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

Wheeler-Ken: Physicist review (Step 2)

Reviewer: PhD physicist role. Scope: the six central claims in claims.md (Ken L. Wheeler, Uncovering the Missing Secrets of Magnetism, 3rd ed., 25 Sept 2014, 236 pp.), plus the Investigator's leads (sites.md, research/ notes, research/files/). Tags: [read] from the project documents (I re-read pp.97-111, 133-142, 168-176 and 207-219 of the extracted text myself; the rest rests on the deep reads, which I cross-checked where noted); [calc] done here with NumPy; [web] with URL; [memory] with confidence.

Stance: I state the mainstream view fairly, then judge. Wheeler is the author, seller and promoter of the book (price $10 on p.2) and has no stated degree or laboratory [read: deep read pp.1-42]; that is his stake. Where the vocabulary is only a relabelling, I say so and keep the material in the record.

Coverage and source limits

One primary document, by the inventor. No measurement, instrument reading, repeat count or control appears in any section I read. Wheeler says the root quantity "cannot be measured" (pp.116, 117, 123 [read via deep read]).
Pages 54-96, 128-132, 143-148, 157-167, 177-219 were read by digest only in Step 1. I read pp.97-111, 133-142, 168-176 and 207-219 directly (text layer, not page images). So the bismuth, RGB LED and light-chapter statements below are from the text layer; figures were not inspected.
The 4th edition (announced p.235) has not been found.

Mainstream view in brief (what the claims are measured against)

Magnets: the force between magnetised bodies and between a magnet and iron comes from the magnetic dipole moments (mostly electron spin and orbital moments) in a field B; the energy is U = -m·B and the force is F = grad(m·B). Between two small dipoles on an axis the force falls off as 1/r^4 [memory: high]. The field obeys div B = 0 and curl B = mu0 J (Griffiths ch. 5-6), so field lines close on themselves and need no electric field. No electric field is needed: a magnet works in vacuum, uncharged and unpolarised.
Domain walls (Bloch walls) are nanometre-scale transition layers inside ferromagnets where the magnetisation rotates [memory: high]. They are not the midplane of the external field.
Proton gyromagnetic ratio: gamma_p/2pi = 42.577 478 461(18) MHz/T, relative uncertainty 4.3e-10 [web: https://physics.nist.gov/cgi-bin/cuu/Value?gammapbar].
Photoelectric effect: a threshold frequency nu_0 below which no electrons are emitted at any intensity, and a stopping potential V0 = (h/e)(nu - nu_0), independent of intensity; classical wave theory predicted intensity dependence and a delay; Millikan's 1914 data confirmed Einstein's linear law [web: https://en.wikipedia.org/wiki/Photoelectric_effect].
Bismuth-209 is not stable: it alpha-decays to Tl-205 with half-life about 2.0e19 years [web: https://en.wikipedia.org/wiki/Isotopes_of_bismuth]. Wheeler's own quoted measurement (1.9 +/- 0.2) x 10^19 y with a 3.14 MeV line is the 2003 Orsay result [read: p.103]; I did not independently verify the 2003 Nature citation beyond the Wikipedia page, which cites a 2003 Physics World report.
Light: vacuum electromagnetic waves are transverse; photons carry E = hf; both are among the best-tested parts of physics (photoelectric, Compton, blackbody, single-photon detection) [memory: high].

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

a. Mathematics. There is none to check. The only formula is "D.V. = M1 x M2 / Distance^2" (p.30 [read: deep read, page image]). It is Newton's gravitational form without G, without defined units, and it is given as a correction of Newton. The deep reader's note stands: for magnets no distance law is stated. The dipole-dipole force goes as 1/r^4, not 1/r^2 [memory: high], so the formula, applied to magnets, is wrong as well as undefined. Wheeler's Faraday "missing three-fourths of the force" (pp.30-31) is just inverse-square falloff re-narrated. Internal tensions: he denies attraction but writes that "magnetism is reaching out in space" (p.12); in 2025 he says "magnetic attraction" is "identical to the phenomenon called gravity" [web: https://kenwheeler.substack.com/p/the-codex-of-gravity, as logged by the Investigator], a different identification from the book's "dielectricity".

