Distinti Robert: Arbiter (Step 3)

Inputs: facts.json, claims.md, physicist.md, mathematician.md, contrarian.md. Source tags: [read], [web] URL, [memory] confidence, [calc] recomputed by me.

Source gap (repeated from Step 1)

All ten PDFs have a text layer, but the derivations the claims lean on are NOT in the folder: ni.pdf (the numerical search that produced New Induction), ni_newman.pdf, ind_jackson.pdf, homopolar.pdf, displacement_dilemma.pdf (the "2J" derivation), Maxwells_dilemma.pdf, ron.pdf, the V1/V2 papers and the promised "Ethereal Mechanics" series. The Paradox 2 raw data (scope captures, videos) and any independent replication are absent. Long derivation chapters (nm.pdf chapters 4-9, nia1.pdf, neThesis.pdf) were only sampled by the reviewers. Where this document says "no independent test is documented", that is absence of evidence in the folder plus failed web searches (the web search quota ran out for all reviewers and for me), not a literature survey. The 2J claim and the antenna-pattern claim in nia1.pdf rest on general physics ([memory]), not on the author's argument as read.

Disagreements I resolved by recomputation or inspection

Web checks of remaining medium/low [memory] facts that verdicts depend on

• Flux rule vs sliding contacts, Lorentz force resolves homopolar generators: [web: https://en.wikipedia.org/wiki/Faraday_paradox]. Confirmed (Faraday's law holds when applied with the correct moving boundary; v x B accounts for the emf). The Feynman II sec. 17-2 page itself stayed unreachable (403 for earlier reviewers).
• Electron size: Penning-trap observation suggests radius limit 1e-22 m, and electron scattering shows no structure [web: https://en.wikipedia.org/wiki/Electron]. Retag medium to [web]. The contrarian/mathematician "<1e-18 m" is consistent (weaker bound).
• Relativistic kinetic energy of electrons up to ~0.987c (0.5-15 MeV), agreement with relativistic, not Newtonian, formula; measurement error about 10% [web: https://en.wikipedia.org/wiki/Bertozzi_experiment]. This still rules out "E independent of motion / no kinetic energy" decisively, but note the 10% calorimetry error, which the physicist did not mention.
• Darwin form and river model: see table above.
• Not checked (search quota gone, no fetch tried): Michelson-Morley/Kennedy-Thorndike/Ives-Stilwell limits on absolute-velocity dependence [memory medium]; PEEC/partial-inductance accuracy [memory medium]; cos(pi/2 cos t)/sin t dipole pattern with on-axis null [memory high; textbook]. Conclusions depending on these are marked.
• No web evidence found or sought for or against independent replication of any Distinti claim: the search tool was exhausted. I make no statement that tests occurred or failed beyond what the folder shows.

Claim-by-claim arbitration

Claim 1: New Induction, F = -K_M Q_S Q_T a_S / r

• Agreement: (a) algebra and units fine; (b) for closed loops it equals Neumann's integral, so the thesis inductance data cannot distinguish it from classical theory (all three); (c) no neutral test of the radial term; (d) a true spherical far-field term would break Lienard-Wiechert/Maxwell far-field transversality.
• Disagreement: physicist and mathematician say the strong form is refuted by calculation; contrarian says "never properly tested" (and "tested, not supported" only if far-field transversality counts as the test).
• Stronger argument: physicist and mathematician. The textbook induction law is already the full a/r form, and the missing piece in the far field is -grad(phi), whose omission is what makes NI "longitudinal"; with the "transverse-only" law no loop would have any inductance (R2). The contrarian's strongest honest point is correct but narrower: closed-loop data cannot discriminate, so a discriminating test is cheap and undone. The mathematician adds that the single-charge angular form is wrong by factor 2 at 90 degrees against exact classical theory (Darwin).
• Tested / never tested / dismissed: Never run as a discriminating test; no record of dismissal without testing (no submission record in folder).
• Settles it: measure the on-axis far field of a short dipole, or equivalently the exact Lienard-Wiechert field for a charge with constant acceleration (calculation already done by mathematician).
• Final verdict (split): weak form (a/r is part of induction, closed-loop emf = Neumann): Established, not new. Strong form (textbook induction lacks it, or the full spherical force is the far-field law): Refuted by calculation (accepted Lienard-Wiechert/Darwin fields; the premise about the classical law is also false). Single-charge angular form: Refuted by calculation. The discriminating radiated-field measurement itself: Never properly tested; not shown dismissed.

Claim 2: New Magnetism

• Agreement: formula (with r-hat) equals Biot-Savart + Lorentz force plus the convective piece of -dA/dt; nothing in the sampled regime differs from classical (all three). "No relative motion" rests on frame dependence of E and B, which relativity already handles. Not Lorentz invariant as written, no covariance claimed.
• Disagreement: contrarian rates the strong form "Tested, not supported / refuted"; physicist and mathematician say "not testable as stated" for the B-obsolete interpretation. I side with the latter: since NM reproduces Lorentz in the cases treated, no experiment distinguishes it, so accelerator data do not test it either; the invariance objection (absolute velocities, contrarian) is a real mathematical problem but an unquantified one.
• Mathematician's extra point: the "third term" was added by hand to match experiment; the section 3 expansion discards an extra +v_S^2 r-hat term without derivation. The co-moving transverse force in section 2.1 is (1 - v^2/c^2) where the exact result is (1/gamma) times static; they agree at low v only.
• Settles it: a calculation of a two-charge beam in the lab frame to O(v^2/c^2) and beyond with NM vs Lorentz; and the Lorentz-force calculation for the homopolar disc with co-rotating magnet.
• Final verdict (split): the pair-force formula as quasi-static law: Established (it is already Lorentz force with A, not new). "B, flux lines, cutting/linking obsolete" and "force without relative motion between charge and field": Not testable as stated (interpretation; frame-dependence of E and B already explains the examples). Any claim that classical theory is wrong because of NM: Internally inconsistent (compares against a truncated classical law).

