Rankings Primer Arbiter Physicist Mathematician Contrarian QE Sites log Claims Sources Research notes

Wheeler-Ken: Arbiter (Step 3)

Subject: Ken L. Wheeler, Uncovering the Missing Secrets of Magnetism, 3rd ed., 25 Sept 2014 (236 pp.). Inputs weighed: claims.md, facts.json, physicist.md, mathematician.md, contrarian.md, deep/, sites.md, research/. Tags: [read] project files; [web] URL given; [memory] with confidence; [calc] recomputed here (NumPy, scratch script).

Stake note: Wheeler wrote the book, sells it ($10, p.2) and promotes a 4th edition; he is the only source for his theory. Timm Vanderelli (ferrocell.us) sells the Ferrocell whose images Wheeler uses and co-authors the Tufaile ferrocell papers. Both are interested parties; their descriptions are reported as theirs.

Source gap: none for the text (one readable 236-page PDF). Coverage limits (superseded by the "QE revision" section at the end: pp.54-96, 97-111, 128-148, 157-176, 177-206 and 207-219 have since been deep-read with page images; read that section first): the three reviewers saw pp.54-96, 128-148, 157-206 as digests only, except pp.97-111, 131-142 and 207-219, which they read from the text layer, not page images. The 4th edition is not in the folder. No document in the project reports a quantitative measurement (one qualitative gauss-meter statement exists, p.145; see QE revision). Every quantitative verdict below therefore rests on arithmetic, units and calculation, not on Wheeler's data.

Where the three reviewers agree

All three agree on: (1) every number in claims 3 and 4 is constructed from phi, none measured; (2) 42.4923 conflicts with the NMR proton value; (3) the book gives no instrument, units, repeats or controls; (4) the Bloch wall is a nanometre-scale internal domain boundary, not the external equatorial plane; (5) "C squared" has the wrong units; (6) the photoelectric "correction" keeps E = hf; (7) the bismuth "overunity" is a misreading of a positive alpha-decay Q-value; (8) the ferrocell patterns are real, with a published optical account that does not use Wheeler's theory. Differences are in verdict labels and in a few details, settled below.

Recomputed numbers (where reviewers differed or I wanted to be sure)

Proton gamma/2pi: NIST value 42.577 478 461(18) MHz/T [web: https://physics.nist.gov/cgi-bin/cuu/Value?gammapbar, cited by all three reviewers; I did not refetch it]. Wheeler's 42.4923 is 0.085178 MHz/T low = 0.2001% [calc]; 0.085178 / 1.8e-8 = 4.73 million standard uncertainties [calc]. The mathematician's 4.7 million is right; the deep read's "10^5 to 10^6" understated it. Offset 128 kHz at 1.5 T, 256 kHz at 3 T [calc]. The CODATA 2018 figure in the deep read is superseded (correction below).
phi constructions [calc]: 180/phi^3 = 42.49224; 180/phi^2 = 68.75388; 360/phi^2 = 137.50776; 360/phi^3 = 84.98447; 2 phi = 3.23607. All three reviewers' tables are right.
Field-ring test (claim 4): I re-ran a Biot-Savart integral for a uniformly magnetised cylinder (side-wall current sheet, coarse 120-segment grid). Radius where B_z changes sign: L/D = 1: 1.25 R at h = 0.1 R, 1.55 R at 0.25 R, 1.97 R at 0.5 R; L/D = 1.618: 1.21 R at 0.05 R, 1.40 R at 0.1 R, 1.82 R at 0.25 R [calc]. This agrees with all three reviewers to about 0.05 R (my grid is coarser). The ring moves outward with height and passes 1.236 R at one low height only.
Bismuth levitation (claim 6): the reviewers differ. The mathematician computed a force between a 3 mm N45 cube and a 1-inch N50 lifter and concluded the bismuth cannot set the height; the physicist called the height difference plausible as an uncontrolled geometry or crystal effect. I recomputed with m(lifter) = 18.6 A m^2, m(cube) = 0.029 A m^2, cube weight 1.99 mN [calc]: the dipole force between the two magnets equals the weight at 113 mm (4.45 in), so the quoted "5 inches" is what the two magnets alone give. Diamagnetic force of a semi-infinite bismuth slab (chi = -1.66e-4) on the cube, F = 3U/h with U = (|chi|/2mu0) pi k^2/(2 h^3) [calc]: 2.6 mN at 1 mm gap, 0.16 mN at 2 mm, 0.03 mN at 3 mm, 0.004 mN at 5 mm, about 1e-8 mN at 127 mm. Refinement of the mathematician's conclusion: the bismuth cannot move the cube from 5 inches away, but the book does not say how far the disk is from the cube. If the disk sits within 1 to 2 mm of it, the diamagnetic force is 8% to over 100% of the weight, so a differently shaped, positioned or oriented disk could shift the float height by millimetres. A larger, "much lower" shift in the photograph pair needs a change in something else (disk position, lifter position, crystal orientation). The book's own caption says only "same lifter magnet", not same lifter position (p.99 [read]). Also the sign of the cube-lifter force is unclear from the caption (hanging below the lifter by attraction, or repelled); the distance estimate is the same either way in this model, and the bismuth is the stabiliser in either case.
Photoelectric example: the reviewers used a sodium work function of 2.28 eV (memory). A web check gives about 2.75 eV for sodium in the Wikipedia table [web: https://en.wikipedia.org/wiki/Work_function], and the value depends on surface and contamination. With 2.75 eV, a 400 nm photon (3.10 eV) gives K_max of about 0.35 eV, not 0.82 eV; a 600 nm photon (2.07 eV) and a 700 nm photon (1.77 eV) give no emission in either case. The qualitative conclusion (a threshold exists, independent of intensity) is unchanged; the quoted 0.82 V stopping potential should not be used. Use "about 0.35 to 0.8 V depending on the work function assumed".

Web verification

Bi-209 alpha decay: half-life (1.9 +/- 0.2) x 10^19 years, 3.14 MeV line, Orsay (Coron et al.), April 2003 [web: https://physicsworld.com/a/bismuth-breaks-half-life-record-for-alpha-decay/]. Retags the physicist's and contrarian's [memory/web-via-search] items. The Gran Sasso 2.01 x 10^19 figure (mathematician) I did not recheck.
Sodium work function: see above (2.28 versus 2.75 eV; retagged [web]).
Not rechecked (do not bear on a verdict): Krafft as author of "The Ether and Its Vortices" [memory: medium]; bismuth susceptibility anisotropy; electrolysis MHD swirl [memory: medium] (a plausible mundane explanation, not a finding); Br values for N42/N52 [memory: medium].

