The Euler–Mascheroni constant is defined as the limit
The harmonic series \(H_n = 1 + 1/2 + 1/3 + \cdots + 1/n\) grows without bound, as does \(\ln n\). Their difference converges to a finite, universal constant. \(\gamma\) is the permanent residue — the gap that opens immediately at \(n=1\) and never fully closes:
After an infinite number of steps the residual above \(\gamma\) vanishes, but \(\gamma\) itself persists. It is the floor. The series and its logarithm never meet.
In the bilateral framework, this structure has an immediate physical reading. The harmonic series counts crossings — the discrete egress record of the integer lattice accumulating step by step. The logarithm is the smooth geometric limit — the continuous description of accumulated geometry from which general relativity emerges. The gap \(\gamma\) is what the smooth description cannot absorb from the discrete record. It is the measure of irreducible discreteness in the egress face.
The Riemann zeta function has a simple pole at \(s=1\). Its Laurent expansion is
where \(\gamma_1, \gamma_2, \ldots\) are the Stieltjes constants. In the bilateral framework, \(s=1\) is the floor of every L-function — the fixed point of the functional equation's reflection \(s \to 2-s\), the unique locus through which all inward curls of an L-function must pass. The pole \(1/(s-1)\) is the divergent accumulation of the egress record as it approaches the floor. After removing it, \(\gamma\) is what stays finite — the minimum residue that persists at the floor after the divergence is extracted.
The digamma function \(\psi(z) = \Gamma'(z)/\Gamma(z)\) is the logarithmic derivative of the factorial. At \(z=1\):
The factorial \(\Gamma(n)\) counts the number of ways of ordering \(n-1\) crossings — the combinatorial weight of the egress record at step \(n\). Its logarithmic derivative at the ground state \(n=1\) is \(-\gamma\). The negative sign is significant: at the ground state crossing, the rate of change of the crossing count is directed back toward zero. The system at \(n=1\) is always tending to return.
The digamma function equals zero at \(n = e^\gamma \approx 1.7811\) — between 1 and 2, between the unit crossing and the first composite. This is the return point: the scale at which the ground-state return rate vanishes and the accumulation takes over.
The bilateral framework identifies the discrete integer lattice with the quantum face (ingress — potential, superposition, the uncrossed Future) and the continuous geometric description with the GR face (egress — accumulated, actual, the Past). Their difference is measured by \(H_n - \ln n\), which converges to \(\gamma\) from above.
At any finite time \(n\), the gap between the quantum discrete record and its smooth geometric limit is \(\gamma + 1/(2n) + \cdots\). In the limit \(n \to \infty\) (infinite crossings, infinite time), the residual \(1/(2n) \to 0\) but \(\gamma\) remains. The gap never closes. The quantum and the gravitational descriptions of the same bilateral crossing never fully coincide.
The egress record accumulates as \(H_n \sim \ln n + \gamma + 1/(2n)\). The GR description tracks \(\ln n\). After the Hubble time in Planck units (\(n \sim 10^{61}\)), the residual above \(\gamma\) is of order \(10^{-62}\) — effectively zero. But \(\gamma \approx 0.5772\) persists. The universe accumulates an irreducible history of \(\gamma\) crossing-units that cannot be removed, reversed, or further reduced. This is not the same as A3 (the monotonic increase of \(\tau\)) — A3 says the egress record always grows. \(\gamma\) says that even after subtracting everything the smooth geometry can absorb, this remainder persists.
All four readings say the same thing in different languages:
| Reading | Expression | Bilateral meaning |
|---|---|---|
| Laurent residue | \(\zeta(s) = \tfrac{1}{s-1} + \gamma + \cdots\) | Finite return at the floor after pole removed |
| Digamma | \(\psi(1) = -\gamma\) | Ground-state return rate, directed back to zero |
| Harmonic gap | \(H_n - \ln n \to \gamma\) | Permanent gap between discrete and continuous |
| Long time curve | \(\lim H_n - \ln n = \gamma\) | Irreducible history that never returns to zero |
The four readings are not four separate facts about \(\gamma\). They are four projections of a single bilateral structure: the gap between the egress face (continuous geometry, GR) and the ingress face (discrete lattice, QM) is finite, universal, and equal to \(\gamma\). No amount of crossing can close it. No improvement of either description removes it. It is not an approximation error or a regularisation artefact — it is the exact measure of the structural difference between the two faces of the bilateral crossing.