Claim 3: Classical electromagnetism is anomalous (bundle)

Items kept separate: - 3a Self-inductance of a zero-radius wire: all agree it is an idealisation (log divergence, finite for finite radius). The thesis itself needs a finite "seed" size. Verdict: Historically superseded but instructive (known and resolved; not a paradox). (Mathematician's "Established idealisation" is the same content.) - 3b Corner emf: all agree the claimed classical prediction is an artefact of dropping the convective term; correct classical result is zero. Contrarian's addition is kept: the experiment was never built (nm.pdf App. A p.98 [read]), so this is neither "tested and failed" nor "dismissed": it is an unperformed test of a mis-stated prediction. Verdict: Refuted by calculation (as a classical prediction); the corner transformer itself: Never tested (never built). - 3c curl H = 2J: derivation absent. Mathematician's calculation gives curl B = mu0(J + eps0 dE/dt) with J counted once for a slow point charge; physicist and contrarian decline to rule. Verdict: Not testable as stated here (source missing), with a strong prior and a calculation against it; this rests on [memory]/own calculation, not on the author's argument. - 3d Antenna item: see claim 6. - 3e Homopolar / Paradox 2 (paradox2.pdf): the best item for a fair hearing. The mathematician's Lorentz-force estimate (radial arms over opposite-polarity magnets add; both-N-up cancels; sign reverses with rotation; ~180 uV with assumed B 0.1 T, radius ~40 mm, omega ~1.6 rad/s, none stated in the paper) reproduces the qualitative pattern; the physicist reaches the same qualitative pattern. The author gives no classical estimate and uses a flux-counting argument on a circuit with sliding brushes, where the flux rule is not the right tool [web: https://en.wikipedia.org/wiki/Faraday_paradox]. The strengths the contrarian credits are real: no-magnet control gives no signal, chopper amplifier, reversal with rotation. Weaknesses: no uncertainty analysis, no repeat runs, no rotor speed or field map, magnets conductive and co-rotating, contact/thermoelectric effects only partly addressed (8 uV Seebeck estimate), "one magnet" cases off from half by factors of 2 to 4.5, 3e-12 W output. The estimate uses assumed inputs, so it shows the observation does not need new physics, not that the number is predicted. Verdict: Never properly tested (single-author bench run; no independent replication documented; no one, including the author, computed the Lorentz-force prediction for this geometry; not dismissed by a test, and nothing in the folder shows it was submitted to anyone). "Anomalous / unexplained by classical theory" is not established. - 3f "curl H = J holds only for infinite currents": not valid for magnetostatics (mathematician, standard result [memory high]): Refuted by calculation. - Settles it: compute emf = integral (v x B).dl around the real Paradox 2 circuit with measured B map and rotor speed; repeat with a conductive-segment control, with magnets replaced by non-magnetic conductors, and in an independent lab; obtain displacement_dilemma.pdf and find the double count.

Claim 4: Mass and energy from charge (BMP)

• Agreement: force-balance algebra gives V_t = c independent of r_P (all confirm; printed lines drop powers of r_P but ratios 1:1:2 are right). Electron mass is an identity (r_e is defined as K_M e^2/m_e); proton radius is the same formula inverted (r_e/1836); hydrogen "prediction" is the identity m_p - 7 m_e; the reference value 1.0000 is wrong; fields must travel faster than c (the paper concedes it) contradicting its own retarded-propagation rule; Newtonian KE at v = c; instantaneous forces at v = c far outside the O(v^2/c^2) validity of the force laws (mathematician).
• Also: the magnetic-moment radius in ne_hydrogen.pdf gives 3.866e-13 m, i.e. the reduced Compton wavelength, 137 times the BMP's 2.82e-15 m, and the author attributes the mismatch to a possibly wrong measured moment (physicist; read).
• Disagreement: only on the label for the electron derivation (physicist "not testable as stated", mathematician and contrarian "internally inconsistent/circular") and on the proton figure (resolved: 550, not 5.5e5). I choose "Internally inconsistent" (it claims to derive what it assumes) with the note that no independent r_P prediction exists.
• Tested? Nothing predicted differs from the input constants. The one external check (measured proton radius 8.4e-16 m) fails by ~550 if r_P is a physical size; the author calls the model "hypothetical", which weakens but does not remove the problem, since the claim reports a "radius". Electron radius limit 1e-22 m [web] vs r_P = r_e: also conflicting.
• Strongest honest case (contrarian, kept): electromagnetic mass (Abraham, Lorentz, the 4/3 problem [memory high]) was serious physics; "part of inertia is electromagnetic" is a legitimate idea. The model does not make it work.
• Settles it: any independent prediction of r_P (for instance from spin or magnetic moment), which the paper cannot supply; or recompute the hydrogen mass with the correct atomic target (done: 0.44% off).
• Final verdict (split): Vt = c force balance: Plausible, untested within its own premises only; the premises are inconsistent (retardation, v = c). "Derives electron mass" and "derives E = mc^2": Internally inconsistent (circular: identities of the definitions K_E = K_M c^2 and r_e). Proton radius 1.5e-18 m: Refuted by calculation and measurement (factor ~550, [web] NIST). Hydrogen 1.0034 g/mol: Refuted by calculation (identity; true 1.00783; paper's target wrong). The wider electromagnetic-mass idea: Historically superseded but instructive.

