O-33 - Massive white dwarfs with very low magnetic fields and ages from 100 Myr to gigayears are observed
These specific stars are the empirical anchor of the white-dwarf bound: each is simultaneously (i) massive enough to stop cosmic-ray-produced neutral black holes, (ii) weakly magnetized enough that cosmic rays reach its surface unscreened, and (iii) old enough that catalyzed decay on <~100 Myr timescales would already have destroyed it. Sirius B (1.0 Msun, ~120 Myr) plays the same role for the interstellar-medium production channel, which is magnetic-field-independent. Quoted mass/age/field values carry ~10% column-density systematics (temperature, composition).
Link to original
Why this is evidence
These stars are exactly the objects that discriminate stability from prompt decay: massive enough to stop cosmic-ray-produced neutral black holes, weakly magnetized enough that cosmic-ray primaries reach the surface unscreened, and old enough that a captured stable hole would already have destroyed them. Under H-24 (stable neutral) their observed persistence at 100 Myr–Gyr ages is strongly unexpected; under H-10 and H-14 it is certain; H-7 is disfavored only in its longest-lifetime regime (holes surviving long enough to be brought to rest and accrete). H-35’s stable behaviours are cut the same way as H-24.
Likelihood
# E-34 (HC-3) — lone edge, one observation O-33 "massive (M >~ 1 Msun), low-field (B_p <~ few e5 G), 100 Myr–Gyr
# white dwarfs". The SHARP white-dwarf anchor of the stable-hole exclusion, covering D<=7 (A-49 approved: D=5,6
# stop in a solar-mass WD, D=7 needs >~1.1 Msun, D>=8 exceeds WD stopping power and is handed to neutron stars →
# E-35). Chain: A-50 copious cosmic-ray production on these low-field stars, A-51 a stopped hole accretes and
# destroys the star in <<Gyr, A-53 no Eddington throttle. Prompt-decay members predict survival with certainty;
# anchor H-10 = 1. Members in HC-3.hypotheses order [H-7, H-10, H-14, H-24, H-35].
lik_wd_H7 = 0.88 # metastable: only its longest-lifetime regime (holes surviving long enough to stop and
# accrete in the WD) is disfavored; the ms-to-years bulk decays first. Slightly less strained
# than by O-22 — WD capture needs the extreme tail specifically
lik_wd_H10 = 1.0 # anchor: prompt Hawking decay, WD survival certain
lik_wd_H14 = 1.0 # near-threshold prompt decay, astrophysically safe — survival certain, matches H-10
lik_wd_H24 = 0.15 # stable neutral: for D<=7 the G-M chain (A-49/A-50/A-51/A-53) makes these old massive
# low-field WDs strongly unexpected — they should already have been destroyed. Nonzero
# because this edge does NOT constrain D>=8 (A-49: those exceed WD stopping power), so H-24
# parameter space with D>=8 survives WDs regardless — that residual room is what E-35 closes
lik_wd_H35 = 0.55 # residual: stable-accreting variants cut as H-24; other unmodelled fates survive. Middling,
# slightly under O-22's value since this is the sharper destructive prediction (rule 3)
t_wd = 0.74 # cap = trust_score(S-37 G-M) = 0.82, docked for the ~10% column-density systematics
# (temperature/composition) on the masses and the cooling-model ages (D-10 SDSS + Hansen
# cooling models) that set both the stopping margin and the age clock — the raw-data-to-
# observation step, not the argument chain (rule 4)
evidence("HC-3", ["O-33"], [lik_wd_H7, lik_wd_H10, lik_wd_H14, lik_wd_H24, lik_wd_H35], t=t_wd)