Claim-by-claim

Claim 1. Magnetic attraction does not exist; it is "dielectric voidance"; a magnet is an electrified dielectric object

Agree: the only formula (D.V. = M1 x M2 / d^2, p.30) has no units, drops G, and the far-field magnet-magnet force falls as 1/d^4, so the formula fails for magnets even on its own terms. Disagree: only on label for the strong form (physicist: Tested, not supported; mathematician: Internally inconsistent for the formula, Not testable for the weak form; contrarian: Tested, not supported, with a note that no one ran Wheeler's version). Stronger argument: the contrarian and physicist, with a caveat from the contrarian that existing force measurements are not tests of Wheeler's hypothesis. A magnet attracting iron in vacuum with no charge, Stern-Gerlach deflection of neutral atoms and Curie-point loss of pull [memory: high] are controlled, replicated results that a "dielectric" cause does not predict. Wheeler states no permittivity-dependent prediction, so it has been ignored, not tested and failed under its own terms. His 2025 Substack reidentifies "magnetic attraction" with gravity [web: https://kenwheeler.substack.com/p/the-codex-of-gravity, via sites.md], a different mechanism from the 2014 book: the claim is not stable across his own writings, and the demonstrations he cites support neither version in particular. What would settle it: weigh the pull on a steel ball from a magnet at fixed distance in air, oil and water (relative permittivity about 1, 2, 80); a dielectric mechanism predicts a change, the magnetic one predicts none (mu is about 1 for all three). Final verdict (split): the renaming form ("voidance" for attraction, no new prediction): Not testable as stated. The strong form (the force is dielectric, not magnetic): Tested, not supported by routine independent measurement; Wheeler's own version was never tested. The formula D.V. = M1 M2 / d^2 as a law for magnets: Internally inconsistent (units) and in conflict with dipole falloff.

Claim 2. Magneto-dielectric hyperboloid, Poincare-disk double vortex, "dielectric inertial plane" = Bloch wall, space a "posterior attribute" of magnetic divergence

Agree: the standard geometry reproduced is correct; the p.125 metric lacks a factor 4 (curvature -4 as printed, -1 with the 4 [calc, mathematician]); the pictured one-sheet hyperboloid is not the two-sheet surface of the Poincare construction; "0 = 1 = infinite inertia" read literally is a contradiction; outside a magnet curl B = 0 and div B = 0, so there is no circulation to be a vortex; the equatorial B_z zero is a symmetry property; the Bloch wall identification is a category error. Disagree: the mathematician gives "Established" to the geometry as geometry and "Plausible, untested" to the Tufaile horocycle bridge; the physicist and contrarian give Not testable. Both are right about different objects. Related material kept: Tufaile, Snyder and Tufaile, "Horocycles of Light in a Ferrocell", Condensed Matter 6(3):30 (2021), which models ferrocell images with hyperbolic geometry and does not cite Wheeler [web: https://www.mdpi.com/2410-3896/6/3/30]. It models how light scatters from field-aligned particle chains ("magneto-optical effects and not just magnetic effects", Tufaile guide [read via Investigator]); it is not a hyperbolic model of the magnet's field. Vanderelli, who co-authors, has a stake. Also kept: Gardi's (2020) Hall-sensor zero at the equator; sympathetic to Wheeler's vocabulary, and the zero is expected from symmetry. Stronger argument: the physicist and mathematician (curl-free field; the quantity "cannot be measured" per Wheeler, pp.116, 117, 123). The contrarian's best case (filings are a 2D slice and mislead; hyperbolic geometry does appear in ferrocell optics) is real but does not reach the field. What would settle it: a three-axis Hall map of a cylinder magnet: any azimuthal component of B (a swirl), or any departure of the field from the calculated current-sheet solution, would count for Wheeler; the calculated field already matches published maps to percent level [memory: high]. Final verdict (split): standard Poincare-disk geometry as mathematics: Established (with the slips listed). Identification of the magnet's field with it, "0 = 1 = infinite inertia", and "space is a posterior attribute": Not testable as stated, and Internally inconsistent where read literally. The sub-claim that the external field has a vortex or spiral structure and that the equatorial plane is the Bloch wall: Tested, not supported by magnetostatic theory and field mapping (not by a test of Wheeler's own measurements, of which there are none). Related ferrocell-horocycle work: Plausible, untested as a bridge, which its authors do not draw. Newton's first law objection (p.113): Not testable as stated (definitional).

Claim 3. phi fixes the geometry: 42.4923 = 180/phi^3, 68.7539, 3.23606:1, 137.5077 with an 85 degree sector

Agree on all arithmetic. The numbers were already in the 2nd edition (1 July 2014, 110 pp) before any ferrocell, bismuth or field-formula material [web: https://web.archive.org/web/20140731145835/http://www.kathodos.com/magnetism.pdf, via Investigator text-layer search]: construction preceded demonstration. Disagree: on whether "omega = gamma e H" counts e twice. The mathematician is right to condition it: Wheeler defines gamma on p.48 as an undefined proportionality constant, so the dimensional error holds only if gamma is the NMR gyromagnetic ratio, which is how the book labels it elsewhere. I adopt the conditional wording. Verdict labels differ slightly (physicist: Not testable for the ratio, Refuted for the spiral; mathematician and contrarian: Never properly tested for the spiral). The spiral is refuted for an axisymmetric static magnet only (B_phi = 0 by symmetry); for the ferrocell optical rosettes it is simply unmeasured, so the contrarian's wording is the fairer one for what Wheeler pictured. Strongest honest points (contrarian, kept): golden ratio does appear in real magnet physics (CoNb2O6 Ising chain, E8 excitation ratio [web: https://arxiv.org/abs/1103.3694; the exact measured number was not confirmed]) and in phyllotaxis; Wheeler says precession is "nothing of my discovery" (p.44). None of this connects to his constants. What would settle it: any proton NMR instrument in a known field. The measured frequency/field ratio is 42.5775 MHz/T to parts per billion; 42.4923 would be off resonance by 256 kHz at 3 T. Reader exercise: convert 42.4923 deg to radians and 42.4923 MHz/T to rad s^-1 T^-1 (2.675e8 for the true value [memory: high]); no physical constant relates them. Final verdict (split): the identities among the constants: Established (as pure algebra). "42.4923 is the proton's true gyromagnetic ratio; the published value is incorrect": formerly "Refuted by calculation", relabelled in the QE revision to Tested, not supported (the refutation is a measured constant, not a calculation from accepted laws); the degrees = MHz/T identification is Internally inconsistent (unit conflation). "Ideal precession angle of a magnet" and "dielectric:magnetic = 3.23606:1": Not testable as stated (no definition or units; C/Wb is siemens, so the ratio needs a stated normalising impedance [calc, mathematician]). 137.5 degree field-line spiral in a real magnet: Never properly tested for ferrocell patterns and Refuted by calculation for the static field of an axisymmetric magnet.