Claim 5: Gravity as induction in a flowing ether

• Agreement: the final time-dilation factor sqrt(1 - 2 K_G M/(r c^2)) and r_S = 2 K_G M/c^2 coincide with GR, but follow from inserting the Newtonian escape speed (the Painleve-Gullstrand "river" form, [web] above) and assuming gravity = inertia (equivalence principle assumed, not derived). Energy treatment E = Mc^2 "regardless of motion" ([read: page image ng.pdf p.14]) contradicts E = gamma m c^2. No field equation, no continuity law; the ether is called a "mathematical abstraction"; the incompressible comparison does not fit v ~ r^-1/2; "two factors" in the dilation text but one in the formula (mathematician).
• Disagreement: physicist rates the E claim "Tested, not supported" (Bertozzi); mathematician "internally inconsistent" (against its own time-dilation); contrarian misreads the formula (corrected above). Both are right on different grounds: the paper's own time dilation implies gamma-like behaviour while its energy implies none, and electron kinetic energy measurements follow gamma m c^2 [web: https://en.wikipedia.org/wiki/Bertozzi_experiment] (10% calorimetry error, but Newtonian is far outside it). Note: the Bertozzi test targets the relativistic energy law, not Distinti's specific model; applying it is an inference.
• Not checked on web: ether-drift and Lorentz-invariance limits (Michelson-Morley, Kennedy-Thorndike) [memory medium]; the author's "Earth rides on the ether" escape is unquantified, so the conflict is stated, not demonstrated.
• Settles it: derive from the model the momentum and energy of a moving BMP (p, E vs V) and compare to gamma m v, gamma m c^2; state a continuity equation for the ether and a prediction that differs from GR (none exists now).
• Final verdict (split): Static-clock factor and Schwarzschild radius as numbers: Established (GR results; reproduced by construction, so no independent support). Ether/induction mechanism of gravity, collapse beyond c, FTL work-around: Not testable as stated. E = Mc^2 independent of velocity: Tested, not supported (relativistic electron energy data) and Internally inconsistent with its own time dilation. Overall: Historically superseded but instructive (Lorentz-ether style, with the river model as the legitimate modern descendant).

Claim 6: Longitudinal waves and engineering payoff

• Agreement: for closed circuits NI = Neumann's integral, so inductance accuracy (mean absolute error 1.9%, std 1.3% without plane; 4.1%, 3.3% with plane, 32 traces; fitted slopes 0.987, 0.939 [read]) cannot separate NE from classical theory; no accuracy comparison with PEEC/field solvers in the sampled pages; "10x faster" claim only compared on processing time against partial inductance; longitudinal components exist in the near field (1/r^2, 1/r^3) but far-field radiation is transverse; the end-on dipole result (-6 dB end-fire, ARRL pattern -9 dB cited, [read: nia1.pdf pp.34-35]) conflicts with the classical null on axis.
• Disagreement: far-field verdict. Physicist/mathematician: refuted by calculation (the omitted -grad(phi) term cancels the longitudinal part in the far zone; Gauss' law in empty space). Contrarian: never properly tested because nobody measured far-field longitudinal polarisation. Reconciling: the claim contradicts accepted laws applied correctly, so "refuted by calculation" is the right label, but the contrarian is right that no neutral end-on measurement is on record in the folder and the ARRL figure is not a clean test (ground, feedline, finite wire fill nulls, physicist). That is a gap in the evidence for Distinti, and also no evidence for it. I note the dipole pattern cos(pi/2 cos t)/sin t and on-axis null are [memory high], not re-checked.
• Settles it: anechoic-chamber measurement of a thin-wire dipole's end-fire field at tens to 100 wavelengths vs. the 1/r longitudinal prediction (-6 dB relative to broadside); and a same-trace comparison of NE vs PEEC/FastHenry vs LCR meter.
• Final verdict (split): Engineering inductance model (about 2-5% on traces in author's tests): Plausible, untested independently (it is Neumann's integral; accuracy measured only by the author; speed advantage never properly tested). Not "Established" as a new method; the physics behind it is Established classical theory. Longitudinal far-field radiation: Refuted by calculation as a propagating free-space far field; the direct measurement Never properly tested (not dismissed). Near-field longitudinal terms: Established (already classical).

Corrected claim list (use this in later steps)

1. New Induction: the force on a target from a source's acceleration is claimed to be F = -K_M Q_S Q_T a_S / r (all of a_S, along r too), which Distinti presents as the cure for the "transverse-only" classical law. It is the a/r part of the textbook -dA/dt; the strong claim (textbook is transverse-only; spherical law governs the far field) is wrong.
2. New Magnetism: the pair force with r-hat (unit vectors required) equals Lorentz force with Biot-Savart plus the convective piece of -dA/dt; with New Induction the whole law is the textbook quasi-static Lorentz force on a point charge (exact symbolic identity [calc], QE revision). Its extra term is postulated (to impose direction-independent BMP time dilation and a known induction effect), not derived, and the theory measures charge velocity against an explicit absolute (ether) frame (nm.pdf pp.35-37), a premise that modern light-speed anisotropy limits (about 1e-17, round 2 [web]) leave only as an undetectable-by-construction frame. The claim that B and flux lines are obsolete, and that force exists without relative motion, is interpretation, not new prediction.
3. Classical EM anomalies (author's assertions): thin-wire self-inductance (idealisation), corner emf (the "classical prediction" omits a term; the proposed transformer was never built), curl H = 2J (source absent; contradicted by a standard calculation), antenna patterns (see 6), the 26 vs 22 AWG loop table (classical theory already predicts gauge dependence; see QE revision), homopolar/Paradox 2 (a single bench run of about 100-200 microvolts, unreplicated, no Lorentz-force prediction computed by the author).
4. Mass from charge: two massless like charges orbiting at speed c balance the three forces for any radius; M = K_M Q^2/r_P with r_P = classical electron radius gives m_e by the definition of r_e, so nothing is derived; proton "radius" 1.5e-18 m is 550 times below the measured 8.4e-16 m; hydrogen 1.0034 g/mol is m_p - 7 m_e, not the true 1.00783.
5. Gravity as ether induction: infall speed sqrt(2 K_G M/r) substituted into a derived dilation factor reproduces the GR static-clock factor sqrt(1 - 2 K_G M/(r c^2)) and r_S = 2 K_G M/c^2 by construction; the model's energy law E = Mc^2 for all velocities contradicts relativity and its own time dilation.
6. Longitudinal EM waves and engineering payoff: the NI integral (the classical Neumann integral) predicts trace inductance to about 2-5% against the author's own measurements, but the with-plane thesis values are corrected readings (per-inductor offset, raw data not published), the nia1 two-plane test is about 10% low once a table typo is fixed, and the 97% plane coefficient is undefended; the far-field longitudinal radiation and improved end-on dipole pattern conflict with accepted far-field theory (the -6 dB end-on figure is the author's vector-potential term with the scalar-potential term omitted, which classical theory cancels; round 2).

