorem (Picard 1879) guarantees that every complex value (with at most one exception) is achieved in any neighbourhood — the singularity is directionless in the strongest possible sense. A simple pole, by contrast, carries a definite orientation through its residue. The reciprocal map 1/z therefore does not merely diverge: it carries a specific, irremovable directional signature. Topological consequence: the punctured sphere The completion maps 0 ↦ ∞, and the point ∞ is thereby distinguished from all other points of $\hat{\mathbb{C}}$. The underlying topological consequence is immediate. Lemma 1 (Punctured sphere is not homeomorphic to the sphere). The sphere S² with a single point removed, S² \ {p}, is homeomorphic to ℝ² via stereographic projection (Ahlfors 1979), and therefore is not homeomorphic to S². In particular, the Euler characteristics differ: χ(S²) = 2 but χ(ℝ²) = 1. The point p at which the reciprocal singularity is located is therefore a topological invariant of the completed space: its removal changes the homotopy type. No continuous deformation of the space can restore S² \ {p} to S². The defect is irremovable. Left: A perfectly symmetric singularity is electromagnetically inert: S = E × H = 0 and nothing propagates. Right: A defect (argument reversal arg (1/z) =  − θ, Res  = 1 ≠ 0) breaks the symmetry; the vacuum responds via Z₀ so that S ≠ 0 and the directed process p → p^(*) begins, developing into the full field structure as depicted in Figure 2. Forced asymmetry: the Poincaré–Hopf theorem We now establish that any non-terminal directed process on S² requires exactly two distinguished singular points. A non-terminal directed process on S² is modelled by a continuous vector field v on S² that is not identically zero. The constraint on singularities of such a field is given by a classical result in differential topology. Theorem 2 (Poincaré–Hopf (Poincaré 1885; Hopf 1926)). Let M be a compact, oriented, smooth manifold without boundary, and let v be a smooth vector field on M with only isolated zeros. Then the sum of the indices of v at its zeros equals the Euler characteristic of M: ∑_(p : v(p) = 0)ind_(p)(v) = χ(M). For the two-sphere, χ(S²) = 2. Applying Theorem 2: Corollary 3 (Forced singularity structure on S²). Every smooth vector field on S² with only isolated zeros has zeros whose indices sum to 2. The minimal realisation of this constraint uses exactly two zeros, each of index  + 1: a source and a sink. No vector field on S² can be zero-free. Proof. A zero-free vector field on S² would have an empty sum of indices, equal to 0 ≠ 2 = χ(S²), contradicting Theorem 2. To achieve the sum χ(S²) = 2 with the fewest possible zeros: a single zero of index  + 2 would require a non-generic tangency; a single zero of index  + 1 leaves a residual sum of 1 ≠ 2; any configuration containing a zero of index  ≤ 0 requires compensating zeros of higher positive index to reach the sum 2, thereby increasing the total count. The unique minimum is exactly two zeros each of index  + 1, one a source and one a sink (Milnor 1965). ◻ Proposition 4 (Non-terminal reciprocal singularity forces two distinguished points). If the reciprocal singularity 1/z at z = 0 generates a non-terminal directed process on S², then the process is represented by a non-zero vector field on S². By Corollary 3, this field has at least two distinguished singular points. The minimal such configuration consists of: 1. a source singularity p at the site of the original reciprocal singularity (the image of 0 under f, which is ∞); and 2. a derived sink singularity p^(*), the antipodal point to p on S², uniquely determined by the index constraint. Neither point can be removed by continuous deformation. The system of {0, 1, process} has exactly three irreducible terms. Remark 5. The uniqueness of p^(*) as the antipodal point follows from the additional constraint that the minimal vector field on S² with the topology of a source–sink pair has its sink at the antipodal point of the source, by spherical symmetry of the base manifold (Guillemin and Pollack 1974; Milnor 1965). Any non-antipodal placement of p^(*) would break the symmetry of the remaining (non-singular) surface in a way that requires additional structure not present in the primitive system. Remark 6 (Historical continuity). Bhaskara II recognised that 1/z as z → 0 is non-terminal. Riemann supplied the geometric completion. The present argument adds the topological content: the completed space supports a non-terminal directed process only if it carries a distinguished defect, and the Poincaré–Hopf theorem requires this defect to come in a pair. The three irreducible terms — origin 0, point at infinity ∞, and the directed process between them — are not posited; they are forced by the topology of the two-sphere. Genesis of the Riemann sphere from a primitive reciprocal singularity. (i) The argument reversal arg (1/z) =  − θ and Res (1/z,0) = 1 ≠ 0 break the perfect symmetry; the Poynting vector S = E × H ≠ 0 and the directed process begins. (ii) Field lines (geodesics) propagate from p; equipotential circles accumulate near the source; the antipodal sink p^(*) is forced to appear by the Poincaré–Hopf theorem (χ(S²) = 2, Theorem 2), shown here as a hollow circle to indicate it is topologically required but not yet fully established. (iii) The complete structure: p = ∞ (source, index  + 1), p^(*) = 0 (sink, index  + 1), logarithmic potential ϕ = log (1/r), and the full complement of geodesic field lines and log-spaced equipotentials detailed in Figure 3. The defect is irremovable and intensifying The simple pole at z = 0 has residue 1 ≠ 0, which is preserved under continuous deformations of the map that do not change the topology of the domain. More precisely: the winding number of 1/z around z = 0 is  − 1, a topological invariant of the map (Ahlfors 1979; Hatcher 2002). No continuous deformation that avoids moving 0 to a different sheet can change this winding number. The defect at p is therefore: 1. Irremovable: no homeomorphism of the domain carries S² \ {p} to S². 2. Intensifying under the process: the logarithmic potential ϕ(r) = log (1/r) generated by the point source at p diverges as r → 0. As the process continues, the distinction between p and all other points of S² increases without bound. The second point invokes a standard result from potential theory: the Green’s function of the Laplacian on ℝ² (and equivalently on the punctured sphere, locally) for a unit point source at the origin is G(r) =  − (1/2π)log r (Evans 1998). This is the unique rotationally symmetric solution, confirming that the logarithmic form of the field is not assumed but forced by the geometry of a point singularity in two dimensions. Summary of Section 2. The reciprocal singularity 1/z at z = 0, when non-terminally completed on the Riemann sphere, forces: 1. a topological defect at p (the puncture), irremovable by any continuous map; 2. a unique derived antipodal point p^(*) (by the Poincaré–Hopf theorem applied to S²); 3. a logarithmic potential field between p and p^(*) (by the Green’s function of the Laplacian); 4. exactly three irreducible terms: origin (0), point at infinity (∞), and the directed process between them. All four consequences follow from standard results in complex analysis, differential topology, and potential theory. No additional postulates are required. State of the derivation at the end of Section 2. The Riemann sphere S² with the completed reciprocal map f(z) = 1/z. North pole p = ∞: the source singularity, index  + 1, carrying residue Res (1/z,0) = 1. The starburst indicates the irremovable defect; the arc θ ↦  − θ records the orientation reversal of [eq:reciprocal_argument_reversal]. South pole p^(*) = 0: the unique antipodal sink, index  + 1, forced by the Poincaré–Hopf theorem (χ(S²) = 2, minimum realisation two poles, total index 1 + 1 = 2). Red circles: equipotential surfaces of the logarithmic potential ϕ(r) = log (1/r) (Green’s function of the Laplacian in 2D), spaced logarithmically to emphasise the intensification of the field near p. Blue curves: geodesic field lines from p to p^(*), each a great circle arc. The three irreducible terms of the system — source, sink, and directed process — are all visible and none can be removed by continuous deformation. Self-similar scaling, the φ-attractor, and the origin of α The logarithmic field ϕ(r) = log (1/r) established in Section 2 generates a natural scaling structure: equipotential surfaces nest geometrically, each related to the next by a multiplicative factor in r. This section shows that the self-similar recursion of that nesting has a unique stable fixed point — the golden ratio φ — and that the irremovable defect at p forces a departure from φ whose geometric measure is α. The logarithmic field generates self-similar nesting On the punctured sphere S² \ {p}, the logarithmic potential ϕ(r) = log (1/r) assigns to each point a depth ϕ measuring its distance from p in the