The Contraction, Priced: FitzGerald's and Lorentz's Reverse-Engineered Factor and the Fringe Shift It Was Built to Cancel

Census pricing (row CC-103) of the contraction hypothesis of FitzGerald, Science 13, 390 (1889), and Lorentz (1892), against a fixed bench (what is postulated, what is derived, what the postulates return numerically, what they cannot reach). The step -- matter in motion through the ether contracts along the motion by exactly sqrt(1 - v^2/c^2) -- is postulated with its magnitude reverse-engineered from the Michelson-Morley null result it exists to explain; no dynamics in either text computes it, and the record's verdict is ad hoc, conceded by Lorentz. Computed in the notebook (14 checks, exit 0) at named inputs (L = 11 m, v = 29.78 km/s, lambda = 589 nm): the uncontracted ether theory predicts 0.3686 fringes on rotation; the contraction of the 11 m path is 5.4271e-8 m (54.27 nm, 0.0921 lambda); with the factor the shift cancels at every velocity tried (residuals at most 1.35e-8 fringes, floating-point zero), and a generic shortening (1 - k beta^2) cancels only at k = 1/2 (+0.3686, +0.1843, +0.0000, -0.1843, -0.3686 fringes at k = 0, 1/4, 1/2, 3/4, 1) -- the coefficient is read off the phenomenon. Remainders at full size: the ether stands unobserved; Einstein 1905 returned the factor as a theorem of a different frame (closure by replacement, credited there); the postulate has no second prediction and its universality across materials is spent, not argued. Scout-grade verification per Batch 6 (no verbatim primary reading). EXTENSION. Vacuum catastrophe NOT SOLVED; quantum gravity NOT SOLVED.