b. Consistency with experiment. A magnet attracts iron in vacuum, with no charge and no applied electric field; Hall probes, torsion balances and magnetometry measure forces that follow F = grad(m·B). Neutron scattering and Stern-Gerlach experiments measure magnetic moments directly [memory: high]. Magnet force is unchanged by electrostatic shielding (a Faraday cage does not stop a magnet) [memory: high]. A "dielectric" origin would predict electric-field-dependent behaviour in the force, which is not seen. Wheeler offers no measurement showing otherwise. His own demonstrations (steel and magnets resting "perpendicular" or "parallel" to the inertial plane, pp.136-142 [read]) are ordinary static equilibria: a soft iron piece magnetises along the local field and aligns with it; a permanent magnet aligns its moment to minimise energy. Both give the resting orientations he photographs without any new entity.

c. Tested? Not by anyone in the record as a separate hypothesis. It is also not set up so that it can be: "voidance" and "attraction" are given the same observed motion (p.30 diagram: iron accelerates toward the magnet). The Tinmanpower tests (Wheeler's 2026 account; videos not viewed; see claim 6) concern vortex patterns, not this. Ufopolitics' 2015 induction test (Energetic Forum) was presented as "full proof" [web: https://www.energeticforum.com/forum/energetic-forum-discussion/renewable-energy/13672-enlightened-magnetism-the-full-proof-of-ken-wheeler-s-theories]; pages 2-27 of the thread are unread, and a magnet-plus-iron-cylinder drop producing a signal is ordinary Faraday induction (flux through coils changes when the iron changes the magnetic circuit) [memory: high], so it does not discriminate.

d. If true. We would need an electric (dielectric) field source for every permanent magnet and for the force between them; the magnetic moment of the electron and nucleus, the Biot-Savart law and the Lorentz force would be wrong or redundant; every Hall-probe field map would need re-explaining. Nothing in the record says what physical measurement of "dielectric inertia" would detect it.

Verdict (split): - Renaming attraction "voidance" with no new prediction: Not testable as stated. - The physical content (a magnet is an electrified dielectric whose electrical state drives the force; force is not magnetic): Tested, not supported (by decades of ordinary magnetometry and force measurements, and by magnets working uncharged in vacuum; no one has run a test of Wheeler's version because he states no prediction). I note honestly that this is "contradicted by established experiment" rather than "tested as Wheeler's hypothesis".

Related material kept in the record: the conjugate magnetic/dielectric pair and "counterspace" come from Eric P. Dollard (2011-2014), who uses Steinmetz's dielectric-field language [read: research/files/Dollard-dielectricity-and-capacitance.pdf; web: https://ericpdollard.com]. The Dollard text says dielectric lines of force "may form closed loops in electromagnetic radiation" and that magnetic lines close on themselves; this is textbook (div B = 0). What I read of that eight-page text offers no claim that attraction is not magnetic. The "voidance" idea is Wheeler's addition. Counterspace as a term traces to George Adams (projective geometry, "Physical and Ethereal Spaces", hosted by Wheeler on kathodos.com 2015) [read: Investigator note]; I did not read Adams in depth. Both are sources of vocabulary, not of data.

Claim 2. The field is a "magneto-dielectric hyperboloid" (Poincare-disk double vortex) with a "dielectric inertial plane" that is the Bloch wall; field lines are Ether pressure gradients; space is a "posterior attribute of magnetic divergence"

a. Mathematics. The textbook geometry he reproduces is mostly correct (Poincare disk, stereographic projection, complex conic algebra, Phi^-3 = 0.23607) [read: deep read pp.112-127]. Wrong or inconsistent: metric on p.125 lacks the factor 4 for the hyperboloid printed beside it (curvature -4, not -1); the magnet pictures are one-sheet hyperboloids (ruled surface) while the Poincare disk comes from the two-sheet hyperboloid, so the two are different surfaces treated as one; "0 = 1 = infinite inertia" read literally is a contradiction, and no quantity is defined to which it applies. The "disk" is stated to be the dielectric and the "hyperboloid" the magnet, with no map to a measurable quantity. "C squared" as a speed limit is dimensionally wrong (c^2 is in m^2/s^2); it is attributed to Maxwell although Maxwell's result is the wave speed c and the limit on bodies is from relativity [read: deep read, section 6; calc: c^2 = 8.99e16 m^2/s^2]. The standard Poincare-disk geometry is a model of non-Euclidean space; nothing about the field of a magnet requires it. The magnetostatic field of a uniformly magnetised body is fixed by Maxwell's equations and the boundary conditions: for a ferromagnet B = mu0 (H + M) with H = -grad(phi_M), and the field of a dipole is B ~ (3(m·r_hat)r_hat - m)/r^3. That is a definite function; it has no hyperbolic metric in it.