natural metric of the field. The equipotential surfaces at depths ϕ₁ and ϕ₂ = ϕ₁ + Δϕ are circles at radii r₁ = e^(−ϕ₁) and r₂ = e^(−ϕ₂), with ratio $$\frac{r_1}{r_2} = e^{\Delta\phi}. \label{eq:ratio}$$ For uniform step size Δϕ = Δ (a constant advance in depth), the successive radii form a geometric sequence: r, r ⋅ e^(−Δ), r ⋅ e^(−2Δ), … The field is therefore self-similar: the geometry at any depth looks identical to the geometry at every other depth, scaled by the constant factor λ = e^(−Δ). Definition 7 (Self-similar scaling ratio). The scaling ratio of the logarithmic field at uniform depth step Δ is λ = e^(−Δ). A process advancing in uniform steps of the logarithmic field is characterised by λ. The field is self-similar if and only if λ is constant across all depths. The self-referential consistency condition and φ A self-similar process that must close — returning the scaling structure to itself after each step — satisfies a consistency condition. If the ratio of one step to the next must equal the ratio of the next step to the one after, then: $$\frac{\lambda_{n+1}}{\lambda_n} \;=\; \frac{\lambda_n}{\lambda_{n-1}}. \label{eq:consistency}$$ In terms of the generating recursion, if x_(n) denotes the ratio at step n, self-consistency requires that x_(n) be expressible in terms of x_(n − 1) alone, with no accumulated memory of earlier steps. A ratio satisfying this condition at every scale simultaneously must have a periodic continued fraction expansion ξ = [a₀; a₁,a₂,…] with all partial quotients equal. The minimum such configuration, with all a_(i) = 1, gives the purely periodic continued fraction [1; 1,1,1,…], whose consistency equation is: $$x \;=\; 1 + \frac{1}{x}. \label{eq:recursion}$$ This is not canonical by convention; it is the unique minimum-information self-referential recursion consistent with the integer winding structure of the defect at p (Section 5.1). Proof. Setting x = 1 + 1/x gives x² − x − 1 = 0, whose positive root is $\varphi = (1+\sqrt{5})/2$. Uniqueness follows from the product of roots equalling  − 1 < 0: the two roots have opposite signs, so exactly one is positive. For global attraction, we exhibit a uniform contraction. The map f(x) = 1 + 1/x satisfies: 1. f maps (0,∞) to (1,∞), since f(x) > 1 for all x > 0. 2. f maps (1,∞) to (1,2), since 1 < 1 + 1/x < 2 for all x > 1. After at most two iterations from any x₀ > 0, the orbit enters (1,2) and remains there. Consider the second iterate g = f ∘ f: $$g(x) \;=\; 1 + \frac{1}{1+1/x} \;=\; \frac{2x+1}{x+1}, \qquad g'(x) \;=\; \frac{1}{(x+1)^2} \;\leq\; \tfrac{1}{4} \quad\text{on } [1,2].$$ The factor $\tfrac{1}{4}$ is a uniform contraction constant. By Banach’s contraction mapping theorem, g has a unique fixed point in [1,2], which coincides with φ; hence f^(n)(x₀) → φ for every x₀ > 0. (Rudin 1976) The local contraction rate at φ is: $$|f'(\varphi)| \;=\; \frac{1}{\varphi^2} \;=\; 2 - \varphi \;\approx\; 0.382 < 1, \label{eq:local_contraction}$$ where the identity 1/φ² = 2 − φ follows from φ² = φ + 1: setting x = 1/φ = φ − 1 gives x² = (φ−1)² = φ² − 2φ + 1 = 2 − φ. ◻ Remark 8. φ is not postulated or chosen for its aesthetic properties. It is the unique limit of the self-similar recursion forced by the logarithmic field on the punctured sphere. Any self-similar nesting process described by a ratio obeying [eq:recursion] converges to φ, regardless of starting value. Rational winding produces field cancellation The self-similar scaling structure also governs the winding of field lines around p. Each circuit around p accumulates a phase advance. If this phase is a rational multiple of 2π — that is, if the winding ratio is p/q for integers p, q — then after q circuits the field returns to exactly its initial configuration. The returning amplitude is then exactly superimposed on the incoming amplitude. Lemma 9 (Rational winding produces a resolvent pole). Let U denote the unitary operator representing one complete circuit around p, with winding ratio ξ. If ξ = p/q is rational, then U^(q) = I, and the resolvent of the q-circuit return map (I−U^(q))⁻¹ does not exist. Proof. If ξ = p/q (in lowest terms, p, q positive integers), the phase advance per circuit is 2πp/q. After q complete circuits the cumulative phase is 2πp—an integer multiple of 2π—so the state returns exactly to its starting configuration: U^(q) = e^(2πip) I = I. Therefore (I−U^(q)) = I − I = 0 identically: every state lies in the kernel of the q-circuit return map, and the resolvent (I−U^(q))⁻¹ does not exist. (Kato 1966) ◻ The physical consequence is exact resonant recurrence: after q circuits, the field returns precisely in phase with its initial configuration. Successive q-circuit passages superimpose constructively, driving unbounded resonant growth. No finite-amplitude bound configuration can persist. Proposition 10 (Matter excludes rational winding). Matter exists (this is an empirical observation requiring no apparatus beyond the fact that the observer is present). A stable bound electromagnetic configuration requires a well-defined q-circuit return resolvent (I−U^(q))⁻¹ for every q. By Lemma 9, rational winding is incompatible with this condition. Therefore the winding ratio ξ is irrational. The unique irrational: Hurwitz’s theorem and the regularity condition Irrationality of ξ is necessary but not sufficient. Infinitely many irrationals exist. The question is which irrational is selected, and whether it is selected uniquely. The relevant constraint is not a maximum or an optimum. It is a regularity condition: the resolvent (I−U)⁻¹ must be uniformly bounded not merely at one scale, but across all scales of rational approximation simultaneously. Definition 11 (Uniform regularity across scales (P10)). A winding ratio ξ satisfies the uniform regularity condition if there exists a constant C < ∞ such that ∥(I−U)⁻¹∥ ≤ C ⋅ q² for all denominators q of rational approximants p/q to ξ simultaneously. Remark 12. Definition 11 is a mathematical condition on the solution space — analogous to requiring solutions to belong to a Sobolev space or to satisfy uniform ellipticity. It is not a physical postulate about dynamics. Lemma 13 (Resolvent bound from Diophantine approximation). For winding ratio ξ, the resolvent norm satisfies: $$\left\|(I-U)^{-1}\right\| \;\leq\; \frac{1}{2\pi\,\min_{p/q}\,|\xi - p/q|}. \label{eq:resolvent_bound}$$ Proof. The operator U has spectrum on the unit circle. At eigenvalue e^(2πiξ), the distance from 1 is |1−e^(2πiξ)| = 2|sin(πξ)|. For ξ near p/q, this equals 2|sin(π(ξ−p/q))| ≈ 2π|ξ−p/q| for small |ξ−p/q|. The resolvent norm is the reciprocal of the spectral gap. (Kato 1966) ◻ Theorem 14 (φ is uniquely selected by the regularity condition). Among all irrational winding ratios ξ > 0, the golden ratio $\varphi = (1+\sqrt{5})/2$ is the unique value satisfying the uniform regularity condition of Definition 11 with minimum constant $C = \sqrt{5}/(2\pi)$. Proof. By Lemma 13 and equation [eq:resolvent_bound]: $$\left\|(I-U)^{-1}\right\| \;\leq\; \frac{q^2}{2\pi\,|\xi - p/q| \cdot q^2} \;\leq\; \frac{q^2}{2\pi \cdot \mathrm{Dioph}(\xi)},$$ where Dioph(ξ) = inf_(p/q) q²|ξ−p/q| is the Diophantine approximation constant. By Hurwitz’s theorem (Hurwitz 1891): for every irrational ξ, infinitely many rationals p/q satisfy $|\xi - p/q| < 1/(\sqrt{5}\,q^2)$, so $\mathrm{Dioph}(\xi) \leq 1/\sqrt{5}$ for all irrationals. Furthermore, $1/\sqrt{5}$ is sharp: the only irrational for which the inequality $|\xi - p/q| \geq 1/(\sqrt{5}\,q^2)$ holds for all p/q simultaneously is ξ = φ (and equivalently 1/φ, since φ = 1 + 1/φ). For every ξ ≠ φ, there exists a denominator q₀ at which $|\xi - p/q_0| < 1/(\sqrt{5}\,q_0^2)$, violating the Hurwitz bound at that scale. Therefore: for ξ = φ, the resolvent satisfies $\|(I-U)^{-1}\| \leq \sqrt{5}\,q^2/(2\pi)$ uniformly for all q — the uniform regularity condition holds with $C = \sqrt{5}/(2\pi)$. For any ξ ≠ φ, the bound $\|(I-U)^{-1}\| \leq \sqrt{5}\,q^2/(2\pi)$ fails at scale q₀: the uniform regularity condition with $C = \sqrt{5}/(2\pi)$ fails for all ξ ≠ φ. Therefore ξ = φ is the unique solution. ◻ Remark 15. The numerical verification confirms the sharpness of the Hurwitz bound for φ. The convergents F_(n + 1)/F_(n) (Fibonacci ratios) achieve $|\varphi - F_{n+1}/F_n| \approx 1/(\sqrt{5}F_n^2)$ to high precision for all n, confirming that φ saturates the bound at every scale simultaneously — it is maximally resistant to rational approximation everywhere. The irremovable defect forces a departure from φ Theorem 14 establishes that a self-similar field on a perfect sphere would require φ as its unique winding ratio. However, the space is not a perfect sphere. By Proposition 4 of Section 2, the space is S² with an irremovable defect at p. This defect modifies the self-similar scaling in a precise way. Proposition 16 (The defect perturbs the self-similar ratio). Let ξ_(ideal) = φ denote the winding ratio of a self-similar field on a perfect sphere. The irremovable defect at p — characterised by Res (1/z,0) = 1 ≠ 0 and winding number  − 1 around p — introduces a perturbation to the effective scaling. The actual winding ratio ξ of the field on S² \ {p} satisfies: ξ = φ − δ,   δ > 0, where δ is determined by the geometric relationship between the local structure at p and the global structure of S². Proof. The defect at p contributes a holonomy to every circuit around p: parallel transport of a vector around p returns the vector with a definite angular deficit. This holonomy is a geometric invariant of the puncture, determined by the residue and winding number of 1/z at z = 0. On a sphere without the defect, the self-similar recursion drives ξ → φ by Lemma [lem:phi_fixed]. With the defect, each circuit around p accumulates holonomy from the puncture. The sign of this holonomy is determined by the winding number of 1/z around z = 0, which equals  − 1 (Section 2, equation [eq:reciprocal_argument_reversal]: the argument reversal θ ↦  − θ is precisely a winding number of  − 1). A circuit around p therefore accumulates negative phase relative to the unperturbed case, reducing the effective winding ratio below the ideal value. Hence ξ < φ and δ = φ − ξ > 0. The sign of the departure is thereby established: δ > 0, meaning the actual winding ratio on S² \ {p} falls strictly below the ideal φ. The exact value of δ is derived in Section 5, Theorem 37, by assembling four established facts about the vacuum structure. The geometric picture here—a departure set by the ratio of local to global field energy on the punctured sphere—and the electromagnetic assembly of Section 5 are two descriptions of the same dimensionless quantity; their consistency is confirmed by exact numerical agreement with the measured value of α. ◻ Definition 17 (The fine-structure constant as the departure). The fine-structure constant α is the dimensionless geometric departure [eq:departure]: α ≡ δ = φ − ξ_(actual), where ξ_(actual) is the winding ratio of the self-similar field on S² \ {p} and φ is the winding ratio on the ideal sphere. Equivalently, α is the ratio of the vacuum’s field impedance $Z_0 = \sqrt{\mu_0/\varepsilon_0}$ to twice the quantum of orbital resistance R_(K) = h/e²: $$\alpha \;=\; \frac{Z_0}{2R_K}, \label{eq:alpha_SI}$$ which is exact in the 2019 SI system (Bureau International des Poids et Mesures 2019). The two expressions are two descriptions of the same geometric object: equation [eq:alpha_departure] is the topological description and equation [eq:alpha_SI] is the electromagnetic description. Remark 18. The identification of equations [eq:alpha_departure] and [eq:alpha_SI] is the content of the remaining sections. Section 4 shows that Z₀ is the primitive 0/1 geometric ratio of the electromagnetic vacuum, and that R_(K) = h/e² is the quantum of the same ratio at the scale of closed electromagnetic loops. The factor of 2 in both α = Z₀/2R_(K) and K_(J) = 2e/h (the Josephson constant) is the minimum winding number — the same topological factor established in this section. Summary of Section 3. The logarithmic field on S² \ {p} has a natural self-similar scaling structure whose recursion [eq:recursion] converges to φ on the ideal sphere. The uniform regularity condition (Definition 11), together with Hurwitz’s theorem (1891), establishes that φ is the uniquely forced winding ratio: no other irrational satisfies the uniform resolvent bound at all scales simultaneously. Rational winding is excluded by the existence of matter (Proposition 10). The irremovable defect at p forces a departure α from the ideal φ. This departure is the fine-structure constant. All four consequences follow from the topology of the punctured sphere, Hurwitz’s theorem (1891), and one empirical observation (matter exists). No additional postulates are required. The electromagnetic vacuum as primitive structure Section 2 established that the reciprocal singularity forces a punctured sphere with two distinguished poles and a logarithmic potential field. Section 3 showed that self-similar scaling on this space converges uniquely to φ, and that the irremovable defect forces a departure α. This section identifies the physical content of that geometry. We show that Maxwell’s two vacuum constants ε₀ and μ₀ are the primitive quantities 1 and 0 of the structure, that $Z_0 = \sqrt{\mu_0/\varepsilon_0}$ is the geometric 0/1 ratio instantiated as the impedance of free space, and that the quantum of orbital resistance R_(K) = h/e² is the same ratio at the scale of closed electromagnetic loops. The fine-structure constant α = Z₀/2R_(K) is then the ratio of the vacuum’s continuous 0/1 expression to twice the minimum quantum of that same ratio — the factor of 2 being the minimum winding number established in Section 3. The two vacuum constants as primitive quantities Maxwell’s equations in vacuum, written in the absence of free charges and currents, take the form (Maxwell 1865; Jackson 1999): $$\label{eq:maxwell} \begin{align} \nabla \cdot \mathbf{E} &= 0, \label{eq:divE}\\ \nabla \cdot \mathbf{B} &= 0, \label{eq:divB}\\ \nabla \times \mathbf{E} &= -\mu_0\,\frac{\partial \mathbf{H}}{\partial t}, \label{eq:curlE}\\ \nabla \times \mathbf{H} &= \varepsilon_0\,\frac{\partial \mathbf{E}}{\partial t}. \label{eq:curlH} \end{align}$$ The two constants ε₀ and μ₀ enter the curl equations [eq:curlE] and [eq:curlH] separately, with structurally distinct roles: 1. ε₀ (vacuum permittivity) governs the capacity of the vacuum to contain electric field energy. It is the measure of how strongly the vacuum holds, stores, and constrains the electric mode. We identify this as the geometric primitive 1: bounded, containing, the quantity whose boundary is what makes the singularity possible. 2. μ₀ (vacuum permeability) governs the openness of the vacuum to magnetic field penetration. It is the measure of the vacuum’s receptivity — how readily it admits the passage and coupling of the magnetic mode. We identify this as the geometric primitive 0: open, penetrating, the quantity whose action initiates the 1/0 process. The identification of ε₀ with the primitive 1 and μ₀ with the primitive 0 is not a declaration made here. It is a recognition established as definitional in Section 5, Theorem 34: the physical definitions of these constants coincide with the mathematical definitions of the primitives, and no additional structure is introduced. Theorem 19 (Non-injectivity of the compression to c²). The mapping $$\Phi: (\varepsilon_0, \mu_0) \;\longmapsto\; c^2 = \frac{1}{\mu_0\varepsilon_0} \label{eq:compression}$$ is exact for all propagation queries but many-to-one: for any fixed c, infinitely many pairs (ε₀,μ₀) satisfy c² = 1/μ₀ε₀, distinguished by the ratio μ₀/ε₀. The compression retains the product (dispersion) while identifying all pairs with the same product (partition). The suppressed coordinate is: $$Z_0 \;=\; \sqrt{\frac{\mu_0}{\varepsilon_0}}. \label{eq:Z0}$$ The pair (c,Z₀) recovers (ε₀,μ₀) uniquely: $$\mu_0 = \frac{Z_0}{c}, \qquad \varepsilon_0 = \frac{1}{Z_0\,c}. \label{eq:reconstruction}$$ Proof. For fixed c² = 1/μ₀ε₀, the locus of admissible pairs is {(ε₀,μ₀) : μ₀ = 1/(c²ε₀)}, a one-parameter family parameterised by ε₀ > 0. Different values of ε₀ give different ratios μ₀/ε₀ = 1/(c²ε₀²). The map Φ is therefore not injective. The reconstruction [eq:reconstruction] follows from c² = 1/μ₀ε₀ and Z₀² = μ₀/ε₀, which together resolve both constants uniquely. (K. B. Heaton and Coherence Research Collaboration 2026b) ◻ Remark 20. Every formalism that passes to the wave equation ∇²E = μ₀ε₀ ∂_(t)²E and sets c = 1 performs the compression [eq:compression] and discards Z₀ as the independent geometric degree of freedom. The suppressed coordinate is the ratio of the primitive 0 to the primitive 1 — it is the geometric content of the 0/1 structure, not merely a convenient engineering parameter. (Heaviside 1893; Kitano 2009). See also (Kelly B. Heaton and Claude Sonnet 4.6 2026). Z₀ as the 0/1 ratio of the electromagnetic vacuum Proposition 21 (Z₀ is the primitive 0/1 geometric ratio). The vacuum impedance $Z_0 = \sqrt{\mu_0/\varepsilon_0}$ is the ratio of the primitive 0 (μ₀, openness) to the primitive 1 (ε₀, containment), expressed as the impedance of free space. It is not derived from or imported into the geometric framework: it is the geometric framework, instantiated as a measurable property of the vacuum. The identification has the following consequences: 1. A propagating photon, which carries equal electric and magnetic energy at every point, has field ratio |E|/|H| = Z₀ exactly. It is a pure expression of the primitive 0/1 ratio, liberated from the topological scar at p. 2. The bound orbital, which has |E|/|H| = Z₀ n/α ≫ Z₀, is deeply mismatched to the vacuum ratio at every n ≥ 1. The scar at p holds the orbital far from the free-vacuum ratio. 