Claim 4. "Magnetic & Dielectric Field Formula" (7-28-2014): radii 0.23606, 0.381966, 1, 1.23606

Agree on all points and the three independent integrals match (including mine). The numbers are exact phi identities (phi^-3, phi^-2, 1, 2/phi). Several captions are wrong ("C2 + C3 = Phi cubed" is sqrt 5 in C3 units). The mathematician adds a useful point I keep: by continuity a zero ring that moves from 1.0 R to beyond 2 R must cross 1.236 R at some height, so a match at one unspecified height has no evidential weight, and standard magnetostatics is scale-free, so fixed ratios to size are not absurd in form. Stronger argument: all three agree; the physicist's caveat (the formula says "mean field" without component or height) is accepted. Wheeler deserves credit for making an exact, falsifiable statement, which the contrarian notes; he then hedges it ("extends both further and shorter", p.152). What would settle it: Hall-probe scan of B_z over a disc magnet (L/D 1 and 1.618) at heights 0.05, 0.1, 0.25, 0.5 R: fixed ring at 1.236 R means Wheeler's strong form; ring moving as tabled means magnetostatics. In a ferrocell, photograph at several gaps and with LED rings of different LED counts and measure ring radii against a scale. Final verdict (split): arithmetic of the four rules: Established (as phi algebra). Height-independent form (these are fixed features of every such magnet): Refuted by calculation. Ferrocell-ring reading (the rosette radii as stated): Never properly tested. Weak form (some ring crosses 1.236 R at some height): Plausible, untested, expected by continuity.

Claim 5. Mainstream foundations are wrong (no electrons, photons, relativity or QM; Newton's first law; "C squared"; Z-axis radial dielectric component of light; photoelectric effect "corrected")

Agree: "C squared" is a unit error; Newton I is definitional; the photoelectric "correction" keeps E = hf (p.217: "h Planck and f frequency remain the same") and Wheeler's own pages concede intensity-independent energy (p.219); the mathematician's derivation that a vacuum plane wave has no longitudinal field component stands, and real finite beams (including twisted light) have a small longitudinal part (about lambda/(2 pi w) [memory: high]) that standard theory derives without a "dielectric carrier"; the 99.9999999999% "empty space" energy remark fails the mass-defect budget (fission about 200 MeV = 0.091% of the nuclear rest energy [calc, mathematician]). Disagree: the physicist gives "Refuted by calculation" to "no photons/electrons" in part; the mathematician and contrarian give Tested, not supported. The latter is more accurate: refutation by calculation applies to specific wrong statements (units, threshold, energy budget), while "no electrons or photons" is contradicted by experiment (electron charge/mass, photoelectric threshold, Compton scattering, single-photon counting [memory: high]) against an alternative that states no calculation to compare. I also drop the physicist's first use of 0.82 V in the sodium example (see recomputed numbers). Points kept in the record: the Poynting-vector picture (energy flows in the space around wires; p.15 quote of Dollard) and 1/c^2 = mu0 eps0 are real physics that Wheeler uses correctly; Eric Dollard's conjugate dielectric/magnetic vocabulary [read: research/files/Dollard PDFs, via Investigator] and the Tesla and Heaviside quotations are a historical lineage, not data. Tone toward Einstein and Feynman is abusive (pp.11, 29, 116 [read]) and is part of the record; it is not evidence either way. What would settle it: a Millikan-type stopping-potential measurement against frequency (slope h/e = 4.136e-15 V s); a Wheeler model must reproduce a threshold and that slope. For the Z-component: any measured longitudinal field in a vacuum plane wave beyond the finite-beam value. Final verdict (split): "C squared" as speed limit: Internally inconsistent. Photoelectric "correction": Internally inconsistent (retains E = hf, then claims quadratic or exponential growth). "Z-axis radial dielectric component" as a longitudinal vacuum field of plane waves: Refuted by calculation; read as the energy-flow (Poynting) axis it is Established but not new; as Wheeler's separate carrier: Not testable as stated. "Empty space" as the source of nuclear energy: Refuted by calculation. No electrons, photons, relativity or quantum mechanics: Tested, not supported. Newton's first law "wrong": Not testable as stated. Restored Ether as metaphysics: Not testable as stated.

Claim 6. Demonstrations confirm the model and promise technology

The reviewers rightly split this claim; I keep six parts apart (the physicist asked that I do).

6a. Ferrocell and ferrofluid patterns. The patterns are real and reproducible. The makers' own account is diffraction and absorption by field-aligned nanoparticle chains. Spikes in his suspension (pp.22-23) are the standard normal-field (Rosensweig) instability [memory: high]. Established as a mainstream optical account; as confirmation of Wheeler's model: Never properly tested (no quantitative comparison, no controls). 6b. RGB LED "discovery" (18 June 2014). Pages 133-136 [read by contrarian and physicist]: white along the equator, red and green along the poles, blue nearly absent; exposure varied by colour ("Exposure times of the camera must also be considered", p.135). Mundane candidates not excluded in the book: stronger blue absorption by the ferrofluid (stated in the Tufaile guide), wavelength-dependent diffraction, LED and camera differences. Confirmation: one photograph by Kris Lang of Kirkland Lake (p.136), no second source found by the Investigator; I record it as one documented photograph, not as fabricated. Wheeler says Vanderelli "confirmed this phenomenon has never been seen before" (p.133): his statement, not seen elsewhere. Open date puzzle: the 18 June 2014 date precedes the 2nd edition (1 July), whose text layer lacks it; page images not checked. Never properly tested. 6c. "Programmed" bismuth. Described in one sentence ("applied charge and field in casting", p.99); no field, voltage, temperature, mould, mass or repeat count. The caption pair is not controlled for lifter or disk position. By my calculation the height of about 5 inches is set by the two magnets; bismuth supplies stability, and can move the float height only if the disk is within a few millimetres. The contrarian's strongest honest case is kept: bismuth diamagnetism is crystal-direction dependent (about -1.24e-4 along a/b, -1.76e-4 along c [web, via search summary, primary table not read: https://www.mrs-j.org/pub/tmrsj/vol25_no1/vol25_no1_069.pdf]), and casting in a field might alter grain orientation; no paper measuring that for bismuth was found. The same chapter says Bi-209 is stable and its measured half-life "spurious" (pp.98, 99, 102, 103, 109) while quoting the Orsay result (1.9 +/- 0.2) x 10^19 y with a 3.14 MeV line (p.103); the Orsay group counted 128 alpha events [web: physicsworld above]. The claim that "bismuth breaks the rule of conservation" mistakes the alpha-decay Q-value (3.14 MeV) for a violation: Refuted by calculation. The claim that bismuth has "the highest dielectric permittivity" (pp.106-108) is the physicist's note, not independently checked by me; bismuth is a metal [memory: high]. Primary record: VIDEO 73 (2014-08-17) [web: https://www.youtube.com/watch?v=LeQ057E8v5w], not viewed by anyone. Verdict for the observed difference: Never properly tested (no replication, no controls, not dismissed). Verdict for the mechanism "dielectric inertia": Not testable as stated. Verdict for a diamagnetic lift at 5 inches: Refuted by calculation (about 1e-8 mN there). 6d. Objects resting at the "gyromagnetic precession angle" (pp.136-142): ordinary static equilibria of magnets and steel; no angle is measured. Never properly tested; an ordinary-equilibrium account is available. 6e. Vortex videos and independent experimenters. Gardi (2020), homemade Hall-sensor gauss meter, south pole, same face, N42 magnets: 1.44 V maximum for 5 mm to 2 in magnets, zero at the equator; stronger grade gives no greater throw [web: https://wiki.naturalphilosophy.org/index.php?title=On_The_Fractal_Nature_of_Magnets_(and_the_universe)., a secondary wiki summary; video not viewed]. Weight: grade effect expected about 10% at the surface and about 3% in distance (throw scales as the cube root of the moment [calc]), so her result rules out a large effect only, and the identical 1.44 V across sizes may indicate sensor saturation (contrarian). Tested, not supported for "a stronger grade has a visibly greater throw", at the resolution of one homemade test. Tinmanpower (an Australian electrician) videos, per Wheeler's 2026 account only ("proved it's 100% accurate"; videos not viewed); a bubble swirl over a pole during electrolysis has a standard Lorentz-force explanation [memory: medium]; the Ufopolitics 2015 induction test on the Energetic Forum (pages 2-27 unread). Never properly tested (no neutral replication with data). 6f. "Vortex technology" (100+ TB drive, vortex communication, tissue regeneration) and the "permanent battery in two dimensions": no device, data or patent (none found on Google Patents by the Investigator). Not testable as stated.