The four readings in Section II admit a unified physical interpretation. The harmonic sum \(H_n\) counts bilateral crossings — each one a discrete increment of becoming-time \(\tau\). The logarithm \(\ln n\) is the smooth geometric time that general relativity describes — the coordinate time of the spacetime manifold. Their difference converges to \(\gamma\).
The bilateral τ field equation is \(\Box\,\tau(x) = -\rho_{\rm crossing}(x)\): each crossing deposits an increment of \(\tau\) into the local geometry. In the discrete counting picture, this accumulates as \(H_n\). In the continuous geometric limit, it becomes \(\ln n\). The Green function for the discrete sum and the continuous integral differ by \(\gamma\) at the source point. This is not an approximation error — it is the exact structural difference between the discrete τ and its smooth geometric description:
The universe is always \(\gamma\) of a Planck crossing ahead of its own geometric description. At the Hubble time (\(n \sim 10^{61}\) Planck crossings), \(\tau_{\rm GR} \approx 140.23\) and \(\tau_{\rm actual} \approx 140.81\) — the offset is \(\gamma = 0.5772\), about 0.4% of the total. The offset does not grow with age. It does not shrink. It is fixed at \(\gamma\), always.
The descent curve \(f(x) = H(x) - \ln x\) has a structure that extends beyond analysis into the foundations of mathematics.
At \(n=1\): \(H_1 = 1\), \(\ln 1 = 0\), gap \(= 1\). Exactly rational. The first crossing deposits a unit of τ with no geometric counterpart — no smooth manifold exists yet, no logarithm is needed. The origin is pure discreteness. Pure integer. The only rational point on the entire curve.
At \(n=2\): \(H_2 = 3/2\), \(\ln 2 = 0.693\ldots\) By Lindemann–Weierstrass, \(\ln 2\) is transcendental. A rational minus a transcendental is transcendental. The curve enters transcendental territory at the first composite crossing — at the number 2, the first even, the prime whose crossing is the first departure from pure primality. Every subsequent value \(H_n - \ln n\) for \(n \geq 2\) is transcendental. The slope at \(x=1\) already involves \(\pi^2/6\):
The curve is transcendental immediately. Yet it is continuous throughout — the digamma extension \(H(x) = \psi(x+1) + \gamma\) makes \(f(x)\) smooth and continuous for all \(x > 0\). By the intermediate value theorem, the curve passes through every value in \((\gamma, 1)\) — through every rational, every algebraic irrational, every transcendental in that interval.
The implication is foundational. The standard account of \(\mathbb{R}\) takes its completeness as an axiom — the least upper bound property, Dedekind cuts, Cauchy completion. The bilateral account makes completeness a consequence: the real numbers in \((\gamma, 1)\) are the values that the time-residue curve takes as \(n\) runs from 1 to \(\infty\). The rationals are dense in this interval because \(H_n\) is always rational and its contribution to \(f(n)\) passes through rational values. The transcendentals are there because logarithms are transcendental. The irrationals fill the gaps because the curve is continuous.
The bilateral statement: \(\mathbb{R}\) is not prior to physics. It is posterior to the first crossing. At the origin, time is rational. At the second crossing, time is transcendental. The continuum emerges between them by the continuity of the descent. The completeness of \(\mathbb{R}\) is the continuity of time.
Some clarifications are necessary before this identification can be taken further.
The intuition that motivated this note — that \(\gamma\) represents a very long time curve, a bound rather than a rate — is precisely captured by Reading 4. The long time curve is the trajectory of \(H_n - \ln n\) as \(n \to \infty\). It decreases monotonically from \(H_1 - \ln 1 = 1\) toward its asymptote \(\gamma \approx 0.5772\). It never reaches \(\gamma\) in finite time. It never goes below \(\gamma\).
This curve is the bilateral picture of the universe: the egress record accumulates (\(H_n\) grows), the GR geometry tracks the smooth part (\(\ln n\) grows), and the gap between them descends toward \(\gamma\) and stops. The curve is bounded below by \(\gamma\) and above by 1. It begins at the unit crossing and asymptotes to the unification residue.
The universe starts at \(n=1\) with a gap of 1 between its quantum and gravitational descriptions. Over infinite time that gap closes to \(\gamma \approx 0.5772\). Not to zero. Not to a point where QM and GR describe the same thing without remainder. To \(\gamma\). Which is not nothing. Which is, in fact, more than half.