Verdict summary

Tested / never tested / dismissed (contrarian distinction kept visible)

• Tested and supported (author, narrow): trace inductance vs LCR meter, 1.9% (no plane, low-frequency readings) and 4.1% (with plane, corrected readings; see QE revision); nia1 ch. 15 direct coupling (123 vs 125 nH), single plane (40.2 vs 41.5-42.3), parasitic loop (74.6 vs 73.5), trace inductance with one plane (165 vs 169 nH), rectangular loops (1.3% and 5.7%). Supports Neumann's integral plus the image method, not the NE interpretation. Not supported: two planes (23.9 nH correct vs 26.6-27 measured, about 10% low). No independent replication in the folder.
• Tested and failed (by measurement or accepted theory, not by a test of NE itself): proton radius (550 times off the measured value), hydrogen mass (0.44% off, wrong target), E independent of motion (contradicts relativistic electron-energy data).
• Never properly tested: Paradox 2 homopolar result; corner-effect transformer (never built); radial longitudinal far field of an accelerated charge; NI vs Darwin on single-charge angular dependence.
• Dismissed without testing: no document in the folder shows submission to a journal or neutral lab, nor a rejection. nm.pdf is said to be condensed from a patent-office submission [read, per Step 1]. I cannot state suppression or dismissal; unknown.
• Applying the contrarian's equal-rigour rule to the researcher: wrong reference value (1.0000), identity presented as derivation, factor ~550 on proton size, claim of a measured-looking corner effect that was never measured, the Paradox 2 write-up without a classical estimate, bold generalisation from one bench run to the end of B.

What would settle it

1. Paradox 2: compute emf = integral (E + v x B).dl for the actual geometry (rotor speed, B map, brush radii), then rerun with a conductive non-magnetic control, shielded leads, uncertainty analysis, independent lab.
2. Far-field longitudinal component: anechoic-chamber end-fire measurement of a thin dipole (about 100 wavelengths); compare with -6 dB (NE) vs deep null (Maxwell).
3. Corner transformer (nm.pdf App. A): build it; the classical prediction is zero, NE (by its own account) also predicts none in steady state, so the test must be set up to separate them using non-steady currents.
4. NI vs exact Lienard-Wiechert/Darwin for a single accelerated charge: already computed (mathematician R3); a particle-beam check would test it.
5. BMP: obtain an independent prediction for r_P (spin, magnetic moment) or drop the claim; reconcile the factor-137 radius mismatch between ne.pdf and ne_hydrogen.pdf.
6. Moving-BMP energy and momentum: derive E(V), p(V) and compare with gamma m c^2 and gamma m v.
7. Same-trace PCB inductance: NE vs PEEC/field solver vs LCR meter, with run times.
8. Obtain displacement_dilemma.pdf, ni.pdf, homopolar.pdf and find the factor 2 and the homopolar derivation (try distinti.com/docs, Internet Archive).

Open questions

• Is the Paradox 2 signal fully accounted for by v x B plus contact/thermoelectric effects once the real geometry and speeds are used? (Calculation absent from both the author and the reviewers' sourced data.)
• Does a fast adaptive-integration Neumann code with a seed-size regulariser really beat partial-inductance tools in speed at equal accuracy?
• Did anyone outside the author review or attempt the experiments? Unknown; not documented in the folder.
• What is the "second" dilation factor (reduced light speed) in ng.pdf: redundant or missing?
• How does the model cope with quantum spin and g = 2 at all? It does not claim to.
• Why does Distinti's NM form match the Lorentz force so well? The reviewers' answer (it is -dA/dt plus v x curl A) leaves open whether his numerical search (ni.pdf, absent) rediscovered known electrodynamics, which would be a fair and instructive way to read the project.

Corrections to earlier steps

• Step 1 and contrarian: ng.pdf factor does contain the square root.
• Step 1 and physicist: proton radius discrepancy is ~550, not 5.5e5.
• Step 1 and physicist: corner formula is sqrt(2) K_M I V_S / h (one power of V_S).
• Step 1 and claims.md: the corner experiment was never built; Paradox 2 "no net flux linked or cut" is the author's assertion; NM expressions need r-hat.
• Physicist (minor): Bertozzi has about 10% measurement error [web]; still decisive against zero kinetic energy.

QE revision (MINOR; deep reads in deep/)

Inputs added: deep/neThesis-9619da.md (pp.43-79), deep/nia1-8f67b3.md (pp.42-91, and in round 2 pp.28-41), deep/nm-3bcad1.md (pp.19-42 and 53-73), deep/anomolies-d00cc0.md (pp.4-16). Tags as above. The three reviewers' headline verdicts survive; no verdict changes category, but several are narrowed and two QE-flagged statements are corrected. Round 2 (below) adds the nia1 pp.28-41 deep read and the web checks that were requested in the first round; the verdict table is updated where those change wording.

