# Reality revision 1.58 — Frozen hydrogen-isotope holdout

Revision 1.57 proved that one H--D 1S--2S residual cannot identify both the
range and coupling of a Yukawa-like central operator. Revision 1.58 freezes a
single member of that family and asks whether it predicts a second transition.

The range is fixed at the electron reduced Compton wavelength and is not tuned
to the target. The coupling is fitted only to the published 1S--2S central
residual. Without further fitting the operator predicts a `-201.996425 Hz`
correction to the H--D 2S--8S isotope shift.
The published H and D transition frequencies imply
`209691233.7 kHz`. A published bound-state parameterization and correlated
uncertainty calculation give `209691226.436128 +/- 10.367652 kHz`. Adding the
frozen operator gives `209691226.234131 kHz`. The signed standardized error
changes from `-0.700628` to `-0.720112`: it gets slightly worse.

This is a useful negative result. The operator earns no predictive credit, but
the target uncertainty is 51.326 times its predicted effect, so the model is
not excluded either. The historical measurement cannot distinguish the two
models. The result uses public 1997 data and is not described as blind or
prospective.

The prior corrections remain unchanged: complete deterministic closure for a
fixed transition family is idempotent; integer-tick point capture does not
enforce continuous non-crossing; angular impulse has no extra time factor; and
finite observed prefixes do not identify arbitrary generators. The recurrence
learner's 2K-sample continuation guarantee still assumes a noiseless
homogeneous constant-coefficient linear recurrence of order at most K from the
initial state. No result establishes seven physical dimensions, P=NP, a
universal physical grid/tick, a physical string particle, or a theological
cause.

Next, rank unscored transitions by frozen operator contrast divided by full
correlated uncertainty. A worthwhile validation target must have enough
resolution to distinguish the models before its central value is examined.