How the claims were treated: nothing suggests data were omitted or publication blocked; Wheeler never entered journals, has no patent and no peer-reviewed paper found, and the Tufaile ferrocell papers do not cite him. "Ignored by the mainstream" and "did not submit" both fit the record.

Absence-of-evidence analysis: the 4th edition and later record

What is missing: the 4th edition, announced "very soon" on p.235 (Sept 2014), still "scheduled" per emediapress in 2018 and "to expand over time" per Substack 2024-2026; Wheeler's original vortex forum thread; the 67 MB kathodos "magnetism" file (2015-10-14); any patent or journal paper; any second record of Kris Lang; the content of the 2018 conference talk's data (a 90-minute talk, not viewed). Competing explanations: 1. The work moved into videos and Substack posts (for: 136 Substack posts and a video series exist [web via Investigator]; against: no document replaces the book with data). 2. The 4th edition exists as the 480/580-page version listed by booksellers (for: StoryGraph lists 580 pages and Spotify blurbs "Pages: 480" for the 3rd edition, which conflicts with the 236-page PDF; against: unchecked, the 3rd edition PDF is 236 pages). 3. Abandoned or indefinitely postponed (for: nine years without publication; against: ongoing activity in other media). 4. The author prefers not to submit work to review (for: hostility to degrees and review, pp.9, 11 [read]; against: none directly). 5. Suppression or blocking: no evidence in the record for it; no refusal by a journal or funder is documented, but the record would be thin by nature for an outside author. Deciding records: the Amazon "Look inside" page count and edition; the 67 MB kathodos file (Wayback, 2015-10); the 2018 talk and VIDEO 73 and Magnetic Egg videos; Kentucky business registration for Darkstar Publications; trademark record for Ferrocell (86185455) to fix Vanderelli's ownership dates. Ferrocell USA's closing notice ("2014 to 2026", sites.md) may take the reference archive offline; Wayback copies should be kept.

Corrected claim list

Magnetic attraction does not exist (dielectric voidance). Final: renaming form Not testable as stated; strong form (force is dielectric) Tested, not supported; the D.V. formula Internally inconsistent. (Wheeler's 2025 version, attraction = gravity, differs and is also unmeasured.)
Poincare-disk magneto-dielectric hyperboloid; inertial plane = Bloch wall. Final: standard geometry Established; identification with the field Not testable as stated and literally Internally inconsistent; vortex/spiral structure and Bloch-wall identification Tested, not supported; ferrocell horocycle work Plausible, untested as a bridge.
phi constants (42.4923, 68.7539, 3.23606, 137.5077, 85). Final: identities Established; proton value replacement Tested, not supported (relabelled, QE revision); angle/MHz-per-tesla identification Internally inconsistent; ratio and "precession angle" Not testable as stated; field-line spiral Never properly tested (ferrocell), Refuted by calculation (axisymmetric static field).
Field-formula radii (0.236, 0.382, 1.236 times the radius). Final: arithmetic Established; height-independent form Refuted by calculation; ferrocell-ring form Never properly tested; weak one-height form Plausible, untested.
Mainstream foundations wrong. Final: split as above; chiefly Tested, not supported (no electrons, photons, relativity, QM), Internally inconsistent (C squared; photoelectric retains E = hf), Refuted by calculation (plane-wave Z-component; empty-space nuclear energy), Not testable as stated (Ether, Newton I, Wheeler's Z-component).
Demonstrations confirm it and promise technology. Final: Never properly tested for ferrocell as confirmation, RGB LED, programmed bismuth (observation), resting positions and vortex videos; Established as the mainstream ferrocell optics; Refuted by calculation for bismuth lift at 5 inches and bismuth "overunity"; Tested, not supported for grade-dependent throw (Gardi, one homemade test); Not testable as stated for vortex technology. QE revision adds 6g (CRT), 6h (gyroscopic weight loss) and 6i (tabletop demonstrations, pp.54-96), and sharpens 6b and 6c: see the QE revision.

Corrections made by the claim list to claims.md / facts.json wording: - Claim 6 in claims.md bundles unlike items; read it as 6a-6f above. - "levitates a 3 mm cube 5 inches above it" (facts.json, claims.md) overstates the bismuth: a 1-inch N50 lifter magnet does the lifting; the bismuth stabilises (p.99).

What would settle it (in order of cheapness)

Proton NMR in a known field (claim 3): 42.5775 versus 42.4923 MHz/T.
Hall-probe scan of B_z over a disc magnet at several heights (claim 4): fixed or moving ring.
Weighing a steel ball's pull in air, oil, water (claim 1).
Ferrocell with ring lights of different LED counts, several gaps, scale in frame; RGB repeat with the magnet removed, LEDs swapped, fixed exposure and white balance, with and without a polariser, and a measured absorbance spectrum of the fluid (claims 3, 4, 6b).
Two bismuth discs cast alike except for Wheeler's field/charge step; same lot and mould; measure mass, size, volume susceptibility (Faraday or SQUID), grain texture (XRD), disk-to-cube gap and lifter position, blind height measurement (claim 6c).
Millikan-type stopping-potential versus frequency measurement (claim 5).

Open questions

What does "applied charge and field in casting" mean in practice, and does a disc cast that way differ in susceptibility or grain texture? (VIDEO 73 may say.)
Does the Faraday-effect reference on p.133 imply a polariser was used in the RGB LED photographs?
What did Gardi's ferrocell-lighting test conclude (the wiki summary omits it)? What do Gardi and Snyder (2024, "Finding the Right Angles of a Ferrocell", blocked) and Dave and Mehta (Optik 2024) measure about ferrocell angles?
What do pages 2-27 of the Energetic Forum thread (2015) report?
Is the 4th edition the 480/580-page work? What is in the 67 MB kathodos file?
Why does the RGB LED discovery (18 June 2014) not appear in the 2nd edition (1 July 2014)? Check the 2nd edition page images.
Is Kris Lang's photograph confirmed anywhere?

