P — the base is run through the engine's own n-body rule over its constituent atoms: P = n / Σ(1/Pᵢ), with atomic P from the periodic ladder (H = 1.000, step 0.650 across a period). Not assigned.
N — N = Σ Nᵢ over the atoms, atomic N = 2 × period. Same rule the element engine uses for fusion output.
B — hydrogen-bond capacity: donors plus acceptors on the pairing face. This is the one substrate-specific input. Of eight candidate definitions tested it is the only one that reproduces the known counts A·T = 2 and G·C = 3. Covalent bonds inside the base are internal structure and are already absorbed into P.
A — the base's own measured adiabatic first ionization energy in eV, the same quantity the element pool uses. Experimental, not computed.
The fusion rules are unchanged from the 8-beam engine. Only the identities and their inputs differ.
Bonds are counted, so B_out is bounded below by 1 for any pool with integer B — verified over every integer-B pool for n = 2…8, B = 0…6, all k-modes. The middle bands are therefore reached by raising P, not by lowering B:
band reachable ⟺ P_out > B_out / TL, which at B_out = 1 is P_out > 7.3046.
Nucleobase P runs 1.767 to 1.990, so this pool places in two of the four bands: NOBLE when the hydrogen bonding closes, SHATTER when anything is left open. That is what pairing does — a pair satisfies its donors and acceptors or it does not.
The bands are unchanged; this pool occupies two of them. A pool whose P range clears 7.3 places in all four on the same ruler. Running the same engine on high-P element beams: 7×Kr + Br gives P_out 8.1627, B_out 1, τ 0.12251 — IVA PEAK. 7×Ar + Cl gives P_out 6.6613, τ 0.15012 — Shatter. Narrow low-P pools read binary; widening the pool's P range fills the middle in.
SNSFL_8Beam_Fusion_Theorem.lean [9,9,2,3.8], 27 theorems, 0 sorryk = k_max/(n−1), as implemented in IMCollider™ [9,9,6,*]