b. Experiment. Measured field maps (Hall probes, NMR magnetometry, ferrofluid and filings patterns) agree with the dipole/current-sheet solutions. Wheeler's claim that iron filings give a "grand error" because they magnetise (p.26) is half right: filings align with the local field direction, which is what is measured; a vortex or spiral in the static field is not supported by anything he shows. A static B field from a stationary magnet has div B = 0 and curl B = 0 outside the magnet (no current there), so there is no circulation to be a vortex. The Bloch wall is a nanometre-scale domain boundary within the material [memory: high], not a plane extending macroscopically outward; to identify the external midplane of the field with it is a category error. The Investigator's lead on Gardi's Hall-sensor work (2020, wiki summary of her video) is relevant: using a homemade Hall sensor, she finds the zero reading at the "inertial plane" (the equator, which is where B_z of a dipole changes sign along a surface and is just a symmetry plane) and that a stronger grade of the same size does not reach further [web: https://wiki.naturalphilosophy.org/index.php?title=On_The_Fractal_Nature_of_Magnets_(and_the_universe).]. That latter result is a measurement that did not support a specific Wheeler claim; it is from a wiki summary of her video (not viewed). Mainstream agrees on the first, and note that a B = 0 or field-reversal ring is a normal property of any dipole's return flux, so finding it does not distinguish theories.

Related work: Tufaile et al. (Univ. Sao Paulo) interpret ferrocell images through hyperbolic geometry ("horocycles", Mobius inversion of the field), peer reviewed, without citing Wheeler [web: https://www.mdpi.com/2410-3896/6/3/30; read: Investigator note and research/files/Tufaile-2024-JOPR-horocycles.pdf, abstract]. This is a different statement: it models how light scatters through a film of field-aligned nanoparticles, not the shape of the magnetic field. It is worth keeping because it shows that hyperbolic geometry does appear in this domain; but it supplies no support for a hyperbolic magnetic field, and the Tufaile guide states plainly that "the light patterns we see at Ferrocell are magneto-optical effects and not just magnetic effects" [read: research/files/Tufaile-FerrocellBookApril2021.pdf]. I am rating the borrowed word, not the optics.

c. Tested? The geometrical identification has no test because it states no measurable prediction ("cannot be measured"). The one measurable corollary (equatorial plane field behaviour; strength vs throw) was measured by Gardi; see above.

d. If true. Static magnetic fields would need a rotational (vortex) structure in a curl-free region and a non-Euclidean metric; Maxwell's equations in the magnetostatic limit would be wrong; space would be derived from fields, not the reverse. The "posterior attribute of magnetic divergence" statement has no operational content I can identify. His Newton's-first-law objection (p.113) is definitional: "rest" and "straight line" are frame-relative, and the law defines inertial frames [read: deep read section 5].

Verdict: Not testable as stated (the Poincare/hyperboloid/Ether-pressure language, the 0 = 1 = infinite equalities, and space-as-attribute). Split: the sub-claim that the external field has vortex/spiral structure and that the inertial plane is the Bloch wall is Tested, not supported by field mapping and magnetostatic theory (curl B = 0 outside the magnet; Bloch walls are internal). The Newton critique: Historically superseded but instructive? No; I rate it Not testable as stated (a relabelling of frames; no measurable difference).

Claim 3. Golden-ratio geometry: gamma = 42.4923 = 180/phi^3, max precession 68.7539 deg, dielectric:magnetic = 3.23606:1, golden angle 137.5077 deg with an 85 deg sector

a. Mathematics. I confirm the deep reader's findings: every number is constructed from phi = 1.618034 (137.5077 = 360/phi^2; 68.7539 = 180/phi^2; 42.4923 = 180/phi^3; 85 = 360/phi^3 = 84.98 rounded; 3.23606 = 2 phi) [read: deep read pp.43-53, section 1]. The arithmetic of golden-ratio identities is correct (phi^3 = 2 phi + 1). The physics is not: MHz/T is a frequency per tesla and degrees an angle; a number can coincide only in an arbitrary unit choice. "omega = gamma e H" counts e twice and is dimensionally wrong (gamma already includes e/2m) [read: deep read §4e]. Internal inconsistency: 42.5 x phi = 68.7539 (really 42.4923 x phi); 42.4932 vs 42.4923; 21.23606 vs 21.246; 360/85 = 4.23606 (really 4.2353); "85 : phi : 1" scale conflicts across pages; 137.5077 = "Phi" and 68.7539 = "1" give ratio 2 not phi.