Corrections to earlier documents

deep/magnetism1small-6bca80.md, pp.43-53, items 4(b) and 8: CODATA 2018 value 42.577 478 518(18) MHz/T should read the current 42.577 478 461(18) MHz/T; and "10^5 to 10^6 times the published uncertainty" should read about 4.7 x 10^6 [calc above].
Same section, item 4(e): "omega = gamma e H counts e twice" holds only if gamma is the gyromagnetic ratio; Wheeler defines it on p.48 as an undefined constant.
claims.md Coverage note and facts.json how_read: pp.97-111, 131-142 and 207-219 were read from the text layer in Step 2; the bismuth setup includes a 1-inch N50 lifter (p.99); page images of these ranges still not viewed.
facts.json: "levitates a 3 mm neodymium cube 5 inches above it" should say the lifter magnet sets the height and the bismuth stabilises it; and the book claims bismuth is stable while quoting its measured half-life (pp.98-103, 109).
facts.json mainstream_conflict: add that the one-sheet versus two-sheet hyperboloid conflation is also incorrect, as is "only three infinite rays can intersect the centre" (p.114).
claims.md says p.46 gives the "incorrectly published" value; Wheeler writes 42.5781, which matches neither CODATA nor the shielded-proton value (42.5764 [memory: medium]).
Physicist and mathematician sodium numbers: the sodium work function used (2.28 eV, [memory]) differs from the Wikipedia table value (about 2.75 eV); a 400 nm photon gives K_max of about 0.35 eV, not 0.82 eV. The threshold conclusion is unchanged.
sites.md, Gardi lead: add the scale-invariance caution (expected grade effect about 10% at the surface, 3% in throw; identical 1.44 V across sizes may be saturation). Investigator's lead itself is not wrong.
Physicist claim 4 text ("1.14 R at h = 0.05 for L/D 1" etc.) matches independent calculations within 0.05 R; no change.

QE revision (Step 4b remediation, 2026-10-08)

Inputs: qe.md (verdict MINOR) and the six new deep reads appended to deep/magnetism1small-6bca80.md: pp.54-96, 97-111 (with page images of the p.99 photos and p.105 castings), 128-148, 157-176, 177-206 and 207-219. Four web checks were made (below). Result: no headline verdict (claims 1-5) changes. One label changes (proton value), one reviewer criticism is withdrawn (p.35 field names), claim 6 gains three sub-parts (6g, 6h, 6i) and 6b and 6c are sharpened. Tags as before; [calc] = recomputed by a deep reader or by QE from the project's example.py, not rerun by me.

Coverage, now

Every page of the 236-page book has been read beyond the digest. Page images were viewed for pp.59, 63, 64, 66, 68, 70, 73, 76-81, 85, 91 (core chapter); 99, 100, 105 (bismuth); 133-136, 138, 141, 145, 146 (RGB, resting positions, "quadrapole"; pp.137, 139, 140, 142, 147, 148 text only); 157-176 (all); 183, 191 (formula table, ring-drop frames); 210, 212, 217 (light). Not viewed as images: pp.211, 213-216, 218-219 (captions only) and the remaining pages of pp.177-206. The p.18 Dollard "10-C" oil-cube demonstration, the p.75 Alaska pipeline remark and the minor factual statements QE listed (item 9) are still covered by the digests only; they bear on no verdict. Unresolved: deep/ line 653 refers to a "section (pp.97-111)" that did not exist when it was written; the new pp.97-111 read now exists, so the reference is satisfied, but the older section's claims were not redone.

Web checks made in this revision

Gyroscope weight loss: Hayasaka and Takeuchi, "Anomalous weight reduction on a gyroscope's right rotations around the vertical axis on the Earth", Phys. Rev. Lett. 63, 2701 (18 Dec 1989): three motor-driven rotors (brass, aluminium, silicon steel) were reported to lose up to about 12 mg when spun in one sense [web: https://www.skeptic.org.uk/?p=49388 and https://journals.aps.org/prl/issues/63/25]. J. E. Faller and colleagues (NIST and Univ. of Colorado) repeated it over six days and found no anomalous weight change depending on rotor speed or direction ("The Gyroscope-Weighing Experiment Revisited - With a Null Result", same sources). A Nature commentary by S. H. Salter argued that gyro imperfections can produce spurious apparent effects [web, same search]. The additional names QE recalled (Nitschke and Wilmarth; Quinn and Picard) were not confirmed by my search: [memory: medium], unchecked. Sources disagree about which sense of rotation lost weight; the paper itself was not read.
Laithwaite: the Royal Institution's own account says he was Professor of Electrical Engineering at Imperial College 1964-1986, gave the Friday Evening Discourse "All Things are Possible" on 8 Nov 1974 on gyroscopes, met "utter hostility", and the text failed peer review for the Institution's Proceedings; it was later published independently [web: https://www.rigb.org/explore-science/explore/blog/gyroscope-engineer-and-christmas-lectures]. Wheeler's "retired in 1981" (p.181) matches a mailing-list quotation (Extropians archive, https://diyhpl.us/~bryan/irc/extropians/extracted-extropians-archive/archive/0208/110459.html) but not the Royal Institution; I record both and prefer the institutional source. The "hostility" part of Wheeler's quotation is corroborated. The 50-pound wheel remark was not checked.
Shielded-proton value: 42.5764 MHz/T is the rounded value for protons in a spherical water sample at 25 C (42.576 384 74(46) in one NMR calibration paper); the free proton is 42.577 478 518(18) (CODATA 2018 as quoted there) or 42.577 478 461(18) (current value I used above) [web: search summaries of arXiv 2410.08279 and an NMR calibration paper; primary tables not opened]. This confirms arbiter correction 6: Wheeler's "incorrectly published" figure 42.5781 (p.46) matches neither.
Electrolysis swirl (6e): bubbles over a magnet in electrolysis are driven by the Lorentz force on the ion current (magnetohydrodynamic flow); rotation at magnet and electrode edges is documented, and the sense reverses with field and current direction [web: https://www.fzd.de/publications/Publ-25616 and the 2017 Weier et al. paper, Int. J. Hydrogen Energy, via search; not opened]. I found no source describing a vortex at a pole face specifically, so the mundane account is plausible and not proven for Tinmanpower's set-up (videos unviewed).