Source gap, updated

nm.pdf chapters 4-5, 7-9, neThesis.pdf pp.43-79, nia1.pdf pp.42-91 and anomolies.pdf pp.4-16 have now been read in depth (the earlier "only sampled" statement no longer applies to them). The dipole derivation behind the -6 dB end-on claim (nia1.pdf pp.28-41) has since been deep-read (round 2, below), so the claim 6 far-field verdict no longer rests on [memory] far-field theory alone. The missing papers (ni.pdf, displacement_dilemma.pdf, homopolar.pdf) are still absent.

New evidence and what it does to each claim

Claim 6 (inductance engineering): verdict unchanged, evidence weaker than it looked. - Thesis data chain [read: neThesis pp.66-70, 72-74; recomputed by the deep reader from all 36 table rows]. Statistics reproduce (no plane 1.91% / 1.27%, slope 0.9874; one plane 4.14% / 3.31%, slope 0.9393; two planes 1.52%, n = 4). But: - No-plane "measured" values are low-frequency readings (about 200 Hz to 1 kHz) taken where the author's no-plane and with-plane curves meet. The frequency is chosen per inductor and not tabulated. - With-plane values (Tables 8-5, 8-6) are the raw 2 MHz reading plus a per-inductor "correction factor" (no-plane reading at the intersection minus no-plane reading at 2 MHz). No raw readings and no factors are published. The factor is not fitted to the model, but it assumes the same stray offset in nH with and without a plane, which the author himself doubts (p.73, parasitic capacitance). - Only Exp 11/12 can be checked from the chart: raw with-plane about 415 nH, correction about +60 nH, corrected 476 nH, computed 431 nH. The correction turns a -3.9% error into +9.4%. For that inductor the raw data fit better than the processed data. So the 4.1% figure cannot be reproduced or bounded from the thesis; it could move either way. - The no-plane drop from about 1100 to 1040 nH between 1 kHz and 2 MHz looks like skin/proximity current redistribution [memory, high], which a uniform-current model cannot represent. The thesis's explanation for no plane effect below about 1 kHz ("a certain M-field intensity required to knock electrons loose") is not the classical one (skin depth, far larger than 34 micron copper at 1 kHz) [memory, high]. - Speed claim: the abstract says "at least an order of magnitude", the body says "nearly an order of magnitude", and Fig 9-2 shows about 70x at the longest bar (PI about 3550 s, NI about 50 s), but only processing times are compared, both codes are the author's, and no inductance values from PI are reported. The accuracy half of the abstract is never compared with PI or a field solver. The zero-seed formula (eq. 6-3) is the classical Neumann/GMD volume formula, so equal accuracy to a correct classical code is expected, not shown. - nia1 ch. 15 and s.8 [read; independent Neumann recomputation by the deep reader, 8 values matched to 3-6 digits]. Measured vs computed: rectangular loops 1.49e-7 and 1.56e-7 H vs 1.578e-7 H (5.7% and 1.3%, at 500 kHz and 310.8 kHz); direct coupling 125.25 vs 123 nH; single plane 41.5-42.3 vs 40.2 nH; outer parasitic loop 73.45 vs 74.62 nH; trace inductance with one plane 169 vs 165 nH; all at 6 MHz. Every computed number is an ordinary Neumann mutual inductance, so these tests cannot separate NE from textbook theory. - Correction to the first QE note on the two-plane case. The earlier QE said "27.5 nH is a defensible Cesaro average of the author's own table (27.9 by my averaging)". The deep read shows the table has a wrong entry: the 8h row prints 40.0 nH, the formula gives 37.99 nH (all other rows reproduce). The error propagates (3.56 nH constant offset); the converged sum with the 97% coefficient is 23.90 nH [calc, 200 image pairs]. So the two-plane "27.5 vs 27 nH, 1.9%" is really 23.9 vs 26.6 nH, about 10% low. The "Guestimation" series averaging is a legitimate technique and was unnecessary (the series converges in 50-100 terms); the problem is the input typo. - The 97% plane reflectivity is not derived in nia1 (deferred to "another book in this series"); no frequency or thickness dependence. Fitting to the author's own data gives 0.945-0.954 (one plane) and 0.921-0.927 (two planes), which is the signature of frequency/thickness dependence (35 micron copper at 6 MHz has skin depth about 26.6 micron [calc]), not a single constant. QE's earlier wording "fitted constant" is an inference; the text never says it was fitted. It is the only free parameter in the plane predictions and is amplified as eta^n in the image series. - Uncertainty is almost never stated in nia1 (no repeats, no error bars, no scope or meter accuracy, no frequency for the nanoinductance meter); the deep reader's honest tolerance is at least 2-5%, which makes the 1.6-4.9% matches uninformative and the 10% two-plane miss marginally significant. Table 2 compares only 6 of 16 measurements; rows 14 and 15 (computed by the deep reader) are 3.9% and 9.8% low. - Verdict, claim 6, inductance engineering: Plausible, untested independently. Narrowed: author-only data; with-plane thesis values are processed, not raw; the author's own two-plane test is about 10% low; the 97% coefficient is a single undefended constant. The physics behind the model remains Established classical theory (Neumann integral plus method of images with an efficiency factor). - Verdict, claim 6, far-field longitudinal waves: unchanged (Refuted by calculation; direct measurement never done). The dipole derivation was deep-read in round 2 (see below).