b. Experiment. The proton gamma_p/2pi = 42.577 478 461(18) MHz/T [web: NIST above]. Wheeler's replacement is 0.2000% lower: 42.577478 - 42.4923 = 0.0852 MHz/T [calc]. For an MRI at 1.5 T that is a 63.866 MHz Larmor frequency against Wheeler's 63.738 MHz, an offset of 128 kHz; at 3 T, 256 kHz [calc]. NMR spectrometers resolve parts per billion and calibrate against the accepted value daily [memory: high], so the claim that 42.4923 is "true" and the published value "incorrectly published" is contradicted by every NMR and MRI instrument. (Wheeler writes "42.5781", which matches neither CODATA 42.5775 nor the shielded-proton value 42.5764; [memory: high].) Larmor precession is the precession of nuclear spins in an external field and its frequency does not depend on the tilt angle; for a bulk ferromagnet the moment is electronic, not nuclear (Fe-56 has zero nuclear spin) [memory: high]. A "precession angle" for a magnet at rest is not a quantity in the theory. His gyroscope remark (precession "is attributed to gravity, however this is incorrect", p.46) conflicts with the standard result Omega = tau/L = m g r / (I omega_s) and the zero precession when pivoted at the centre of mass [memory: high].

Golden-angle field line spiral (137.5 deg, "sunflower phyllotaxis"): the field of an axially symmetric magnet has no azimuthal component, its field lines are planar closed curves in meridian planes, so there is no spiral to measure [calc: the cylinder current-sheet model has B_phi = 0 by symmetry; memory: high]. The p.50-53 figures are sunflower photos, computer-generated tori, moire patterns and Belousov-Zhabotinsky scroll waves from Winfree [read: deep read §6]. None is a magnet measurement. The Investigator found that the 2nd edition (1 July 2014, 110 pp) already contained 42.4923 and 137.5077 [read: research/files/kathodos-magnetism-20140731.pdf, contents list as logged], consistent with the numbers being constructed first. Investigator also found a 2025 Substack statement that "the plane of inertia is the dielectric 45 degrees from magnetic flux and 90 degrees from gyromagnetic precession" [web: https://kenwheeler.substack.com/p/...]; no measurement behind it, and the 2014 sections do not give that angle.

c. Tested? Only by implication: the proton value is tested every day, and Wheeler's value fails. The 137.5 spiral and 3.236:1 ratio have no test described because no method of measuring "dielectric:magnetic" is given. Dave and Mehta (Optik 2024) and Gardi and Snyder (preprints.org 2024, "Finding the Right Angles of a Ferrocell"; blocked) are independent ferrocell angle studies still to read; I cannot state what they find.

d. If true. The most precise measurements in physics (proton NMR, magnetic moment ratios, the SI definition chain for the tesla used in some practical standards) would need a 0.2% revision; all NMR/MRI would be off. I find nothing else in physics that needs phi.

Verdict (split): - Proton gamma = 180/phi^3 and "published value incorrect": Refuted by calculation (and by routine instrument use). The "degrees = MHz/T" identification is Internally inconsistent (unit conflation). - Golden-angle 137.5 deg field-line spiral and 3.236:1 dielectric:magnetic ratio in real magnets: Not testable as stated for the ratio (no way to measure "dielectric"); Refuted by calculation for a spiral in the static field of an axisymmetric magnet (B_phi = 0 by symmetry). - phi appearing in biology (phyllotaxis) is real mathematics [memory: high] and is not in dispute; it does not transfer to magnets.

Claim 4. "Magnetic & Dielectric Field Formula" (7-28-2014): fixed field boundaries at 0.23606, 0.381966, 1, 1.23606 times the magnet radius

a. Mathematics. The numbers reproduce as phi identities (C2 = 1 + phi^-3 = 1.23607; C1 = phi^-2 = 0.381966; C4 = phi^-3 = 0.23607) [read: deep read pp.149-156]. Errors: "C2 + C3 = Phi cubed" fails in either unit choice (2.236 or 5.854); "C1+C2 = Phi cubed" holds only if C1 = 1; "85/360 = Phi cubed" inverted; "1/Phi^3 = Phi cubed" missing a minus sign; C2 labelled "Phi 1.618033" against the rule's 1.236; two different "1" units for angles on different pages. No hypotrochoid parameters or equations are given.

b. Experiment (my check). The claim is testable and I tested it against standard magnetostatics. I modelled a uniformly magnetised cylinder (radius R, top face at z = 0) as a surface current sheet on its side wall, integrated Biot-Savart numerically, and found where B_z changes sign on a plane at height h above the face [calc; reproduces the deep reader's table to 0.01 R]:

L/D	h (in R)	radius of B_z = 0
1	0.05	1.14 R
1	0.10	1.26 R
1	0.25	1.56 R
1	0.50	1.97 R
1.618	0.05	1.26 R
1.618	0.10	1.44 R
1.618	0.25	1.84 R
1.618	0.50	2.36 R