3. The gap at IE/4 is the spectroscopic record of the unique transition energy at which the bound field attempts to achieve |E|/|H| = Z₀ — a configuration simultaneously required by the impedance match and forbidden by the closure condition. Proof. Statement (i) follows from the plane-wave solution to Maxwell’s equations: for a plane wave propagating in the k̂ direction, H = (k̂×E)/Z₀, giving |E|/|H| = Z₀ identically. (Jackson 1999) Statement (ii) is established in §4.3 below. Statement (iii) follows from statements (i) and (ii) together with the Diophantine impossibility of §4.4. ◻ The orbital field impedance from Coulomb and Biot–Savart We now derive the field impedance of a bound electromagnetic closure mode at principal quantum number n, without quantum mechanical postulates beyond the closure velocity. Theorem 22 (Orbital impedance). Consider the electromagnetic closure mode at principal quantum number n, with closure velocity v_(n) = αc/n. From Coulomb’s law and the Biot–Savart law, the ratio of electric to magnetic field amplitude at the orbital radius satisfies: $$Z_{\mathrm{orb}}(n) \;\equiv\; \frac{|\mathcal{E}_n|}{|\mathcal{H}_n|} \;=\; \frac{Z_0\,n}{\alpha}. \label{eq:Zorb}$$ No free parameters and no quantum mechanical input beyond v_(n) = αc/n enter this derivation. Proof. Let r_(n) = n²a₀ be the orbital radius and v_(n) = αc/n the closure velocity (confirmed independently by the quantum mechanical expectation value ⟨v⟩_(n) = αc/n from the virial theorem applied to exact hydrogen wavefunctions (Bethe and Salpeter 1957)). The orbital current generated by the closure mode is: $$I_n = \frac{e\,v_n}{2\pi r_n} = \frac{e\,\alpha c}{2\pi n^3 a_0}. \label{eq:current}$$ The radial Coulomb electric field at r_(n) is: $$|\mathcal{E}_n| = \frac{e}{4\pi\varepsilon_0 r_n^2} = \frac{e}{4\pi\varepsilon_0 n^4 a_0^2}. \label{eq:Efield}$$ The azimuthal magnetic field at the loop centre, by the Biot–Savart law, is: $$|\mathcal{H}_n| = \frac{I_n}{2 r_n} = \frac{e\,\alpha c}{4\pi n^5 a_0^2}. \label{eq:Hfield}$$ The impedance ratio is independent of both r_(n) and e: $$Z_{\mathrm{orb}}(n) = \frac{|\mathcal{E}_n|}{|\mathcal{H}_n|} = \frac{1}{\varepsilon_0\,v_n} = \frac{\mu_0 c^2}{v_n} = \frac{\mu_0 c^2 n}{\alpha c} = \frac{\mu_0 c\, n}{\alpha} = \frac{Z_0\, n}{\alpha}, \label{eq:Zorb_derivation}$$ using μ₀c = Z₀, which follows directly: $\mu_0 c = \mu_0/\sqrt{\mu_0\varepsilon_0} = \sqrt{\mu_0^2/(\mu_0\varepsilon_0)} = \sqrt{\mu_0/\varepsilon_0} = Z_0$. Numerical verification: for n = 1, Z_(orb)(1) = Z₀/α ≈ 376.730/7.2974 × 10⁻³ ≈ 51 625.6 Ω, consistent with direct calculation. (K. B. Heaton and Coherence Research Collaboration 2026c) ◻ Corollary 23. Every bound orbital satisfies Z_(orb)(n) ≥ Z₀/α ≈ 137 Z₀ ≫ Z₀. No bound orbital is impedance-matched to the vacuum. Photon emission is mode conversion from the high-impedance bound configuration to the matched free-propagating configuration with |E|/|H| = Z₀. The impedance-match boundary at IE/4 The impedance match condition Z_(orb) = Z₀ is never satisfied at any individual orbital. It is satisfied, however, at a specific transition energy in the spectrum. Theorem 24 (Universal gap at IE/4). The transition energy ΔE = IE/4 is the unique energy at which the electromagnetic closure mode would achieve Z_(orb) = Z₀. No bound-to-bound transition can carry this energy. The spectroscopic consequence is a universal structural gap: a zero-observation interval at IE/4 in every ion’s transition spectrum. Proof. Part 1: Why IE/4. The ionization energy satisfies IE = E₀Z²μ̂ where E₀ is the Rydberg energy. The ionization energy of the n = 2 shell is IE/4 exactly, by the hydrogen level formula E_(n) = E₀/n². A transition carrying energy IE/4 would take the field from a bound closure mode to the n = 2 continuum threshold in a single step, passing through the configuration where Z_(orb) = Z₀ — neither fully bound nor fully propagating. Part 2: Diophantine impossibility. No bound-to-bound hydrogen transition carries energy IE/4 because no positive integers (n_(f),n_(i)) with n_(f) < n_(i) satisfy: $$\frac{1}{n_f^2} - \frac{1}{n_i^2} = \frac{1}{4}. \label{eq:diophantine}$$ The three cases are exhaustive: 1. n_(f) = 1: requires n_(i)² = 4/3, not an integer. 2. n_(f) = 2: requires n_(i) → ∞, not a bound state. 3. n_(f) ≥ 3: maximum transition energy E₀/9 < E₀/4. No solution exists. The gap is therefore not a selection rule: it is a number-theoretic impossibility, valid regardless of orbital angular momentum or atomic species. Part 3: Universality. For multi-electron ions, screening and configuration interaction modify the absolute energies but not the ratio IE/4, since IE is defined as the first ionisation energy of the ion and the impedance-match boundary scales with IE exactly. The nuclear charge Z cancels in the ratio ΔE/IE. The gap is therefore universal across all ions and all subshell types. Empirical confirmation. Direct test of the NIST Atomic Spectra Database confirms zero-observation at IE/4 across 71 ions spanning Z = 1–82, all four subshell families, and multiple ionisation stages, with zero falsifications (K. B. Heaton and Coherence Research Collaboration 2026c). The coordinate κ_(q) = log 4/| logα| ≈ 0.2818 places the impedance boundary at the same depth in every spectrum. ◻ R_(K), K_(J), and the quantum of the 0/1 ratio The preceding subsections established Z₀ as the continuous vacuum expression of the 0/1 primitive ratio. We now show that the quantum scale has two complementary expressions of the same ratio, and that neither h nor e is an independent primitive. R_(K) = h/e² as the quantum orbital resistance Definition 25 (von Klitzing constant). The von Klitzing constant R_(K) = h/e² (exact in the 2019 SI system (Bureau International des Poids et Mesures 2019; von Klitzing, Dorda, and Pepper 1980)) is the quantum of resistance. In the same Ohm’s-law structure as Z₀: - h (Planck’s constant, action) plays the role of the bounded containing quantity: the minimum action required to complete one closed electromagnetic loop. This is the quantum-scale “1”. - e² (charge squared) plays the role of the penetrating quantity: the coupling strength of the charge through the loop. This is the quantum-scale “0” squared. - R_(K) = h/e² is the resistance of one closed loop: the quantum 0/1 ratio. From α = Z₀/2R_(K) (exact, 2019 SI (Bureau International des Poids et Mesures 2019)): $$R_K \;=\; \frac{Z_0}{2\alpha}. \label{eq:RK_geometric}$$ R_(K) is therefore entirely determined by Z₀ (the continuous vacuum 0/1 ratio) and α (the departure from ideal φ scaling established in Section 3). It is not an independent primitive. The factor of 2 in equation [eq:RK_geometric] is the minimum winding number: the 2π closure established in Section 3 appears here as the factor distinguishing one full loop (2π) from a half-loop (π). K_(J) = 2e/h as the quantum dual Definition 26 (Josephson constant). The Josephson constant K_(J) = 2e/h (exact in the 2019 SI (Bureau International des Poids et Mesures 2019; Josephson 1962)) is the frequency-to-voltage ratio of the Josephson effect: a junction maintained at voltage V oscillates at frequency f = K_(J) ⋅ V. Theorem 27 (K_(J) and R_(K) are dual expressions of the quantum 0/1 ratio; h and e are not primitive). The constants R_(K) and K_(J) are related by: $$K_J \cdot R_K \;=\; \frac{2}{e}, \label{eq:KJ_RK}$$ which resolves both h and e exactly: $$h \;=\; \frac{4}{K_J^2\,R_K}, \qquad e \;=\; \frac{2}{K_J\,R_K}. \label{eq:h_e_resolved}$$ Neither h nor e is an independent primitive of the electromagnetic structure; both are determined by the two complementary quantum expressions of the 0/1 ratio. Proof. K_(J) ⋅ R_(K) = (2e/h)(h/e²) = 2/e, establishing [eq:KJ_RK]. From [eq:KJ_RK]: e = 2/(K_(J)R_(K)). Substituting into h = R_(K)e² = R_(K) ⋅ 4/(K_(J)²R_(K)²) = 4/(K_(J)²R_(K)). Numerical verification: h = 4/(K_(J)²R_(K)) = 6.626 070 × 10⁻³⁴ J⋅s (exact against 2019 SI); e = 2/(K_(J)R_(K)) = 1.602 177 × 10⁻¹⁹ C (exact against 2019 SI). (Bureau International des Poids et Mesures 2019) ◻ Corollary 28 (Charge is geometric). From Theorem 27 and R_(K) = Z₀/2α: $$e \;=\; \frac{4\alpha}{Z_0\,K_J}. \label{eq:e_geometric}$$ The elementary charge e is determined by α (the departure from φ scaling), Z₀ (the vacuum 0/1 ratio), and K_(J) (the quantum dual ratio). It is not a free parameter of the electromagnetic vacuum. Numerical verification: 4α/(Z₀K_(J)) = 1.602 177 × 10⁻¹⁹ C, exact. Remark 29 (The factor of 2 is the minimum winding number). The factor of 2 appears identically in both: $$\alpha = \frac{Z_0}{2R_K}, \qquad K_J = \frac{2e}{h}. \label{eq:factor_two}$$ In both cases, this 2 is the minimum winding number from the 2π closure condition of Section 3. In the Josephson effect, the factor 2e corresponds to the Cooper pair — the minimum closed winding of the superconducting order parameter carries charge 2e, completing exactly one circuit of the order-parameter phase. The same topological factor that appears in α = Z₀/2R_(K) (one full loop = 2π) appears in K_(J) = 2e/h (one full winding = 2e). This is not a coincidence; it is the same topological content expressed at the continuous vacuum level and the quantum