Claim 2 (New Magnetism): verdicts unchanged; three points settled. - "Added by hand": confirmed and sharpened [read: nm pp.19-28; SymPy by the deep reader]. Route 1 (BMP time dilation in all directions) imposes V_t^2 = C^2 - V^2, which is the relativistic result, and shows only that some extra term is needed; -(v_T . r-hat) v_S, -(V . r-hat) v_S, or deleting a term, satisfy it equally. Route 2 ("reverse engineering" New Induction) is analogy. The author's final justification is the known effect that moving a source wire induces an emf (p.27). Better wording: "postulated to satisfy an imposed time-dilation constraint and a known induction effect, not uniquely derived". Using relativity's dilation factor as an input and then citing agreement with relativity (pp.21, 28) is circular. - The author's remark that the term "seems to have been omitted from the traditional models" (p.27) is wrong. NM + NI equals the textbook quasi-static Lorentz force with A = K_M Q v / R exactly (symbolic residual zero, [calc]): the term is the part of -dA/dt due to source motion. So claim 2 amounts to the textbook law plus an absolute frame. - Absolute frame [read: nm pp.35-37]. The author states that charge velocity is "always measured relative to the ether" and that Gaussian currents "allow one to measure ethereal properties such as velocity, density and acceleration". This is an explicit preferred-frame theory, not just an inferred one. It is in tension with the claimed consistency with Michelson-Morley and relativity (a measurable absolute velocity contradicts the relativity postulate). It makes NM testable in principle (lab-velocity modulation of a Gaussian current's magnetic effects), but no size or experiment is given and it is unknown whether the ether comoves with the Earth. The premise that the magnetic attraction between co-moving electrons exists for every observer assumes the conclusion; standard theory splits E and B frame-dependently. The Michelson-Morley/Kennedy-Thorndike limits were checked in round 2 (see below). - Stationary/Gaussian current [read: nm pp.29-43]. The stationary-current algebra is correct (V_w and V_O cancel because Q_p = -Q_e) and is the textbook statement that a neutral wire's current is Galilean invariant; the author's "same to all observers" contradicts the O(v^2/c^2) net charge a current-carrying wire shows in other frames [memory, high]. The resistive-wire Gaussian current (Eq. 5-1 to 5-4) comes from integrating step 5 to Q = eps A E, which equals eps A [E(l) - E(0)] = 0 for the constant field the author assumes (he concedes Poisson gives zero charge density, p.38, then dismisses Poisson as "only for potential fields", which is not standard). Internally inconsistent. Its own ratio I_G/I_NG is about 6.5e-22 for 1 V over 1 m of copper [calc], negligible even on its terms. The doped-semiconductor example is wrong (doped material is neutral). - Charged-disk emf (p.35) and concentric-coil Exp. 2 (pp.32-33): predictions or thought experiments with no data. Classical theory (Rowland convection current plus Faraday induction) predicts the same emf, about 1.2e-12 V per pickup turn for a 10 cm disk at 7e-8 C and angular acceleration 1e3 rad/s^2 [calc], so neither discriminates. Exp. 2's "no net emf" ignores the Tolman-Stewart inertial current [memory, high]. - Beam coherency (ch. 7, pp.53-54): the formula F = (Q^2/d^2)(K_E - K_M V^2) equals (Q^2 K_E/d^2)(1 - V^2/c^2) = the standard 1/gamma^2 factor [calc]; a relabelling that attributes to magnetic cancellation what relativity attributes to time dilation. As written (V relative to an unspecified "stationary reference point") it gives different forces in different inertial frames. Nothing was measured; the premise that beams do not spread is overstated (real beams spread from space charge). At 10 keV the cancellation is only 3.8%. - Edge-current magnets and MagnaView film (ch. 8, pp.55-67) [read]: qualitative photographs only (bright edge lines, 3-magnet film arrangement, steel-ball and test-disk edge seeking). The film interpretation is circular (Fig 8-6) and the author lists optical artefacts (pp.61-62). No numbers, repeats or controls. "Magnet strength in amps" is the textbook surface current K = M x n. - Homopolar derivation (ch. 9, pp.68-73) [read]: the author performed no homopolar experiment on these pages; the four-mode Faraday-disk results are quoted, uncited. The algebra is correct, but the cancellation of the magnet velocity V_M is built in (the lattice charges are assigned V_M and the parallel-motion term is bilinear), and the result dV_K = -(K_M Z Q_e / r^2)(V_e V_C) dL is v x B from Biot-Savart: textbook, relabelled. No emf per rpm is computed. The author concedes back-torque and eddy currents are not treated. - Verdicts for claim 2 are unchanged and extended (see table): formula Established (extra term postulated); interpretation Not testable as stated; absolute-frame premise Not testable as stated, with an internal tension; resistive-wire derivation Internally inconsistent; beam formula Established (relabelling of 1/gamma^2).

