The Second String: TP-3's Pre-Registration, Paid on a Second System

In Closed Accounts (TP-3, October 29, 2026) I priced string breaking in the units the string sets for itself. With σ ≈ 0.18 GeV² and a breaking radius of 1.2–1.4 fm, the dimensionless product came out at √σ·r_break ≈ 2.80, inside a band of 2.580–3.010 built from the radius range alone. One system gave that number. A threshold is not one system's number, so I filed the number as a pre-registration instead of a result: a second string-breaking system, measured in the same units, either lands in the band or the sever rhyme abroad dies. This paper pays that filing. The second system is stated before any arithmetic: lattice QCD with N_f = 2+1 dynamical flavours, the N200 ensemble of Bulava and collaborators, where the string tension and the breaking radius come out of one fit to one ensemble. The sea differs from the N_f = 2 determinations behind TP-3's range. The pion is heavier than nature's, at 280 MeV. The breaking radius is defined as a level crossing, stated in the open. In lattice units the spacing cancels, and the product is √σ·r_c = 2.883(23). It lands in the filed band. The analog hardens one grade, to CONSTRAINS, and no further. The same ensemble's strange channel sits at 3.044(24), just past the band's top edge, and I print that beside the result at the same size: the product moves when the created pair gets heavier, so what repeats is a light-channel number, not a constant of nature. The home headline is untouched by any of this, in either direction. Grades as printed in the paper: √σ·r_c = 2.883(23) inside the band [2.580, 3.010]; verdict CONSTRAINS.