discrete level. (Josephson 1962; Tinkham 1996) Planck’s constant as the cleaving invariant Theorem 30 (h is the invariant of the self-similar cleaving process). On the punctured sphere, the logarithmic field generates two distinct α-scaling relations, empirically confirmed for hydrogen: $$\begin{aligned} E(\kappa) &= E_0\,\alpha^\kappa \quad\text{(energy ladder)}, \label{eq:energy_ladder}\\ \nu(\kappa) &= \nu_0\,\alpha^\kappa \quad\text{(thread law, $\beta = \log_{10}\alpha$, $r^2 = 1.000$, $9$ pairs)}. \label{eq:thread_law} \end{aligned}$$ The ratio E(κ)/ν(κ) is independent of κ: $$\frac{E(\kappa)}{\nu(\kappa)} = \frac{E_0\,\alpha^\kappa}{\nu_0\,\alpha^\kappa} = \frac{E_0}{\nu_0} \;=\; h. \label{eq:planck_invariant}$$ Planck’s constant h is the invariant preserved at every cleaving event — the quantity that cannot change because both energy and frequency scale identically with α^(κ). The constancy of h across all transitions is the self-similarity of the process expressed as a ratio. Proof. Equations [eq:energy_ladder] and [eq:thread_law] are empirically confirmed. The cancellation of α^(κ) in the ratio is exact algebra. The identification E₀/ν₀ = h follows from the definitions of the Rydberg energy and the corresponding frequency. (K. B. Heaton and Coherence Research Collaboration 2026a, 2026c) ◻ Summary of Section 4. The vacuum constants ε₀ and μ₀ are the geometric primitives 1 (containment) and 0 (openness); their ratio $Z_0 = \sqrt{\mu_0/\varepsilon_0}$ is the 0/1 primitive ratio expressed as the impedance of free space. This identification is definitional, not axiomatic, as established in Section 5. The orbital impedance Z_(orb)(n) = Z₀n/α, derived from Coulomb’s law and the Biot–Savart law with no free parameters, is always much larger than Z₀: no bound orbital is impedance-matched to the vacuum. The impedance-match condition Z_(orb) = Z₀ corresponds to the transition energy ΔE = IE/4, which no bound-to-bound transition can carry — a consequence of number theory (no integer solution to equation [eq:diophantine]) rather than a selection rule, confirmed across 71 ions with zero falsifications. At the quantum scale, R_(K) = h/e² and K_(J) = 2e/h are dual expressions of the same 0/1 ratio: together they determine h and e exactly, neither is a primitive of the structure, and both carry the same factor of 2 as α = Z₀/2R_(K) — the minimum winding number from the 2π closure condition of Section 3. Planck’s constant h is the invariant of the self-similar cleaving process. Closure: definitional identity and the necessity of α Sections 2–4 establish, respectively: the punctured sphere with its irremovable defect and logarithmic potential field; the unique φ-attractor of self-similar scaling and the departure δ forced by the defect; and the electromagnetic realisation of the 0/1 geometric structure including the orbital impedance, the universal gap at IE/4, and the quantum dual constants R_(K) and K_(J). Three questions remain open: 1. Gap 1. The self-referential recursion x = 1 + 1/x was motivated in Section 3 by continued fraction structure, but the Markov spectrum theorem — which makes the argument rigorous — was deferred. 2. Gap 2. The departure δ was proved to exist and be positive, but its value was identified with α = Z₀/2R_(K) by citing the 2019 SI identity rather than deriving it from the geometry; and the identification of ε₀, μ₀ with the mathematical primitives 1, 0 was asserted rather than established. 3. Gap 3. The Josephson constant K_(J) = 2e/h was cited as an established experimental fact; its topological necessity was stated but not formally derived. This section closes all three gaps. The central result is that no additional structure is introduced at any step: the physical constants ε₀, μ₀, c, Z₀, R_(K), and K_(J) are not imported into the mathematical argument — they are the mathematical argument, expressed in physical vocabulary. Gap 1: the Markov spectrum theorem forces x = 1 + 1/x The logarithmic field ϕ(r) = log (1/r) assigns a depth coordinate to every point of S² \ {p}. The self-similar scaling ratio at depth γ is the ratio of successive equipotential radii: λ = r(γ)/r(γ+Δ), where Δ is the uniform step. For the scaling to be self-similar across all depths simultaneously — not merely at any single depth — the ratio λ must be expressible as a function of itself alone at every scale. This is the defining property of a continued fraction. Lemma 31 (Self-similar scaling requires a continued fraction ratio). A scaling ratio ξ > 0 is self-similar across all depths of the logarithmic field simultaneously if and only if ξ has a purely periodic continued fraction expansion ξ = [a₀; a₁,a₂,…], where the partial quotients {a_(i)} are constant at every level. The unique ratio satisfying this condition with minimum partial quotients (all a_(i) = 1) is $$\varphi = [1;\,1,1,1,\ldots] = \frac{1+\sqrt{5}}{2}. \label{eq:phi_cf}$$ Proof. A ratio ξ is self-similar across all depths simultaneously if and only if the description of the ratio at any depth requires the same information as the description at every other depth. This is the statement that ξ satisfies a recursion of the form ξ = a₀ + 1/ξ′ where ξ′ has the same description as ξ — the definition of a purely periodic continued fraction. For the recursion to be consistent with the integer winding structure of the defect at p (which requires integer partial quotients), the minimum non-trivial case is all a_(i) = 1, giving ξ = 1 + 1/ξ, whose positive solution is φ. That φ = [1;1,1,1,…] is uniquely characterised by minimum partial quotients is the content of the Markov spectrum theorem (Markov 1880): the Lagrange value $$L(\xi) \;=\; \limsup_{q\to\infty} \frac{1}{q^2\,|\xi - p/q|}$$ satisfies $L(\xi) \leq \sqrt{5}$ for all irrationals ξ, with equality $L(\varphi) = \sqrt{5}$ achieved uniquely by φ (and 1/φ). The partial quotients of φ are uniformly 1 at every scale; for every other irrational, at least one partial quotient exceeds 1, introducing scale-dependent asymmetry inconsistent with uniform self-similarity. ◻ Corollary 32. The recursion x = 1 + 1/x is the unique self-referential consistency condition for a ratio with minimum uniform structure across all scales of the logarithmic field. It is forced by the Markov spectrum theorem. Gap 1 is closed. Gap 2: why the vacuum must have these primitives Before establishing that ε₀ and μ₀ are the mathematical primitives, we must address a prior question: why must the vacuum have any primitives of type 1 and 0 at all? The answer is already in the proof. Proposition 33 (A non-terminal process requires both primitives). The existence of matter (Proposition 10) implies a non-terminal reciprocal singularity 1/0. A non-terminal reciprocal singularity requires: 1. a containing quantity (the primitive 1): something that can be divided, bounded, and entered — the denominator of the reciprocal relation; and 2. a penetrating quantity (the primitive 0): the quantity that enters and drives the process toward infinity — the numerator that approaches zero. Both are necessary for the relation 1/0 to be non-trivially defined. A vacuum that lacked either would support neither the reciprocal singularity nor the process it generates. Proof. The reciprocal singularity 1/0 requires a denominator (which must be capable of being divided: a bounded, containing quantity) and a numerator approaching zero (which must be capable of penetrating: an open, admitting quantity). Remove the containing quantity and the denominator has no boundary to define a singularity. Remove the penetrating quantity and no process initiates. Both are constitutive of the relation. Since the non-terminal process exists (matter exists), both primitives must be realised in the physical vacuum. ◻ The question is then: which physical constants play these roles? The answer is established not by declaration but by the content of the definitions themselves. Theorem 34 (Primitive Identification — the primitives are definitional, not axiomatic). The identification ε₀ ↔ 1,   μ₀ ↔ 0 is not an additional structure imposed on the mathematical framework. It is the content of the physical definitions of these constants. Proof. The vacuum permittivity ε₀ is defined as the constant governing the capacity of the vacuum to permit, contain, and store electric field energy. Its defining role is the capacity for containment: the bounded, holding, finite-extent property. This is precisely what the mathematical primitive 1 means in the present framework. ε₀ does not represent the primitive 1; by its own physical definition, it is the containing quantity of the vacuum. The vacuum permeability μ₀ is defined as the constant governing the permeability of the vacuum — its openness to penetration, the ease with which the magnetic field passes through. Its defining role is openness to entry. This is precisely what the mathematical primitive 0 means: the open, penetrating, admitting quantity. μ₀ is, by its own physical definition, the open