Claim 3 (classical anomalies): one new item, one item sharpened. - 3g, 26 vs 22 AWG loop table [read: anomolies pp.6-7, page image p.7]: circle 2253 / 2055 nH, square 2144 / 1950 nH (48 in perimeter; 26 / 22 AWG). The author concludes that "the classical model for intrinsic inductance is incorrect" because wire thickness matters. The only classical statement he attacks is the internal term mu/(8 pi) per metre, which is indeed gauge independent at low frequency [memory, high]; he overlooks that the total loop inductance, mu0 R [ln(8R/a) - 2] plus the internal term, already depends on wire radius a. Deep-reader calculation [calc] with R = 0.194 m: 1754 nH (26 AWG) and 1641 nH (22 AWG), difference about 113 nH against the measured 198 nH (circle) and 194 nH (square). Classical theory gets the sign and about 57% of the size. The same offset of about 500 nH above the classical absolute values in both gauges is of the size of leads, gap and fixture; the paper gives no leads, gap, instrument, frequency, repeats or uncertainty, and quotes the data second-hand ("found in ni.pdf"). The residual of about 85 nH in the gauge difference is unexplained (lead inductance changing with gauge, skin effect, dimensions) but not evidence against the classical model. Verdict: Refuted by calculation as an anomaly (the stated classical prediction of gauge independence is wrong). Residual: Never properly tested. - 3a, self-inductance of a thin loop: the author's note that Jackson's L = mu0 a [ln(8a/b) - 7/4] "is actually the mutual inductance between two parallel loops" is asserted without derivation (derivation is in the absent ind_jackson.pdf); the formula is the standard uniform-current loop result including the internal term [memory: high physics, medium for the equation number]. The $50 per hour challenge was a recruiting device (anomolies p.5), which the author says himself. - Other anomolies.pdf sections (Faraday and point charges, reciprocity, flux anomaly, Maxwell vs Kirchhoff, pp.8-16): no data, argument only, and each misstates the classical prediction in effect (reciprocity M12 = M21 follows from the Neumann formula; the Maxwell-Kirchhoff point is a known non-conservative-field bookkeeping issue that the paper concedes). Verdicts unchanged. Page 16 prints curl B = (mu/eps) dE/dt, dimensionally wrong; likely a typo.

Claim 4 (BMP) and Claim 5: small additions. ne_hydrogen.pdf's appendix cites particleadventure.org for r_e about 1e-18 m and r_p about 1e-15 m, the reverse of the model's radii (QE, [read]); paradox2.pdf admits its scope captures are labelled 100 mV/div while the real scale is 100 uV/div (QE, [read: paradox2 line 439]). No verdict changes.

Changed verdicts, summary

Corrections to earlier steps (added)

• qe.md section 3b: "27.9 nH by Cesaro averaging of the author's table" is superseded; the table has a typo at 8h and the correct converged sum is 23.9 nH.
• qe.md section 1(c) is right that nia1 contains measured tables; the deep read lists them above.
• facts.json "10x faster": the thesis text says "nearly an order of magnitude" in the body; Fig 9-2 shows about 70x at the longest length, on processing time only.

Round 2 (second-round QE fixes)

New deep read: deep/nia1-8f67b3.md, section "Deep read (QE round 2): nia1.pdf chapter 6 Antennae Modeling, pp.28-41" (text layer plus page images pp.30-37, with SymPy/NumPy checks by the deep reader [calc]). The other four deep-read files had no new material in this round. Web tool budget ran out partway; items that could not be verified are marked [memory].

nia1 pp.28-41 deep read: claim 6, end-on -6 dB

• What was read: the dipole derivation, Eq. 6-1 to 6-4, the general-pattern code (Code Fragment 1, pp.39-40) and the comparison figure (Fig 6-6, pp.37-38) [read: nia1 pp.28-41; page images pp.30-37].
• Does -6 dB follow from the author's integral? Yes. Broadside |VK| = K_M C^2 I0/(f d) (Eq. 6-3) and end-on |VK| = K_M C^2 I0/(2 f d) (Eq. 6-4); the ratio 1/2 = -6.02 dB follows exactly [read: pp.32, 35; calc]. Only slip: end-on phase should be sin, not cos; magnitude unaffected. Retardation is applied consistently for the magnitude (first-order path differences end-on are kept; broadside ones are second order) [calc].
• What the kernel is: the author's New Induction sum (K_M = 1e-7 = mu0/4pi) is the classical retarded vector potential A integrated along the receiving wire, i.e. -dA/dt [read: pp.31, 39; calc]. Classical theory adds -grad(phi) from the charge the same current deposits on the arms (continuity). On axis at 100 wavelengths |-jwA| = |-grad phi| = 0.4712 V/m per amp and the total is 0.0015 (0.3% residual, falling as 1/r^2) [calc, deep reader, author's own current model]. The author drops phi on the grounds that electric effects are "inverse square" (p.28-29); that holds for a static Coulomb field, not for the retarded potential of oscillating charge.
• Restated verdict basis: the claim 6 far-field verdict, Refuted by calculation, now rests on this deep read of the derivation plus the calculation above, not on [memory] far-field theory. The earlier statement that the derivation was unread no longer applies. The classical dipole pattern cos(pi/2 cos t)/sin t itself is still [memory high] textbook; the deep reader reproduced it numerically.
• Does the verdict stay or change? It stays, with firmer support. "Never properly tested" still applies to the direct measurement (nobody measured on-axis, free-space, far-field polarisation); the folder has no such measurement [read: pp.28-41].
• ARRL -9 dB narrowed: the classical pattern evaluated along the wire's azimuth at 25 degrees elevation is -9.19 dB, and the author's ratio 1:2.828 is -9.03 dB. The author's own ground trace is taken "25 degrees from the horizontal" [read: p.38; calc]. So the ARRL figure plausibly agrees with classical theory, not NE (-6 dB). Red (classical, 0 degrees, free space), blue (ARRL, ground, angle unstated) and green (NE, 0 degrees, no ground) are not like-for-like [read: page image p.37; calc]. The elevation link is an inference (confidence medium); the web check of the ARRL figure failed. This also corrects the earlier ARRL cite: use pp.33, 36 (numbers) and p.38 (source), not pp.34-35.
• Other findings: the printed peak-search code (p.40) uses dp = lambda/12 as a time and, as printed, returns +1.3 dB, not -6 dB; a real phase sweep gives 0.50000 [calc], so the plotted trace used a different setting (reproducibility defect, not physics). "Low frequencies travel farther" does not follow: 1/f is antenna size at fixed current, with 1/d at every frequency [read: p.32; calc]. "Radiation is purely magnetic" (pp.28, 41) is inconsistent with computing an emf from -dA/dt; B = curl A on the axis is zero by symmetry [memory high]. No measurement is reported in chapter 6.

