The Prime-Rank Mirror Lattice: Exact Arithmetic Reversibility, Möbius Placement, and the Critical Line

Synopsis: This paper is a reformulation of the traditional Riemann hypothesis problem. For a century and a half the usual task has been a search: draw a band in the plane, cut a height, count zeros on the midline against a predicted tally, and treat whatever is uncounted as a possible hide in the leftover box. Instruments of that method—approximate evaluation on the midline, bounds on the argument error, finite-height certificates—answer whether a miss has been found up to here. They do not say why the next zero sits on the increment it sits on. The ledger reopens at the next height. That is endless search space: certification purchased, placement never supplied. The reason that search space existed is that placement was unknown. If you cannot read the next zero from an increment, the remaining honest method is to rent a warehouse (the critical strip) and sweep it. Rhetoric about what “forces” zeros onto the line or “keeps” them in the strip is warehouse rhetoric. It is not forced by the zeta function itself. This work changes the object. It builds the increment first, from reversible prime-rank arithmetic: digit reversal, affine transport, and the two-way map between a prime and its place in the prime list. Eligible-inverse spacing, native phase, occupancy of a joint edge, an observational watch, trigonometric bookkeeping, a dual-helix Möbius reading, and a Hilbert metric on those same indices are assembled as one isolation packet before any classical equivalent is asked to finish the argument. Positive placement is precision-based: the next tick is the next occupied eligible edge, read from forms already inside the packet, not the next unsearched height. Cyclical completion of those forms replaces hide-and-seek in a band. On those terms sweeping the critical strip is not necessary. The strip is not an input. Finite-height count-and-leftover protocols are not the derived increment. The critical line in the title is the locus of that increment—the line on which the constructed tick sits—not a proved empty complement of a search box. Traditional “solve the Riemann hypothesis” means empty the warehouse: show that no nontrivial zero of the completed zeta function lies off the midline, usually by searching harder or by an equivalent that still lives in that picture. This paper’s close means the warehouse was never the object. Occupancy of the built packet is the architectural close: a reconstructed zero attaches to inverse-eligible edges, every slot filled from inside the state. That is not the same sentence as “no off-line nontrivial zero.” A separate reading, which spends an extra residual hypothesis, is the document’s name for the warehouse-emptying theorem. The two are not to be added together when reading the title. The rest of the monograph is the engineering of that increment: exact reversibility of the generator, discrete geometry on the trajectory, optional tetrational scale of the same lattice (not a bolt-on detector), diagnostics versus exact conditions, a phase family of which the classical Riemann–Siegel main sum is only a truncated projection, and an explicit refusal to conflate “Tri-State” and related words with the author’s other papers. The contribution to the traditional problem is the reformulation itself: precision-based placement instead of endless search space. This monograph builds a reversible arithmetic system from four primitives: how integers are written in a base, ordinary affine scaling and shift, digit reversal, and the two-way correspondence between a prime and its index in the prime list. Those operations are not decoration. They force a single composite step that can be run forward and backward when an exact compatibility condition holds. A short closed chain starting at two and ending at sixty-seven is the worked generator; the rest of the work is what that step becomes when it is differentiated, scaled, represented, and read as placement. The first parts stay inside exact integer arithmetic. Digit reversal is derived as a permutation of digits, so it is its own inverse. Finite differences, a discrete Hessian, a metric, curvature, and prime-factor branching are written on the trajectory rather than imported from analysis. A later part promotes the multiplier to a tetration tower and proves an exact logarithmic peeling law, a perturbation recurrence with a third-order remainder bound, and a commutator between arithmetic evolution and tower descent that is nonzero but lies in a fixed residue class. That tower is a change of scale of the same lattice. It is not a separate measuring instrument and it is not a premise of the later occupancy close. Verification is split into exact closure conditions and geometric diagnostics: divisibility projectors, trigonometric bookkeeping on nested