So the return-flux ring passes through 1.236 R only at one height (about 0.1 R above a 1:1 magnet, about 0.05 R above a 1:phi magnet) and moves outward steeply with height. A fixed boundary at 1.236 R is not a property of the magnet; it depends on where you look. The standard model has no feature at 0.382 R or 0.236 R: B_z varies smoothly across the face [calc; deep reader's table]. The edge at r = R (C3) is a real feature, as expected. The ferrocell images he cites (pp.150-151, 168-176) are optical: Tufaile et al. show pattern lines are perpendicular to field lines, can cross, and colours "are strongly linked to the colors of the light sources" [read: research/files/Tufaile-FerrocellBookApril2021.pdf]; the number of circles in a ring rosette is set by the number of light sources (Ferrocell USA sells a "ring of 36 RGBW LED's" [web: https://www.ferrocell.us/]; the circle-count link to LED number is my inference from the geometry, [memory: medium], not stated in the sources I read).

Caveat: Wheeler's formula says "mean field" and does not say which component or height, so a literal reading of my table may be unfair. Even so, no height is named, and the numbers cannot be read off the model at all heights.

c. Tested? Not by Wheeler: no Hall scan or scale bar appears on pp.149-156. The Gardi Hall tests (above) concern strength vs throw, and found zero at the equator; they do not test these fractions. No replication exists.

d. If true. A scale-free golden ratio in the mean field of every magnet of aspect 1 to phi, cube or sphere would contradict the magnetostatic solution above; you would need a mechanism that selects phi.

Verdict: Refuted by calculation in the strong form (fixed fractions independent of height, on the standard model). Never properly tested against a gaussmeter scan by anyone (a cheap test exists; see below), so the weaker form "for a specific plane/height" is Plausible, untested only in the trivial sense that a ring crosses 1.236 R at some height.

Claim 5. Mainstream physics is wrong in its foundations: no electrons, photons, warped space; relativity and QM false; Newton's first law wrong; c^2 limit; Z-axis radial dielectric component of light; photoelectric effect "corrected"

a. Mathematics. The text has no derivation. What is stated as quantitative is wrong or undefined: "c squared" as speed limit (units); "Light through a dielectric medium is the square root of the inverse of the change of capacitance" (p.212, fragment; not a usable formula); the p.35 field-name/unit assignment is crossed (Phi in Wb called "dielectric field"; Psi in C called "magnetic field") [read: deep read §9]. The photoelectric section replaces E = hf by "Replace E with energy of the dielectric, h Planck and f frequency remain the same" (p.217 [read]), i.e. the same equation with a different name for the symbol, so the formula is identical to Einstein's. Wheeler's own pages concede the experimental facts: "the energy of the charge did not depend on the intensity" (p.219 [read]), and "shorter wavelengths = more power = more energy" (p.219 [read]).

b. Experiment. - Photoelectric effect: his model (more "dielectric pulses" at shorter wavelength) can be written to give energy rising with frequency, but gives no threshold and no relation between stopping voltage and frequency; the observed behaviour is a threshold: sodium (work function about 2.3 eV [memory: medium, check]) emits for 400 nm (3.10 eV photon, V0 ~ 0.82 V [calc]) and does not for 700 nm (1.77 eV [calc]) at any intensity. That is the whole content of the effect, and he gives no equivalent. He also states "NOTHING ... shot off the surface material, and certainly no electrons" (p.217): electrons from a photocathode are detected directly (charge, mass in deflection, individual counts) [memory: high]. - "Light is transverse and has an extra radial dielectric Z-component": in vacuum, the plane wave is transverse; a longitudinal component appears for guided and tightly focused beams, which is mainstream optics (orbital-angular-momentum and structured light), none requiring a separate "dielectric carrier" [memory: high]. The coaxial analogy (Heaviside's cable has a TEM mode) is a property of a guided structure, not of light in free space. - Photons, relativity: each is confirmed by many experiments (single-photon detectors, GPS timing, muon lifetime, particle accelerators) [memory: high]. Wheeler cites none and responds with abuse of Einstein, Feynman and "the cult of Quantum" (pp.11, 29, 116 [read]); that is rhetoric, not evidence, but it is part of the record. - Newton's first law: definitional, as above. - Related: Tesla's rejection of curved space is quoted (pp.104, 181 [read]). Tesla's view is a historical position; the 1919 and later light-bending and gravitational-wave measurements fit general relativity [memory: high]. Not checked against Tesla's source here. - Investigator: the p.228 caption claims twisted-light (OAM) communication "could not be possible" without the Z-axis dielectric component [read: deep read]. OAM beams are Laguerre-Gauss solutions of the ordinary paraxial wave equation [memory: high]; the press articles Wheeler reprints do not mention his model.

c. Tested? Each mainstream point has been tested extensively (and by the standards of the field replicated many times); Wheeler's alternative has not been tested because it makes no distinct prediction that I can find.

d. If true. Compton scattering, photoelectric thresholds, blackbody, photon counting and relativistic kinematics would all need alternative explanations that reproduce the same numbers; Wheeler gives none.