Web checks requested in round 1

1. Ether-drift limits (applied to claim 2, absolute-frame premise). - Michelson-Morley 1887: expected fringe shift 0.4, observed at most 0.02 and on average well under 0.01; velocity "probably less than one-sixth" of Earth's orbital speed [web: https://en.wikipedia.org/wiki/Michelson%E2%80%93Morley_experiment]. Modern optical-resonator experiments confirm no aether wind "at the 10^-17 level" (2009) [same URL]. - Kennedy-Thorndike 1932 gave 10 +/- 10 km/s and bounded the RMS velocity-dependence parameter at about 1e-2; later bounds: 1e-5 (Hils and Hall 1990), 1e-7 (Wolf et al. 2003-04), 1e-8 (Tobar et al. 2009), 7e-12 (Gurzadyan and Margaryan 2018) [web: https://en.wikipedia.org/wiki/Kennedy%E2%80%93Thorndike_experiment]. The page is a secondary source and I did not open the primary papers. - Application: Earth moves at about 3e4 m/s, v/c about 1e-4, v^2/c^2 about 1e-8 [calc]. Any ether frame that affects light propagation or clock rates at order v^2/c^2 is excluded by many orders of magnitude. Distinti's NM is a force law on charges, not a light-speed measurement, so these limits do not directly test it; connecting them is my inference. Verdict for the absolute-frame premise stays Not testable as stated, but the tension is firmer: a preferred frame that changes magnetic effects measurably with lab velocity would also have to leave light speed unchanged to 1e-17, and an ether comoving with the Earth is the only escape, which also makes the "ethereal velocity" measurement of nm pp.35-37 null.
2. PEEC accuracy (applied to claim 6). - Could not be verified. The Wikipedia PEEC page describes a full-wave version with time delay and quasi-static reductions for electrically small structures, but states no accuracy figures or validation studies [web: https://en.wikipedia.org/wiki/Partial_element_equivalent_circuit]. An attempt to open the IEEE record of the original paper returned nothing. - What stays [memory, medium]: PEEC and FastHenry-class partial-inductance or field solvers reproduce measured PCB/package inductance to a few percent when geometry is accurately known, and use the same Neumann partial inductances. Application: this makes Distinti's 2-5% accuracy unremarkable (the same integral), reinforces that his accuracy test cannot separate NE from classical theory, and leaves the "10x faster" claim a processing-time comparison against his own PI code only. The same-trace NE vs PEEC/field solver vs LCR comparison remains the missing test; verdict unchanged.
3. Guala-Valverde and Tramaglia, Am. J. Phys. 70, 1052 (2002) (applied to the homopolar discussion). - What the folder shows: anomolies.pdf p.26 quotes "Guala-Valverde & J.A. Tramaglia" ("Customary physics don't prescribe how the forces produced by the closing wire on the magnet are distributed on its bulk, which make impossible the direct calculation of the relevant torque") and says Jorge's work shows by energy conservation that a back torque must exist when magnet and disk move together, while classical physics does not say where it acts [read: anomolies pp.26]. The Am. J. Phys. reference is the arbiter's/deep reader's earlier note, not on that page. - Web: the Wikipedia Faraday paradox page lists Am. J. Phys. 70(10), 1052-1055 (Oct. 2002) as "Jorge Guala-Valverde, Pedro Mazzoni, Ricardo Achilles, The homopolar motor: A true relativistic engine", described as arguing "only the Lorentz force can explain Faraday's disc" [web: https://en.wikipedia.org/wiki/Faraday_paradox]. So that journal citation lists different co-authors and a different title than the quote in anomolies.pdf; the Tramaglia work was not found and is [memory, low] as a separate paper. I could not open the paper itself, so the back-torque content is unverified. - The same Wikipedia page states that the magnet's own motion is irrelevant to the emf in the usual analysis; the relative motion that matters is disk versus return path [web: same URL]. - Application: if the listed paper is the relevant one, it argues for a Lorentz-force treatment, which is the textbook reading and what the arbiter found Distinti's homopolar algebra reduces to (v x B). The back-torque point does not support NE; it is a classical accounting question (where the reaction acts) that Distinti's chapter concedes it does not treat. The Paradox 2 verdict stays Never properly tested.

Round 2 verdict changes

No verdict changes category. Wording changes, reflected in the verdict table and Corrected claim list above: | Item | Before | After | |---|---|---| | Claim 6, far-field longitudinal | Refuted by calculation, rested on [memory] theory | Same label; now rests on a deep read and checked calculation (scalar-potential term cancels the author's -dA/dt term, 0.3% residual at 100 wavelengths) | | Claim 6, ARRL -9 dB | Cited as conflicting with the classical null | Not like-for-like; matches the classical pattern at 25 degrees elevation (-9.19 dB) [calc], so it is not evidence for NE | | Claim 2, absolute frame | Not testable as stated, tension with relativity | Same label; tension tied to sourced ether-drift limits [web], inference flagged | | Claim 6, inductance engineering | Plausible, untested independently | Unchanged; PEEC accuracy comparison could not be web-verified | | Claim 3, homopolar | Never properly tested | Unchanged; the Guala-Valverde citation is unverified and, per the journal listing, differently authored |

Still unchecked

• Not verified on the web (budget exhausted; see Round 2): PEEC accuracy figures; the content of the Guala-Valverde paper cited by the author (authorship in the folder differs from the journal listing); the ARRL Antenna Book figure conditions. These stay [memory] with the confidence given below.
• No independent replication of any Distinti measurement is documented in the folder or was found; no suppression or dismissal can be stated.