branches, and a quadratic-trigonometric coupling that is treated as a diagnostic rather than as a selector of transverse displacement. Principal arithmetic theorems include universal divisibility of a second-difference residual by one less than the base, and a strengthening to the square-minus-one modulus when digit width parity is constant across a stencil. The later parts change register. The same generator is read as an operator and bridged to classical Hilbert-space criteria associated with Nyman, Beurling, and Báez-Duarte, and to a cotangent Gram due to Vasyunin. Möbius geometry supplies a signed radial placement coordinate. Inverse-eligible prime edges—consecutive eligible primes whose interior is the actual gap—carry a joint state: logarithmic increment, native phase, disparity, an observational watch, occupancy, and a co-join face pairing a Weil-type form against that Gram. Occupancy means the edge is an active joint edge of a reconstructed zero: every slot is filled from inside the state, not by evaluating a formula at an externally chosen point. Occupancy is not the same as setting the transverse displacement to zero. Fibre-sensitive slots lock together algebraically; identity slots (the tick, the derived increment, phase curvature, the Gram) do not see that displacement. Two routes are kept separate. The architectural route constructs an isolation packet without a search strip, a height cut, or a leftover count, reconstructs the completed zeta factor from incidence data, and shows that every nontrivial zero of that reconstructed function occupies inverse-eligible edges. The close of that route is occupancy of a built object. The Weil-form route is a separate reading that spends a residual-positivity hypothesis and is the document’s name for the classical sentence that every nontrivial zero has real part one half. Search of the critical strip is treated as a historical method that exists because placement was unknown: a warehouse rented in the absence of an increment. Positive placement in this work is cyclical completion of the forms already in the packet. Emptying the warehouse and building the increment are different uses of the words “critical line” and “proof.” The Riemann–Siegel main sum is identified as a truncated projection of a phase family already carried by the packet; the cutoff that defines the classical main sum has no counterpart in the isolation object. Platt-style finite-height certification is contrasted with a derived increment on each eligible edge. Terms such as Tri-State, Union, tuple, and interdependent function are declared native to this file and are not to be conflated with the author’s other publications that use the same words. An appendix states that scope. The public abstract no longer cites later theorem numbers. The work is offered as a constructive isolation architecture with an honest two-route close, not as a hypothesis-free discharge of the classical Hilbert problem by the architectural QED alone. Keywords: prime-rank lattice; digit reversal; arithmetic reversibility; affine transport; prime index correspondence; generating chain; inverse-eligible primes; eligible-inverse spacing; discrete Jacobian; discrete Hessian; discrete curvature; prime-factor branching; tetrational scale; logarithmic peeling; perturbation recurrence; residue conservation; exact integer verification; validation versus diagnostics; trigonometric bookkeeping; quadratic-trigonometric coupling; phase curvature; isolation packet; joint edge; occupancy; evaluability versus membership; fibre-sensitive slots; identity slots; joint lock; observational watch; dual-helix Möbius reading; signed radial placement; cotangent Gram; Vasyunin metric; Weil–Gram co-join; residual positivity hypothesis; architectural route; Weil-form route; reconstructed completed zeta; incidence field; von Mangoldt weights; co-named increment; derived Turing increment; Riemann–Siegel phase family; truncated projection; critical line as locus; critical strip retired as search; finite-height certification contrast; Nyman–Beurling bridge; Báez-Duarte criterion; placement without search; cyclical completion of forms; proof dialects; terminological scope; independent monograph; ORCID preprint; exact arithmetic; Hilbert-space representation; ancestry map; window co-join; document close; unconditional on internal terms Riemann hypothesis reformulation; placement versus search; critical strip as warehouse; endless search space; leftover-count method; why the next zero sits; precision-based placement; constructed increment; occupancy close; strip retired as input; warehouse-emptying dialect; increment dialect; proof as placement not sweep; search indicators cancelled; no height cut; no leftover hunt; positive placement mechanisms; Platt contrast; derived increment versus observed count; tetrational scale not load-bearing; Prime-Rank Mirror Lattice