Verdict: Refuted by calculation/accepted-law consistency for specific statements (c^2 units; photoelectric threshold), and Tested, not supported for "no photons/electrons" (the alternative has no test and the mainstream has many). The "Z-axis dielectric component" is Not testable as stated (no measurable difference defined). Splitting: the Ether restoration as a metaphysical thesis is Not testable as stated.

Claim 6. The model is confirmed by demonstrations and promises technology

Sub-items, each separately.

6a. Ferrocell patterns (hypotrochoids, "bowl" depressions). Photographs are by Snyder and Vanderelli, who sell/supply the device [read: pp.168-176]. The makers and the Tufaile group describe the patterns as magneto-optical diffraction by chains/needles of nanoparticles [read: Tufaile guide; web: https://www.mdpi.com/2410-3896/4/2/35]. The patterns are real; the interpretation is Wheeler's. Vanderelli's own paper says magnetic anisotropy "in agreement with Maxwell's Equations" [read: deep read, p.4] and the Ferrocell comes from a 2006 patent application (US20070241745A1) [web: https://patents.google.com/patent/US20070241745A1].

6b. RGB LED "discovery" (18 June 2014). Wheeler says white light lies along the equator and red/green along the poles, which he attributes to wavelength-dependent "dielectric capacitance" [read: pp.133-134, 174]. Mainstream: the ferrofluid absorbs blue more strongly and diffraction depends on wavelength, particle concentration, carrier and field [read: Tufaile guide, "Ferrocell's ferrofluid solution can mainly absorb colors closer to blue"; "the concentration of nanoparticles or the type of liquid carrier may favor some wavelengths"]. A ring of discrete RGB dies at different positions would naturally give coloured, position-dependent patterns. Wheeler himself notes "Exposure times of the camera must also be considered" (p.135) and states the unfiltered light is "intensely white" (p.133). No spectral measurement, repeated exposure series or control with a non-magnet appears. The "third-party confirmation" by Kris Lang is one photograph caption (p.136); no further record of Lang found by the Investigator. I do not call it fabricated; it is one documented photograph, no second source found. Gardi (sympathetic) tested "continuous lighting around a Ferrocell" in her 2020 video and calls it a claim she "politely debunks" [web: https://www.youtube.com/watch?v=B8rl1PXybMw, description as logged]; her result is not given in the notes I have. Also: the edition comparison shows the 3rd edition dates this discovery before the 2nd edition's date yet the 2nd edition's text layer lacks it [read: Investigator note]; I did not check page images.

6c. "Programmed" bismuth disk (pp.97-111, which I read). The claim: a disk cast with "applied charge and field in casting" lets a 3 mm N45 cube float 5 inches below a 1-inch N50 lifter cube, "much lower levitation" with a standard disk [read: p.99]. No casting procedure, field strength, voltage, mass, thickness, repeats or control photos beyond one pair appear. Mainstream: diamagnetic stabilisation of a magnet in a field gradient (Earnshaw's theorem evaded by a diamagnet) is standard; stability depends on bismuth's anisotropic diamagnetic susceptibility, grain orientation, purity, disk thickness and shape, and exact lifter geometry [memory: high]. Wheeler's own p.100 says convex castings were chosen "to conform with the field geometry". So "higher" is plausible as an uncontrolled geometry/crystal-orientation effect. The same chapter contains wrong statements that bear on reliability: Bi-209 "stable" and a "spurious" half-life (p.99, 102, 109) vs measured alpha decay [web above]; bismuth has "the highest dielectric permittivity" (p.106-108) (it is a metal; a metal screens fields, permittivity is not defined this way [memory: high]); "bismuth exhibits overunity ... breaks the rule of conservation" (p.103): the Q-value of Bi-209 -> Tl-205 + alpha is positive (his own quoted line 3.14 MeV is the alpha energy released), meaning alpha decay is allowed and energy is conserved [read: p.103; memory: high]; a "golden ratio seat" (p.97); hopper crystals as "anti-cubes" produced by dielectric inertia (pp.104-110), whereas hoppering is standard crystal growth (faster growth at edges/corners than face centres) [memory: high]. Video 73 "M-D-Programmed Bismuth" (2014-08-17, about 7,300 views) is the primary record; not viewed by the Investigator or me [web: https://www.youtube.com/watch?v=LeQ057E8v5w, metadata only].