quantity of the vacuum. The identification [eq:defn_identity] therefore requires no external declaration. It is the recognition that the physical definitions of ε₀ and μ₀ coincide with the mathematical definitions of 1 and 0. The map from the mathematical primitive system to the physical vacuum constants is injective at each point — 1 ↦ ε₀ and 0 ↦ μ₀ — and no fiber variation is possible because the definitions are identical. In the language of (K. B. Heaton and Coherence Research Collaboration 2026b): the query “what is ε₀?” is not fiber-sensitive under the compression (ε₀,μ₀) ↦ c², not because of any closure rule, but because the answer is determined by the physical definition of ε₀ itself, independently of c. The definitions are the structure; no quotient is involved. ◻ Corollary 35 (The constancy of c within this framework). Within the mathematical framework established by Theorem 34, the propagation speed $c = 1/\sqrt{\varepsilon_0\mu_0}$ is constant: the primitives 0 and 1 are fixed and unique, therefore their product ε₀μ₀ is fixed, and therefore c is fixed. Proof. By Theorem 34, ε₀ IS the mathematical primitive 1 and μ₀ IS the mathematical primitive 0. These have fixed, unique values by definition of what a primitive is: 1 is not a variable and neither is 0. Therefore ε₀μ₀ is fixed, and $c = 1/\sqrt{\varepsilon_0\mu_0}$ is fixed. ◻ Remark 36 (Scope of Corollary 35). The claim is precisely this: within the framework where ε₀ and μ₀ are the fixed mathematical primitives, c is fixed. This is not a derivation of special relativity from arithmetic, nor a claim that all possible universes must have a fixed c. It is the observation that the same mathematical structure that forces α also fixes c: both are consequences of the primitives being what they are. Special relativity takes c as a postulate; the present framework identifies the deeper source of that constancy. Theories proposing a varying speed of light correspond, in this language, to theories in which the primitives themselves vary — which requires the supply of additional structure not present in the primitive system. Gap 2: the departure δ equals Z₀/2R_(K) With Theorem 34 established, the departure δ from φ can be derived rather than identified. Theorem 37 (The departure is α = Z₀/2R_(K)). The departure δ = φ − ξ_(actual) of the self-similar winding ratio from its ideal value, forced by the irremovable defect at p, equals: $$\delta \;=\; \frac{Z_0}{2R_K} \;=\; \alpha. \label{eq:alpha_derived}$$ This is not an identification from measurement. It is a consequence of four facts, each established in the preceding sections. Proof. Fact 1. The defect at p has residue Res (1/z,0) = 1 (Section 2, equation [eq:residue]). Exactly one unit of the vacuum’s fundamental structure is implicated per circuit around p. Fact 2. By Theorem 34, ε₀ and μ₀ are the mathematical primitives 1 (containment) and 0 (openness) by their own physical definitions; no additional structure is introduced. Their ratio $Z_0 = \sqrt{\mu_0/\varepsilon_0}$ is therefore the primitive 0/1 geometric ratio of the vacuum, definitionally. By Theorem 19, Z₀ is precisely the degree of freedom suppressed by the compression (ε₀,μ₀) ↦ c²: it is independent of c and cannot be recovered from c alone. Therefore “one unit of the vacuum’s fundamental structure” is one unit of Z₀. Fact 3. The minimum quantum of the closed process is R_(K) = h/e² (Section 4.5), where h is the invariant of the self-similar cleaving process (Theorem 30) and e is the minimum topological charge of the defect. By the Dirac quantization condition (Dirac 1931) on the punctured sphere: a charge circling the defect at p once satisfies eg = nℏc/2 for integer n; the minimum case n = 1 establishes e as the fundamental charge quantum of the defect. Therefore R_(K) = h/e² is the minimum resistance of one closed circuit. Fact 4. The minimum winding number is 2 (Section 3, 2π closure condition), the factor distinguishing one full loop from a half-loop, appearing in the denominator of α = Z₀/2R_(K). Assembly. The departure δ per unit winding is the ratio of one unit of the vacuum’s structure (Z₀, by Facts 1 and 2) to twice the minimum quantum of the closed circuit (2R_(K), by Facts 3 and 4): $$\delta \;=\; \frac{Z_0}{2R_K} \;=\; \frac{\sqrt{\mu_0/\varepsilon_0}}{2h/e^2} \;=\; \frac{e^2}{2\varepsilon_0 hc} \;=\; \alpha. \label{eq:delta_assembly}$$ The step from the second to the third expression uses $\sqrt{\mu_0/\varepsilon_0} = 1/(\varepsilon_0 c)$ (equivalently Z₀ = 1/(ε₀c)), which follows from $c = 1/\sqrt{\mu_0\varepsilon_0}$: multiplying numerator and denominator of $\sqrt{\mu_0/\varepsilon_0}$ by $\sqrt{\varepsilon_0}$ gives $\sqrt{\mu_0\varepsilon_0}/\varepsilon_0 = (1/c)/\varepsilon_0 = 1/(\varepsilon_0 c)$. The final step δ = α is a definitional identification. The entire chain is an algebraic identity given the definitions. No measurement is introduced. Gap 2 is closed. ◻ Remark 38 (No quotient is introduced). Every element of [eq:delta_assembly] is either definitional (Z₀, by Theorem 34) or derived from the geometric structure (h from Theorem 30; e from the Dirac condition; 2 from the minimum winding number). The identification of δ with α is therefore not a quotient in the sense of (K. B. Heaton and Coherence Research Collaboration 2026b): it is a derived identity, not a closure rule imposed to collapse fiber variation. Gap 3: the Josephson constant from minimum winding Theorem 39 (The Josephson phase equation is topologically necessary). The Josephson constant K_(J) = 2e/h follows necessarily from the Aharonov–Bohm phase accumulated by the minimum closed winding of the self-similar process. Proof. A charge q moving through potential V for time t accumulates Aharonov–Bohm phase (Aharonov and Bohm 1959): $$\Phi_{\mathrm{AB}} = \frac{qVt}{\hbar}. \label{eq:AB_phase}$$ The minimum closed winding of the superconducting order parameter carries charge 2e (the Cooper pair). This is topologically forced: the order parameter ψ = |ψ|e^(iϕ) is bilinear in the electron field (Tinkham 1996), so one winding of ψ corresponds to one winding of the pair, carrying charge 2e. The factor 2 is the minimum winding number of Section 3: one closed circuit of the self-similar process requires charge 2e, not e. Setting q = 2e in [eq:AB_phase]: $$\frac{d\Phi_{\mathrm{AB}}}{dt} = \frac{2eV}{\hbar} = 2\pi K_J V, \label{eq:josephson_phase}$$ giving K_(J) = 2e/h. The derivation uses only the Aharonov–Bohm phase (consequence of minimal electromagnetic coupling), the minimum winding charge 2e (forced by bilinear order parameter structure and the minimum winding number), and the invariants h and e established in the preceding sections. No free parameters enter. Gap 3 is closed. ◻ Corollary 40 (The factor 2 is topologically unified). The factor 2 appearing in both α = Z₀/2R_(K) and K_(J) = 2e/h is the same topological factor throughout: the minimum winding number of the self-similar process on the punctured sphere, now seen to be unified across the continuous vacuum level and the quantum discrete level. The main theorem: complete statement Theorem 41 (Necessity of α). The fine-structure constant α is the unique departure from the golden ratio φ that is: 1. forced by the irremovable defect at p on the Riemann sphere (Section 2); 2. consistent with the uniform regularity of the self-similar field at all scales (Theorem 14); 3. compatible with the existence of matter (Proposition 10); and 4. realised in a vacuum whose constitutive constants ε₀ and μ₀ are the mathematical primitives 1 and 0 by their own definitions (Theorem 34). Its value is $$\alpha = \frac{Z_0}{2R_K} = \frac{\sqrt{\mu_0/\varepsilon_0}}{2h/e^2} = \frac{e^2}{2\varepsilon_0 hc} \approx \frac{1}{137.036}, \label{eq:alpha_final}$$ derived from the mathematical structure with zero free parameters, zero postulates, and one empirical observation (matter exists). Proof. Conditions (i)–(iii) establish that ξ_(actual) = φ − δ with δ > 0 uniquely determined. Condition (iv), established in Theorem 34, is the recognition that the physical vacuum IS the mathematical structure: not an approximation to it, not a model of it, but the same structure expressed in physical vocabulary. By Proposition 33, both primitives are necessarily present in any vacuum that sustains a non-terminal process; by Theorem 34, they are ε₀ and μ₀. Given (iv), the departure δ = Z₀/2R_(K) follows by Theorem 37. The identification δ = α follows from the definition of α as the dimensionless coupling constant of the electromagnetic interaction, which is precisely the departure of the physical vacuum from ideal self-similar scaling. The formula [eq:alpha_final] is an algebraic identity in the 2019 SI system (Bureau International des Poids et Mesures 2019). No measurement is used to derive it; measurement confirms what the structure requires. ◻ Summary of Section 5. All three gaps are closed without introducing additional structure. Gap 1: the recursion x = 1 + 1/x is the unique minimum-information