Verdict changes

Proton value (claim 3). "42.4923 MHz/T is the true gyromagnetic ratio": was Refuted by calculation (all four files). QE is right: the scale reserves that label for what accepted laws rule out with no experiment. Here the ruling-out is a measured constant (NMR, parts in 10^9, replicated worldwide), and the 0.2 % gap (4.7 million standard uncertainties) is arithmetic on that measurement. New label: Tested, not supported. The substance is unchanged and no weaker. The reasoning "degrees are not MHz per tesla" stays Internally inconsistent.
p.35 "swapped field names": withdrawn. The deep read (pp.1-42 section) and mathematician.md called Phi/(L x T) "dielectric field" and Psi/(L x T) "magnetic field" swapped. Check: Wb/(m s) = V s/(m s) = V/m, electric field strength; C/(m s) = A/m, magnetic field strength H [calc, QE and the pp.177-206 reader agree]. This is the Steinmetz/Dollard dual convention, used consistently with "E = Phi/time, I = Psi/time" (p.183). The siemens point stands: a dielectric:magnetic ratio has units C/Wb, so 3.23606 needs a stated impedance. No verdict used the "swapped" point.
Gauss-meter statement (p.145). "As any gauss meter will register to you, the highest pressure points on a magnet are its divergent centrifugal edge, and secondly its convergent centripetal point at the center of every polarized side." The pp.128-148 reader confirms it is an assertion: no meter, probe, magnet, distance or value. It is, however, a qualitative instrument claim, so "no document reports a measurement" is corrected to "no quantitative measurement". For the edge half the textbook agrees: example.py gives |B| at r = R of 0.405 Br against 0.402 Br on the axis at h = 0.1 R, and the edge peak grows nearer the face [calc, QE]. That is a point in Wheeler's favour as an observer; the "quadrapole" reading (p.146: "every dipolar magnet is in reality a quadrapole") is his own, uses "quadrupole" in a different sense from physics, and is illustrated with borrowed meteorological diagrams (one watermarked Dennis Tasa), not magnet data. Verdict: observation Established (qualitatively, agrees with magnetostatics); "quadrapole" reading Not testable as stated. Same page-143 rule "higher gauss = shorter reach": see 6i.

New and sharpened sub-parts of claim 6

6b (RGB LED), sharpened. Page images and text of pp.133-136: a white RGB package, "greatly filtered light" (filter unspecified), N50 1-inch cube magnet; no exposure time, f-number, ISO, camera, LED current, LED-to-cell distance or polariser anywhere; the book itself says exposure must be considered and blue needs "longer exposure times" (pp.133, 135). The Faraday effect Wheeler invokes (rotation of polarisation) needs polarised light and an analyser; none is mentioned. Internal differences: p.133 says green is "far less" present than red and blue barely present at the poles, p.134 says red and green lie along the poles; the book does not reconcile them. The colour scale figure (2.95 ... 1.77 "dielectric energy") is photon energy in eV, borrowed and relabelled. Confirmation: Kris Lang's photo (p.136) is almost entirely red-brown with no visible colour separation, no settings or date, and the word "confirmation" is Wheeler's caption; Vanderelli's confirmation is Wheeler's statement only, with no quote or date. All of this is recorded as documented; none of it is a refutation. Mundane candidates remain unexcluded (blue absorption in the fluid, per Tufaile; per-colour exposure; camera channels). Verdict unchanged: Never properly tested. The control that would decide it (magnet removed, fixed exposure per colour, with and without polariser) is not shown.

6c (programmed bismuth), sharpened by the p.99 and p.105 image reads. - p.99 photo pair [read: page image p.99, deep reader's pixel estimates, rough]: in both photos the 3 mm cube sits within roughly 0.6 to 0.9 mm of the disk dome (scaled from the cube's width; no ruler in frame). No lifter is visible on the left; on the right an unidentifiable dark object is in a corner. The two set-ups differ in background (grey versus towel), camera and resolution (822x374 against 427x275 px, the right looks like a video still), and by cube-scale about 40 mm against 34 mm disk width. So the pair does not show the "much lower levitation" the caption claims, nor that the set-ups are "identical". This does not contradict Wheeler; it means the one piece of evidence for the effect does not show it. - Effect on my earlier bismuth calculation: the model above put the cube at 113 mm from the lifter pair and found diamagnetic force negligible at 127 mm and 8 % to over 100 % of the weight within 1 to 2 mm. The photos put the cube within about 1 mm of the bismuth, i.e. in the regime where the bismuth force (about 2.6 mN at 1 mm in my semi-infinite-slab model, which overestimates for a finite domed disk) is comparable to the cube's weight (1.99 mN). So the bismuth is plausibly the stabiliser at that gap, as I said; what remains unshown is any difference between a "programmed" and a standard disk. Disk shape and thickness alone change the field the cube sits in (Wheeler says convex castings are chosen for this, p.100). - p.105 angles [read: page image p.105, deep reader's by-eye estimate; I did not view it]: the "137.5077, 137.5077 and 85 degrees" are an overlay: the left and right photos are two views of what looks like one cast (not "multiple casts"); the drawn Y has arms at roughly 103, 130 and 130 degrees by the deep reader's estimate, not 85/137.5/137.5; the labels are not attached to vertices; the disk is photographed obliquely; and 360 - 2 x 137.5077 = 84.985 (checked: golden angle 360 x (1 - 1/phi) = 137.5078), so "85" is the remainder. Four-decimal angles cannot be read from a photograph. They are the golden-angle partition, drawn and labelled, not measured. Verdict for "casts show 137.5077/85 degree marks, now confirmed": Never properly tested (asserted and drawn; the drawing does not match its labels by one reader's estimate; needs a second reader to confirm that estimate). - Radii (p.97): "143, 176, 190, 154, 156" carry no unit. They match the calculated-radius column of the standard tables in pm (Bi 143, Ti 176, Pb 154, Fe 156) [memory: medium, unchecked]; the empirical column ranks bismuth larger than titanium and iron, which reverses Wheeler's ordering. The same reader notes the book says Bi-209 is stable while quoting the Orsay measurement; unchanged from the verdicts above. - Verdicts otherwise unchanged: observed difference Never properly tested; "dielectric inertia" mechanism Not testable as stated; lift at 5 inches Refuted by calculation; "bismuth breaks conservation" Refuted by calculation. New, from the same chapter: "tesseract has 12 anti-facets" (p.110) is numerology contradicting the real count (8 cubic cells, 24 square faces) [memory: high]: Internally inconsistent with the standard definition it uses; and the "Pacioli: only two copies exist, untranslated" remark (p.106) is a factual statement unsupported in the book [memory: medium, unchecked]. Litvinenko attribution (p.99) is an assertion without a source in the book; I record it and do not weigh it.