6d. Objects "resting at the gyromagnetic precession angle" (pp.136-142): static equilibria of magnets and steel, with the equator marked by a green line drawn by Wheeler. No angle is measured. A sphere magnet resting on a ring magnet settles at the equilibrium orientation of the two dipoles, set by geometry and magnetisation; a "42.5 degree" value is asserted. Bismuth disk "perpetually" settling at the same place relative to a 2 x 1 inch cylinder is expected for a diamagnet in a field gradient settling to a stable position [memory: high]; no measurement is shown.

6e. Vortex videos / Tinmanpower. Wheeler's account (2026 post): the Australian YouTuber tested whether "every side of a magnet has a vortex magnetic field" and "proved ... it's 100% accurate", the same with an electromagnet [web: https://www.youtube.com/post/Ugkx3fsEFLIdNufEn1aDXQhzi6sxlb4ZbHur]. The videos were not viewed; this is Wheeler's account only (single source, which I report as his characterisation, with the stated stake that he promotes the work). What "vortex" means in the observation (hydrogen bubbles over a pole, the book p.226) is a flow of electrolysis bubbles: a magnet in an electrolyte with current produces a magnetohydrodynamic (Lorentz force) swirl [memory: medium]; I did not find a source for the specific setup, so this is a mainstream reading I flag as low-to-medium confidence.

6f. Technology promises. "100+ terabyte drive" data written between platters in vacuum, vortex communication, tissue regeneration (pp.61, 72, 226-228 [read: digest/deep]): no device, data, patent (none found on Google Patents by the Investigator) or demonstration exists in the record. The "permanent battery in two dimensions" and the terawatts that "vanish" through grounding (pp.33, 77 [read]) are not claims with numbers; a permanent magnet stores no extractable energy in its static field in the sense of a battery (the force does no net work over a closed cycle) [memory: high].

c. Tested? Ferrocell patterns: yes, extensively, by others, with a different interpretation and no mention of Wheeler. Programmed bismuth: no controlled test or replication on record. RGB LED: one photograph by a third party; no spectral data. Hall-sensor test by Gardi: partly sympathetic third party, one claim not supported (strength vs throw).

d. If true. The ferrocell would need to be a detector of "dielectric inertia" that the makers' optics do not need; the bismuth effect would need a casting dependence on electrical state during solidification, a measurable and testable effect (and one that could simply be a crystal-structure or shape effect).

Verdict: ferrocell/RGB/bismuth as confirmation of the model: Never properly tested (not dismissed without testing; no neutral, controlled test of Wheeler's specific predictions has been published, and he gives no apparatus detail to test it). The mainstream optical account of the ferrocell is established (Established for diffraction/absorption by field-aligned chains, in the Tufaile series; peer reviewed, 2017-2025). Technology promises: Not testable as stated. The programmed-bismuth levitation-height effect as an observed fact: Never properly tested; its explanation by "dielectric inertia" has no supporting evidence.

Investigator's findings, weighed (kept whatever the verdict)

2nd edition (1 July 2014, 110 pp, research/files/kathodos-magnetism-20140731.pdf): the phi numbers and Bloch-wall identification came first; bismuth, ferrocell, RGB LED and the 7-28-2014 formula were added for the 3rd edition (text-layer search only). Bears on claims 3, 4, 6: the demonstrations were added after the theory, as illustration. A caution: RGB discovery is dated 18 June 2014, before the 2nd edition's date, yet absent from it; page images not checked.
Substack 2024-2026: "attraction is gravity" (2025) departs from "dielectric voidance" (2014); no new measurements. The theory is evolving by relabelling.
Gardi Hall-sensor test: a measured, instrumented third-party result; same 1.44 V maximum for 5 mm to 2 in N42 magnets; zero at the equator; grade does not extend throw. The measurement supports the mainstream view that equatorial B_z = 0 is a symmetry property, and does not support one Wheeler claim. Source is a wiki summary of her video, not viewed.
Tufaile et al.: peer-reviewed ferrocell optics including a hyperbolic-geometry description of images (horocycles), no Wheeler citation; the strongest independent link to his Poincare-disk vocabulary, but as optics of the film. Stake: Vanderelli sells the Ferrocell.
Ufopolitics/Energetic Forum 2015 thread: induction test with fixed magnet and dropped iron; pages 2-27 unread, so the forum's objections and results are unknown to me.
Dollard/Adams/Krafft ("The Ether and Its Vortices", [memory: medium]): sources of vocabulary (conjugate dielectric, counterspace, vortex atom). These are historical positions of Ether-era electrical engineering; Steinmetz's dielectric field language is legitimate period usage that modern textbooks fold into the electric field D and E [memory: high].
Absence-of-evidence: the 4th edition (announced Sept 2014, "scheduled" 2018, still unreleased) has not been found; competing explanations in sites.md (moved to video/Substack; the 480/580-page version; abandoned). I cannot decide between them; the Amazon page count and the 67 MB kathodos file (2015-10-14) would. No patent on any vortex technology; no peer-reviewed publication by Wheeler found.
Not reached by the Investigator: Kentucky business registry, Scribd/Academia copies, trademark records, Peaceful Science thread, Gardi and Snyder 2024.