self-referential consistency condition, forced by the Markov spectrum theorem. Gap 2: a non-terminal vacuum must contain both a primitive 1 (containing mode) and a primitive 0 (penetrating mode), and the physical definitions of ε₀ and μ₀ are these primitives; the departure δ = Z₀/2R_(K) then follows by assembly of four previously established facts with no quotient. Gap 3: the Josephson phase equation follows from the Aharonov–Bohm phase with minimum winding charge 2e, topologically forced by the same minimum winding number as α = Z₀/2R_(K). The factor of 2 in both α and K_(J) is the same topological factor throughout. The proof is complete. Its honest characterisation is: The fine-structure constant α is the departure from φ uniquely forced by the irremovable defect of the non-terminal reciprocal singularity 1/0, realised in an electromagnetic vacuum whose constitutive constants are the mathematical primitives by their own definitions. It is not derived from physics; the physics is the expression of the mathematics. Discussion The preceding proof derives the fine-structure constant from the non-terminal reciprocal singularity 1/0 on the Riemann sphere. Hydrogen’s spectrum was not used. This section identifies where hydrogen — and by extension all of atomic and nuclear spectroscopy — lives inside the derived structure. The empirical record does not support the proof; it instantiates it. The distinction matters. Hydrogen as the minimal instantiation The proof requires the minimum closed electromagnetic geometry consistent with Maxwell’s two independent curl equations. Section 4 established that each curl equation requires a minimum of three non-degenerate nodes for closure, and that the two cycles must share exactly one node (the impedance-match point at IE/4), giving five nodes in total. Hydrogen, as the first element formed cosmologically and the simplest atom, instantiates this minimum geometry exactly. Its lowest five quantum states are: {1s, 2s, 2p⁻¹, 2p₀, 2p₊₁}, with n = 1 providing the ground state and n = 2 providing the first excited shell. The states {2s, 2p} together are the first configuration in which both Maxwell curl equations are simultaneously active: - 2s (ℓ = 0): purely radial, capacitive E-sector — the ε₀ mode only. - 2p (ℓ = 1): the first mode supporting angular momentum, activating the inductive H-sector — the μ₀ mode. The right angle between the energy axis (ℓ = 0) and the angular-momentum axis (ℓ = 1) at the 2s vertex is the geometric record of Maxwell’s orthogonality condition E ⊥ H at the impedance-match point. The proto-triangle and the integer scar Mapping the three states {1s, 2s, 2p} into binding-energy and angular-momentum coordinates (E_(bind),ℓ), measured in units of (IE/4, Δℓ=1), places the vertices at (4,0), (1,0), and (1,1). This forms a right triangle with: - horizontal leg  = 3 (energy: 4 − 1 = 3 units of IE/4); - vertical leg  = 1 (angular momentum: ℓ(2p) − ℓ(2s) = 1 − 0 = 1); - right angle at 2s, forced by E ⊥ H at the impedance-match point. The horizontal leg 3 is an integer — the universe’s first electromagnetic closure occurred at integer ratio 3 : 1 rather than at the ideal irrational ratio φ² : 1. The departure of the integer closure from the φ-attractor is: 3 − φ² = φ⁻², an exact algebraic identity from φ² = φ + 1. This quantity φ⁻² ≈ 0.382 is the proto-scar: the fossil of the universe’s first integer closure, visible in the hydrogen energy levels and related to the same φ-departure that generates α. The proto-triangle {1s, 2s, 2p} is not a geometric input to the proof. It is a geometric consequence, readable directly from the NIST hydrogen spectrum (Kramida et al. 2024). The derivation precedes the triangle; the triangle instantiates it in the simplest possible atom. The proto-triangle {1s, 2s, 2p} in binding-angular space, legs 3 and 1 in units of (IE/4, Δℓ=1), right angle at 2s. The right angle records E ⊥ H at the impedance-match point (§4.3). The deviation 3 − φ² = φ⁻² (equation [eq:proto_scar], exact) is the proto-scar. The hypotenuse touches the IE/4 boundary — the impedance-match point no bound-to-bound transition can reach (§4.4). The impedance-match boundary in atomic spectra The proof establishes in Section 4.4 that no bound-to-bound transition can carry energy IE/4, a consequence of number theory independent of atomic species or nuclear charge. The empirical survey of Heaton (K. B. Heaton and Coherence Research Collaboration 2026c) confirms the spectroscopic consequence: 71 of 80 ions surveyed show a zero-observation interval at IE/4 with zero falsifications; the remaining 9 have catalogs that do not bracket IE/4 from both sides and cannot be tested without additional spectral coverage. The gap is confirmed across Z = 1–82, all four subshell families, and multiple ionisation stages; in 42 of the 71 confirmed ions it ranks in the top 10% by width among all consecutive-pair intervals. The empirical record does not establish the theorem; it confirms that the theorem’s prediction is not falsified by any known spectrum. The thread law as direct observation of the logarithmic scaling The self-similar logarithmic scaling of Section 3 — equipotential surfaces nesting with ratio α at each depth — is directly visible in photon frequencies via the empirical thread law (K. B. Heaton and Coherence Research Collaboration 2026a, 2026c): log₁₀ν(κ) = χ + β κ,   β = log₁₀α. For hydrogen, the fit to NIST spectral data gives β_(fit) =  − 2.1368346704388017 against log₁₀α =  − 2.1368346704397490, a residual of 9.47 × 10⁻¹³ with r² = 1.000 over 9 pairs. The thread law is not evidence for the proof; it is the proof made visible in the spectrum of the simplest atom. Recovery of fundamental constants from the gap The impedance-match boundary encodes the electron rest mass and the Bohr radius. From Heaton (K. B. Heaton and Coherence Research Collaboration 2026c): $$m_e c^2 \;=\; \frac{8G}{\alpha^2}, \qquad a_0 \;=\; \frac{\hbar c\,\alpha}{8G}, \label{eq:mass_bohr}$$ where $G = \sqrt{\mathrm{gap}_{\mathrm{lo}} \cdot \mathrm{gap}_{\mathrm{hi}}}$ is the geometric mean of the gap boundaries, derived from the NIST photon catalog with no free parameters and without using m_(e) or a₀ as inputs. The proof derives α from the geometric departure; the gap encodes m_(e) and a₀ through α; the constants define the atom. Nuclear spectroscopy: the same structure at a different scale The same mathematical structure governs nuclear shell closure. The four doubly magic nuclei ¹⁶O, ⁴⁸Ca, ¹³²Sn, and ²⁰⁸Pb show complete depletion of observed levels below S_(n)/4; all 20 non-doubly-magic nuclides surveyed show no gap; zero nuclides falsify the prediction; the two populations do not overlap (K. B. Heaton and Coherence Research Collaboration 2026c). The fraction 1/4 appears at both atomic and nuclear scales because it follows from the Diophantine impossibility of equation [eq:diophantine], which is purely mathematical and indifferent to physical scale. The hierarchy of confirmation The relationship between the proof and its empirical record is strictly hierarchical: 1. The mathematical structure (Sections 2–5) is primary and self-contained. 2. The electromagnetic vacuum instantiates this structure (Theorem 34). 3. Atomic spectra are the observable record of the structure’s consequences in the simplest bound systems. 4. The NIST data confirms that no known spectrum falsifies any prediction. The empirical agreement across 71 atomic ions, 24 nuclear nuclides, and 13 orders of magnitude in energy scale does not prove the mathematical argument. What the empirical record establishes is that the universe we inhabit is the one in which ε₀ and μ₀ are the mathematical primitives 1 and 0 — not by coincidence, but because those words mean what they mean. Conclusion We have derived the fine-structure constant from the non-terminal reciprocal singularity 1/0 on the Riemann sphere. The argument proceeds in four sections, each building on the last without importing new structure. Section 2 establishes that the non-terminal reciprocal singularity forces a punctured sphere S² \ {p} with an irremovable source singularity at p, a unique antipodal sink p^(*) forced by the Poincaré–Hopf theorem (χ(S²) = 2), and a logarithmic potential field between them forced by the Green’s function of the Laplacian in two dimensions. The defect at p carries residue 1 and winding number  − 1 and cannot be removed by any continuous deformation. Section 3 shows that the logarithmic field has a self-similar scaling structure whose consistency recursion, derived from the Markov spectrum theorem, forces the unique continued fraction [1;1,1,1,…] = φ as the winding ratio on the ideal sphere. Rational winding is excluded because it produces a resolvent pole incompatible with stable matter. The uniform regularity condition (Definition 11), together with Hurwitz’s theorem (1891), establishes that φ is the unique irrational satisfying the resolvent bound at all scales simultaneously. The irremovable