6g (new). CRT demonstration (pp.157-162). Documented in Wheeler's own photographs: a camcorder feed of a spiral or square-grid test pattern on a CRT, a hand-held neodymium magnet (2 x 1 inch, 1 inch cube, a round and a rectangular-with-hole magnet) held "very close" or touching; the image twists CW with one magnet face and CCW with the other; with a vertical magnet four closed colour lobes appear ("four loops", which he reads as centrifugal edges and rejects the "four poles" reading of unnamed others). Missing: CRT make, voltage, camera, distance in any unit, pole identification, a measured angle, a monochrome-tube or LCD control. The p.159 "inverse pole" pair shows two different magnets (round and rectangular), so it is not one magnet turned over. Standard account [memory: high, from the deep reader]: an axial field rotates an electron-beam image by an angle proportional to the integral of B along the beam, sign reversing with field direction; a transverse field gives sideways deflection and lobed distortion; and the coloured contours around a magnet on a colour CRT are the known purity (beam landing) error that degaussing coils exist to remove. Illustration by the deep reader (their numbers, anode voltage and field assumed): 25 kV, integral B dz = 5e-4 T m gives about 0.47 rad, about 27 degrees. So the pole-dependent twist and the opposite-sense centre and rim are what ordinary electron optics predicts; the observation is real and ordinary, and no control separates it from Wheeler's "dielectric deflection". One fact stated plainly: the device works by an electron beam, which Wheeler elsewhere denies exists, and he calls it a "cathode ray dielectric projector" (p.157). Verdicts: the phenomenon Established (as standard electron optics); as confirmation of opposite vortex spin at the poles Never properly tested (an ordinary account is available; no neutral control run); his reading of the "four loops" Not testable as stated.

6h (new). Gyroscopic weight loss (pp.60, 181, 190). Wheeler quotes Laithwaite (p.181) and asserts "gravitational polarization is the principle behind gyroscopic weight loss ... nothing to do with torque or angular momentum" (p.190). No test of his own, rotor, rpm, weight reading or person is described anywhere on pp.181-190; the p.60 "dead gyroscope spins up" passage is a thought experiment. Record, with sources above: the anomaly was reported in a refereed journal (Hayasaka and Takeuchi, PRL 1989, up to about 12 mg, sense-dependent); a controlled repeat (Faller et al.) found no effect at that size; the Royal Institution episode shows Laithwaite's demonstrations were rejected for publication by that body, not tested by it. I did not find, and the record does not contain, a test of Laithwaite's own larger claim by a neutral party in this review; the Faller null result addresses Hayasaka and Takeuchi, not Laithwaite. Verdicts: Hayasaka-Takeuchi's report: Tested, not supported (one controlled repeat on record, from web summaries; not independently read here; further repeats named by QE unconfirmed); Laithwaite's large-effect claim: Never properly tested by the record I have (not dismissed without argument, but the sources describe rejection of the lecture text rather than a controlled measurement; low confidence); Wheeler's mechanism: Not testable as stated.

6i (new). Tabletop demonstrations and statements in the core chapter (pp.54-96). The deep read finds no measuring instrument used anywhere in the chapter: no gauss, force, torque, time, angle or voltage reading; repeats stated only in words; no control described. Numbers present are object sizes plus "1.5+ Tesla" (stacked apex magnets), "500 pound" refrigerator, "3 to 5 inch" discharges, "20 bits", "100+ terabyte", "400 % beyond N55", "80 to 90 %". Item by item (all [read: pp.61-91] with [calc]/[memory] from the deep reader): - Copper square (p.76), thin disk magnet against steel disk at the equator of a 2x2x1 block (pp.76-78, pictures on p.78), the "500-pound refrigerator" (p.77), magnetic putty around a 1 cm cube (p.63), the broken ring magnet that rejoins flipped (pp.80-81, with one unbroken control picture, the only comparison object in the chapter), the ring of sphere magnets weakening pull on iron (p.91), two magnets on viewing film (p.91). Wheeler's reported outcome in each matches standard magnetostatics: eddy-current drag depends on flux through the sheet's face (so horizontal drag, little vertical), a permanent disk aligns its moment with the local field while soft steel's plane contains it, the hand supplies the torque that shakes the refrigerator (disk moment about 0.12 A m^2, torque up to about 0.02 N m [calc, disk size assumed]), a ferromagnetic putty flows toward field maxima, fragments stack antiparallel, a closed loop has no net dipole. Correction to qe.md item 6: it is the refrigerator that shakes with the brick on it (p.77), not "a disk shakes a brick". Verdict: phenomena Established (reported direction agrees with standard theory); as confirmation of "dielectric" primacy Never properly tested (no discriminating test: a high-permittivity insulating square against copper would separate the accounts; none was run). No change to claim 1. - Broken-magnet "electrical charges ... shot out like lightning ... 3 to 5 inches" (p.78, "on several occasions"): first-hand, one published source, no photograph or measurement. Mundane candidates: sparks from fresh pyrophoric NdFeB fragments (manufacturers warn of this) [memory: medium] and fracto-emission, normally far smaller than inches [memory: medium]; an electrical arc of 3 to 5 inches in air would need tens of kilovolts (about 3 kV/mm), for which no mechanism is known [memory: high]. Deciding test: break identical magnets in air and in argon, filmed at high speed, with a field mill or electrometer nearby. Never properly tested; not dismissed, and not refuted. - Withheld tests: "I have already written 20 bits of data to a magnetized subject and reproduced the results, by means I will not explain here" (p.61) and the 7/15/2014 inertial-plane discovery "using a testing medium as yet undisclosed" (p.79). Kept as Wheeler's statements; Not testable as stated (methods withheld). Writing 20 bits on any magnet is routine; the novelty claimed ("another plane") is not described. The "100+ TB in a 9.5 mm drive" projection (6f) is covered above. - "400 % beyond N55 ... would fold to 1/6th its surface area" (p.78): N55 is about 55 MGOe, Br about 1.5 T; the ceiling for the highest-magnetisation alloy (Fe-Co, about 2.45 T) is about 150 MGOe [calc, deep reader, Js from memory: high]. Not testable as stated (requires a material beyond the known ceiling; the unit mixes surface and volume). - Pole-linked spin of pyrolytic graphite and of a magnet on bismuth (pp.68, 72, 73, 90): graphite "thrown CW off the north and CCW off the south", the 3 mm cube on bismuth "always" spinning off CW from one pole and CCW from the other ("proof after proof without end"); no trials counted, no camera record, no control. Plain diamagnetic and eddy-current forces go as B squared and do not tell N from S; a polarity-linked sense needs something odd in B. The deep reader proposes a Hall-angle eddy-current effect (both materials have high carrier mobility; Hall angle of order 14 to 27 degrees at 0.5 T for mobility 0.5 to 1 m^2/V s [calc; mobilities memory: medium]) as a mainstream candidate [inference, not established]. This is the one report in the chapter that standard textbook forces do not obviously cover, and I keep it flagged for that reason, without crediting Wheeler's explanation. Verdict: Never properly tested. Deciding test: 20 releases per pole with a fixed push, then a low-mobility diamagnet as control. - "Higher gauss = shallower field" (pp.66-68; "Gauss ratings ... are a complete misnomer"): at fixed shape B scales linearly with Br, so a higher grade gives more field at every distance with the same profile: for a 25 mm diameter by 25 mm cylinder, N42 (Br 1.30 T) gives 0.581, 0.534, 0.359, 0.206, 0.049 T at 0, 1, 5, 10, 25 mm and N52 (1.45 T) gives 0.648, 0.596, 0.400, 0.230, 0.055 T; the half-field distance is 6.9 mm for both [calc]. (I checked the first value: 0.65 T x 25/sqrt(25^2 + 12.5^2) = 0.581 T.) Field depth is set by size and shape. For grade at fixed shape: Refuted by calculation. For cones and stacked "apex" magnets, the shallow high-gauss tip is ordinary flux concentration: Established, not new. This fits Gardi's equal readings across grades (6e) and, by the same calculation, does not need a vortex. - Formula "Psi x Phi = Q" (pp.18, 54, 57, 64; "electrification, in Plancks"): coulomb x weber = (A s)(V s) = J s, the unit of action, not of charge (QE item 7) [calc]. Standard physics has such a product: e x (h/2e) = h/2 for the superconducting flux quantum, 3.313e-34 J s [calc, deep reader]. As written it states a unit and predicts nothing, and it uses Q, the symbol for charge, for an action. Not testable as stated (consistent in units; no content). - Formula "one half the product of dielectric and magnetic density" (p.76): read literally as 1/2 D.B it has units J s/m^4 and is not a pressure (Pa = J/m^3) [calc]; read as Maxwell's 1/2 B.H it is the ordinary magnetic pressure (B^2/2 mu0, about 1.0e5 Pa at 0.5 T) and contains no dielectric quantity, which defeats the use made of it. Around an uncharged magnet D = 0, so the literal reading predicts no force on the steel disk, which is turned strongly. Internally inconsistent, adding a second formula to the claim 1 verdict (the D.V. formula) without changing it. A test: charge the magnet to high voltage; the literal reading predicts a changed torque on the steel disk, standard theory none [inference]. - Not weighed because they are asserted without data or source: the "80 to 90 %" and "90 %" dielectric fractions (pp.60, 76): Not testable as stated; healing arrangements of three block magnets (p.72, no test); Alaska pipeline "dielectrically charged oil" (p.75; geomagnetically induced currents in Alaskan pipelines are a studied effect [memory: medium], Wheeler's mechanism is not); the "100 nanometer scale proof" (p.85) sits over panels whose scale bars read 128 micrometres, so the caption's scale should not be repeated. The p.85 numbers phi^-3 = 0.23607 and phi^3 = 4.23607 are correct arithmetic.