Corrections to Step 1

claims.md claim 3 says the proton value is "said to be 'incorrectly published' (pp.46)". Correct as a report, but Wheeler writes 42.5781, which matches neither CODATA 42.5775 (42.577 478 461(18)) nor the shielded-proton 42.5764; facts.json attributes "CODATA 42.5775 MHz/T [memory: high]". I confirm 42.577 478 461(18) MHz/T [web: NIST above]; the memory tag can be retagged [web].
facts.json summary says "'85 deg' is 360/phi^3 = 84.98 rounded", correct. It says the Poincare "metric lacks a factor 4", which I accept; but "the standard Poincare-disk ... algebra he reproduces is correct" should add that the one-sheet/two-sheet hyperboloid conflation is not correct (deep read §3.1 item 7); the facts.json mainstream_conflict row omits it.
facts.json says the book "states" bismuth is stable; the deep-read claim list reports "digest p.99". I read pp.97-111 directly: the book says Bi has "a half life of 20 billion billion years" and that this "conclusion is spurious" (pp.98, 99, 102, 103, 109) and cites the Orsay result (1.9 +/- 0.2) x 10^19 y [read: p.103]. So the book both quotes the measured decay and denies it. claims.md should say "claims bismuth is stable while quoting its measured half-life".
claims.md says pp.97-111 "is digest-only"; it is now read in the text layer by me (not page images). The bismuth "programming" is given in one sentence ("applied charge and field in casting", p.99) with no method; the question the triage posed ("what was done to program the disk") is answered: nothing is described.
claims.md says a deep-read passage "refers to an earlier pp.97-111 section ... not in deep/". Confirmed: the deep file has four sections (112-127, 220-236, 43-53, 1-42 and 149-156); there is none for 97-111. The "Corrections" line at the end of the pp.43-53 section also mentions it.
facts.json lists "p.136 [digest]" for Kris Lang; I read it: the caption reads "Independent third party confirmation by Kris Lang of the RGB LED discovery. Photo courtesy of Kris Lang of Kirkland Lake, Ontario Canada." One photograph, as stated.
facts.json states the deep read pp.149-156 table for B_z crossing; I reproduced it (differences up to 0.01-0.02 R from grid). One entry (L/D 1.618, h = 0.25) is 1.84 R here vs 1.83 R there; immaterial.
Claim 6 in claims.md bundles ferrocell, bismuth, LED and technology. For review they behave differently (above); the arbiter should keep them apart.

What would settle it (cheap tests, from the claims' own terms)

Hall-probe scan of B_z across a disc or cube magnet (L/D 1 to 1.618) at heights 0.05, 0.1, 0.25, 0.5 R: does a zero ring stay at 1.236 R? The standard model says it moves (table above). Look for any feature at 0.236 R or 0.382 R.
Ferrocell with ring lights of different LED counts and with a single white LED: do the colour bands and circle counts follow the light sources?
Programmed vs ordinary bismuth: same mass, thickness, shape, lifter geometry, several castings, with and without Wheeler's casting step; blinded height measurement; also check crystal orientation.
Any proton NMR instrument: measure the resonance in a known field; 42.4923 would be off by 0.2%.

Source tags and web checks done

Web calls used: 3 (NIST CODATA, Bi isotopes, photoelectric page). [memory: medium] items not rechecked: sodium work function (about 2.3 eV; result depends on it), the 2003 Orsay/Nature citation of Bi-209, hopper-crystal growth mechanism and bismuth diamagnetic anisotropy (high), electrolysis-bubble swirl interpretation (medium-low), Krafft authorship of the vortex book (medium).