defect at p forces a departure δ > 0 from this ideal ratio. Section 4 identifies the electromagnetic realisation of the structure. The two vacuum constants ε₀ and μ₀ are the mathematical primitives 1 (containment) and 0 (openness) by their own physical definitions; no additional structure is introduced. Their ratio $Z_0 = \sqrt{\mu_0/\varepsilon_0}$ is the 0/1 geometric ratio of the vacuum. The orbital impedance Z_(orb)(n) = Z₀n/α is derived from Coulomb’s law and the Biot–Savart law with no free parameters. The impedance-match boundary Z_(orb) = Z₀ corresponds to ΔE = IE/4 by a number-theoretic impossibility (no integer solution to 1/n_(f)² − 1/n_(i)² = 1/4), confirmed across 71 atomic ions with zero falsifications. The von Klitzing constant R_(K) = Z₀/2α and the Josephson constant K_(J) = 2e/h are dual quantum expressions of the same 0/1 ratio; together they resolve both h and e, neither of which is a primitive of the electromagnetic structure. Planck’s constant h is the invariant of the self-similar cleaving process. Section 5 closes the three remaining gaps. The recursion x = 1 + 1/x is forced by the Markov spectrum. The identification of ε₀ and μ₀ with the primitives is definitional, not axiomatic, so no quotient is introduced. The departure δ = Z₀/2R_(K) follows by assembling four previously established facts: the residue  = 1, Z₀ as the 0/1 ratio, R_(K) = h/e² from the Dirac condition and the cleaving invariant, and the minimum winding number 2. The Josephson phase equation follows from the Aharonov–Bohm phase with minimum winding charge 2e. Within the framework where ε₀ and μ₀ are the fixed primitives, the constancy of c follows as a mathematical consequence rather than a physical postulate (Corollary 35). The result is: $$\boxed{ \alpha \;=\; \frac{Z_0}{2R_K} \;=\; \frac{\sqrt{\mu_0/\varepsilon_0}}{2h/e^2} \;=\; \frac{e^2}{2\varepsilon_0 hc} \;\approx\; \frac{1}{137.036}, } \label{eq:alpha_conclusion}$$ derived with: ------------------------ ------------------------------- Free parameters zero Postulates zero Mathematical premises one (uniform regularity, P10) Empirical observations one (matter exists) ------------------------ ------------------------------- The proof is not a derivation of physics from mathematics. It is the recognition that the physical vocabulary of electromagnetism — permittivity, permeability, impedance, resistance, charge, action — names the mathematical structure that the non-terminal reciprocal singularity 1/0 necessarily generates. The physics does not explain the mathematics. The mathematics is what the physics means. Provenance The research that gave rise to this paper was conducted by Kelly B. Heaton through an interactive collaboration with Claude (Sonnet 4.6) (large language model developed by Anthropic) and ChatGPT (various models, OpenAI), collectively referred to as the Coherence Research Collaboration. All research design, direction, interpretation, and conclusions were determined by the human author who is the sole owner of the intellectual property. DOI: 10.5281/zenodo.19157339 Version History v1.0: March 25, 2026. Initial publication. v1.1: March 29, 2026. Corrected citation metadata for reference [30]: updated title to The Encounter of Two Primitives and One Flaw: Geometric Unification at the Electromagnetic Scale, updated status from forthcoming to published. Removed an unused draft bib entry carrying an incorrect subtitle for this paper. No changes to scientific content, proofs, or conclusions. [] Lucerna Veritas Follow the light of the lantern. Aharonov, Y., and D. Bohm. 1959. “Significance of Electromagnetic Potentials in the Quantum Theory.” Physical Review 115 (3): 485–91. https://doi.org/10.1103/PhysRev.115.485. Ahlfors, L. V. 1979. Complex Analysis. 3rd ed. McGraw-Hill. Bethe, H. A., and E. E. Salpeter. 1957. Quantum Mechanics of One- and Two-Electron Atoms. Berlin: Springer. https://doi.org/10.1007/978-3-662-12869-5. Bhāskara II. 1150. Lı̄lāvatı̄. Brahmagupta. 628AD. Brahmasphutaṣiddhānta. Bureau International des Poids et Mesures. 2019. The International System of Units (SI). 9th ed. Sèvres, France: Bureau International des Poids et Mesures. https://www.bipm.org/en/publications/si-brochure. Conway, J. B. 1978. Functions of One Complex Variable. 2nd ed. Springer. https://doi.org/10.1007/978-1-4612-6313-5. Dirac, P. A. M. 1931. “Quantised Singularities in the Electromagnetic Field.” Proceedings of the Royal Society of London, Series A 133 (821): 60–72. https://doi.org/10.1098/rspa.1931.0130. Eddington, A. S. 1946. Fundamental Theory. Cambridge University Press. Evans, L. C. 1998. Partial Differential Equations. Vol. 19. Graduate Studies in Mathematics. American Mathematical Society. Feynman, R. P. 1985. QED: The Strange Theory of Light and Matter. Princeton, NJ: Princeton University Press. Guillemin, V., and A. Pollack. 1974. Differential Topology. Prentice-Hall. Hatcher, A. 2002. Algebraic Topology. Cambridge University Press. https://pi.math.cornell.edu/~hatcher/AT/ATpage.html. Heaton, K. B., and Coherence Research Collaboration. 2026a. “Hydrogenic Alignment as a Coordinate Principle for Atomic Spectra.” Zenodo, January. https://doi.org/10.5281/zenodo.18167643. ———. 2026b. “Determinacy Under Quotient Representations.” Zenodo, March. https://doi.org/10.5281/zenodo.18868210. ———. 2026c. “Electron Mass and Bohr Radius from a Universal Spectroscopic Gap: Vacuum Impedance as the Organizing Principle of Atomic Spectra.” Zenodo, March. https://doi.org/10.5281/zenodo.19164224. Heaton, Kelly B., and Claude Sonnet 4.6. 2026. “The Encounter of Two Primitives and One Flaw: Geometric Unification at the Electromagnetic Scale.” https://doi.org/10.5281/zenodo.19194036. Heaviside, Oliver. 1893. Electromagnetic Theory. Vol. 1. London: The Electrician Printing; Publishing Company. Hopf, H. 1926. “Vektorfelder in n-Dimensionalen Mannigfaltigkeiten.” Mathematische Annalen 96: 225–50. https://doi.org/10.1007/BF01209164. Hurwitz, A. 1891. “Ueber Die Angenäherte Darstellung Der Irrationalzahlen Durch Rationale Brüche.” Mathematische Annalen 39: 279–84. https://doi.org/10.1007/BF01206656. Jackson, J. D. 1999. Classical Electrodynamics. 3rd ed. Wiley. Jentschura, U. D., and I. Nandori. 2014. “Attempts at a Determination of the Fine-Structure Constant from First Principles: A Brief Historical Overview.” European Physical Journal H 39: 591–613. https://doi.org/10.1140/epjh/e2014-50044-7. Josephson, B. D. 1962. “Possible New Effects in Superconductive Tunnelling.” Physics Letters 1 (7): 251–53. https://doi.org/10.1016/0031-9163(62)91369-0. Kato, T. 1966. Perturbation Theory for Linear Operators. Vol. 132. Grundlehren Der Mathematischen Wissenschaften. Springer. https://doi.org/10.1007/978-3-642-66282-9. Kitano, M. 2009. “The Vacuum Impedance and Unit Systems.” IEICE Transactions on Electronics E92-C (1): 3–8. https://doi.org/10.1587/transele.E92.C.3. Kramida, A., Yu. Ralchenko, J. Reader, and NIST ASD Team. 2024. “NIST Atomic Spectra Database (Ver. 5.12).” Gaithersburg, MD: National Institute of Standards; Technology. https://physics.nist.gov/asd. Markov, A. A. 1880. “Sur Les Formes Binaires Indéfinies.” Mathematische Annalen 17: 379–99. https://doi.org/10.1007/BF01446234. Maxwell, J. C. 1865. “A Dynamical Theory of the Electromagnetic Field.” Philosophical Transactions of the Royal Society of London 155: 459–512. https://doi.org/10.1098/rstl.1865.0008. Milnor, J. 1965. Topology from the Differentiable Viewpoint. University Press of Virginia. Picard, É. 1879. “Sur Une Propriété Des Fonctions Entières.” Comptes Rendus de l’Académie Des Sciences 88: 1024–27. Poincaré, H. 1885. “Sur Les Courbes définies Par Une équation Différentielle.” Journal de Mathématiques Pures Et Appliquées 1: 167–244. Riemann, B. 1851. “Grundlagen für Eine Allgemeine Theorie Der Functionen Einer Veränderlichen Complexen Grösse.” PhD thesis, Georg-August-Universität Göttingen. Robertson, B. 1971. “Wyler’s Expression for the Fine-Structure Constant.” Physical Review Letters 27: 1545–47. https://doi.org/10.1103/PhysRevLett.27.1545. Rudin, W. 1976. Principles of Mathematical Analysis. 3rd ed. McGraw-Hill. Singh, T. P. 2022. “Quantum Gravity Effects in the Infra-Red: A Theoretical Derivation of the Low Energy Fine Structure Constant and Mass Ratios of Elementary Particles.” European Physical Journal Plus 137: 664. https://doi.org/10.1140/epjp/s13360-022-02868-4. Sommerfeld, A. 1916. “Zur Quantentheorie Der Spektrallinien.” Annalen Der Physik 356 (17): 1–94. https://doi.org/10.1002/andp.19163561702. Tinkham, M. 1996. Introduction to Superconductivity. 2nd ed. New York: McGraw-Hill. von Klitzing, K., G. Dorda, and M. Pepper. 1980. “New Method for High-Accuracy Determination of the Fine-Structure Constant Based on Quantized Hall Resistance.” Physical Review Letters 45 (6): 494–97. https://doi.org/10.1103/PhysRevLett.45.494. Wyler, A. 1969. “L’espace Symétrique Du Groupe Des équations de Maxwell.” Comptes Rendus de l’Académie Des Sciences, Série A-B 269: A743–45.