Corrected claim list: changes

Claim 3: proton value replacement is now Tested, not supported (was Refuted by calculation).
Claim 6, added: 6g CRT twist: phenomenon Established (standard electron optics); as confirmation Never properly tested; "four loops" reading Not testable as stated. 6h gyroscopic weight loss: Hayasaka-Takeuchi report Tested, not supported (one controlled repeat from web summaries); Laithwaite's large-effect claim Never properly tested (low confidence); Wheeler's mechanism Not testable as stated. 6i core-chapter demonstrations: phenomena Established, as confirmation Never properly tested; broken-magnet discharges Never properly tested; pole-linked spin Never properly tested (one B-odd report worth a controlled test); withheld 20-bit and 7/15/2014 tests and "400 % beyond N55" Not testable as stated; "higher gauss = shallower" Refuted by calculation for grade, Established for cones and stacks; p.76 pressure formula Internally inconsistent (units); Psi x Phi = Q Not testable as stated (units J s).
Claim 6b: unchanged (Never properly tested), with the missing exposure, polariser and internal inconsistencies recorded. Claim 6c: observed difference Never properly tested, now with the note that the p.99 pair does not show the claimed difference; casting angles Never properly tested (drawn, not measured).
Claim 5 additions: equivalence principle rejected Tested, not supported; "beta is not particles" Tested, not supported; zero-capacitance rule and frequency-proportional-to-capacitance Not testable as stated.
Claims 1, 2 and 4: no change.

What would settle it: additions

Break identical hard-drive magnets in air and in argon, filmed, with a field mill (6i discharges).
Release a magnet or graphite piece 20 times off each pole with a fixed push, then repeat with a low-mobility diamagnet (6i pole-linked spin).
Repeat the CRT test with a monochrome tube, an LCD, marked poles and a measured twist angle against distance (6g).
Cast two bismuth discs alike except for Wheeler's step and photograph both under one fixed lifter with a ruler in frame (6c; already item 5 above, now with the photo evidence added).

Corrections to earlier documents (this revision)

physicist.md, mathematician.md, contrarian.md and arbiter.md (claim 3): the verdict "Refuted by calculation" for "42.4923 is the true proton value" should read Tested, not supported (a measured constant, not a calculation from accepted laws). Arbiter's own text is fixed above.
deep/magnetism1small-6bca80.md (pp.1-42 section, line about swapped field names) and mathematician.md: "the field names on p.35 are swapped" is wrong; Phi/(L x T) is electric field strength in V/m and Psi/(L x T) is magnetic field strength in A/m (see Verdict changes 2). The siemens point stands.
arbiter.md "Coverage limits" and "No document in the project reports a measurement": superseded; p.145 contains a qualitative gauss-meter statement (no numbers), and pp.54-96, 97-111, 128-148, 157-176, 177-206 and 207-219 have now been deep-read. The same overstatement is in primer.md (opening and section 6) and claims.md "Coverage note"; the orchestrator should apply the wording "no quantitative measurement".
facts.json: still says the bismuth levitates a 3 mm cube "5 inches above it" and lists pp.97-111, 133-176 and 207-219 as digest only; apply arbiter corrections 3 and 4 above and the new how_read ranges. claims.md claim 6 "levitates a magnet unusually high" should be replaced by the lifter wording and split 6a-6i.
qe.md item 6 / pp.76-77: it is the 500-pound refrigerator that shakes with the brick magnet on it (p.77), not a disk shaking a brick (p.76). The "picture 1" and "picture 2" figures cited as p.77 are on p.78.
digest p.2: only D. Anderson is "the late"; Martin and Vanderelli are thanked (source lines 105-107). The digest for p.82 places list item 1 on p.82; it begins on p.81.
sites.md: Leads 1 (2nd edition) "no bismuth matches" should read "no bismuth chapter or levitation claim; one passing sentence"; add PDF CreationDate 2014-06-28; Internet Archive row: item created 2014-06-25, current file mtime 2014-09-19; add US8246356 and archive.org item history to recheck and "Hayasaka Takeuchi gyroscope weight" to related work.
Wheeler's date for Laithwaite's retirement (1981, p.181): the Royal Institution gives 1986; a mailing-list quotation gives 1981. Report both; do not state either as settled by this book.
Wheeler's p.159 text presents the left and right photos as the two ("inverse") poles of a magnet; the page images show two different magnets (round on the left, rectangular with a hole on the right). Any later file implying one magnet turned over should be corrected.

Not done in this revision

The Magnetic Egg and VIDEO 73 videos, the Energetic Forum thread pages 2-27, the 4th edition and the 67 MB kathodos file remain unviewed or unobtained. The pp.54-96 pole-linked-spin Hall-angle estimate, the CRT rotation number and the p.105 angle estimates are single-reader results, flagged above; none of them carries a verdict weight beyond "Never properly tested".