Rankings Primer Arbiter Physicist Mathematician Contrarian QE Claims Sources

Distinti Robert: Science Contrarian review (Step 2)

Reviewer role: Science Contrarian (PhD standard; judges by experiment, not consensus; same rigour applied to the researcher and to his critics).

Source gap (repeated from Step 1)

All ten PDFs have a text layer. Not available: the numerical search paper ni.pdf, ni_newman.pdf, ind_jackson.pdf, homopolar.pdf, displacement_dilemma.pdf, the raw Paradox 2 scope data and videos, and any independent replication. Conclusions below that concern independent testing rest on absence of evidence in the folder plus a failed web check (see "Web checks"), not on positive documentation. Equations were read from pdftotext -layout output of ne.pdf, nm.pdf, ne_hydrogen.pdf, paradox2.pdf and neThesis.pdf; no page images were needed for the passages I checked.

Web checks (budget used: 3 fetches; search tool was out of quota)

Proton rms charge radius = 8.4075(64)e-16 m (CODATA 2022). [web] https://physics.nist.gov/cgi-bin/cuu/Value?rp (Step 1 said ~0.84 fm: confirmed.)
1H relative atomic mass = 1.00782503223 u; standard atomic weight of H = [1.00784, 1.00811]. [web] https://physics.nist.gov/cgi-bin/Compositions/stand_alone.pl?ele=H (Step 1 said 1.00783: confirmed.)
Check failed: web search for independent replication of Distinti's work, for the electron size bound (<1e-18 m), and for Feynman's discussion of the flux-rule exceptions. Quota exhausted; fetching feynmanlectures.caltech.edu returned 403. These stay [memory] and are labelled below. I therefore make no claim that anyone has or has not tested NE beyond what the folder shows.

Corrections to Step 1

claims.md claim 3 and facts.json (nm.pdf) say classical theory predicts corner emfs "that are not observed", implying a negative experiment. [read] nm.pdf Appendix A (p.98) says: "Since it is scientifically accepted that the corners of loops do not have measurable 'hot spots', there is no need to build this experiment." The corner-effect transformer was never built or measured. The "proof of the spherical field" rests on an unperformed test plus an assumed classical prediction. Step 1 hedged ("no measured data seen") but the claims.md wording should be "never built".
facts.json for ne.pdf: BMP derivation. I re-did it. With r = 2 r_P, Coulomb : NM : NI = K_E : K_M Vt^2 : -2 K_M Vt^2 after multiplying by 4 r_P^2/Q^2, so Vt = c is correct algebra (the printed step 2 drops Q^2 and mixes denominators but the ratios are right). Step 1's note that the electron-mass result is an identity is correct and I confirm it: r_e = K_M e^2 / m_e c^0 in these units, since K_E = K_M c^2, so M = K_M e^2/r_e = m_e exactly. [read: ne.pdf pp.26-27] The paper's "more precise" variant (charges e/2, r_P = r_e/4) is the same identity with a free choice of r_P (K_M (e/2)^2 / r_P times 2 / ... gives m_e again at r_P = r_e/4). Not a correction, an addition.
ne_hydrogen: Step 1 wrote 1.0034 vs "1.0000" and flagged the 1.0000 as wrong. Confirmed [web]. Sharpening: the true 1H molar mass is about 1.00783 g/mol, so the model's 1.0034 is 0.44% below the real value, and the paper's claimed 0.34% "error" is measured against an incorrect target. Neither figure is evidence for the model; 12 pair terms with signs and one fitted radius (r_P of proton fixed by inverting the same formula) is a fit.
Step 1 said Paradox 2 "output about 150-180 uV". [read] paradox2.pdf lists 180 uV (both magnets), 20 uV, 90 uV, 80 uV, and the Seebeck section compares 8 uV with "150 uV". Fine; but the amplifier is a 1000x chopper DVM with load 10 kOhm and a 1 Hz low-pass; the "no magnets, no signal" control is real and is the strongest point of that paper (see claim 3).
neThesis: Step 1 noted comparison is only to Neumann-type formulas. I checked: the thesis compares NE with Partial Inductance only on processing time (sec. around line 2600), not on accuracy against an independent solver (no FastHenry/PEEC/commercial field solver accuracy table found by grep). Accuracy figures (1.9% and 4.1% mean absolute error) are NE vs LCR-meter measurements only.

No other factual errors found in Step 1.

Claim 1. New Induction: F = -K_M Q_S Q_T a_S / r, spherical (acts along r too)

a. Math consistent? Internally yes as a postulate; it is simply the Liénard-Wiechert acceleration term with the transverse projection removed. It is not derived from Maxwell. Distinti's own V1 "derivation" was a numerical search (ni.pdf, not in folder), so the law was fitted to induction examples, then declared fundamental. Consequence worth noting: a force along r from acceleration is a longitudinal acceleration field whose Coulomb-like 1/r piece, combined with the real 1/r transverse piece, would change the radiation pattern. For a steady current in a closed loop it must sum to zero net, and the paper shows this only for chosen geometries. b. Consistent with experiment? The transverse form -K_M q [a - (a.r^)r^]/r is the far-field acceleration term; it reproduces measured antenna patterns and radar/radio to high accuracy [memory high]. A radial 1/r force from acceleration would give a longitudinal component of the radiated field at 1/r, which far-field measurement of dipole polarisation (purely transverse in the far zone) does not show [memory medium]. No document shows a measurement isolating the radial term. c. Tested? Only by Distinti, as a modelling check against LCR-meter inductance measurements of PCB traces (thesis; see claim 6). No neutral lab has tested the radial term. Dismissed without testing? No evidence in the folder either way: nothing here shows submission to a journal or a neutral lab (the nm.pdf says it was condensed from a patent-office submission). The corner effect that is the "crucial test" was never built. d. If true: Maxwell's far-field radiation would need a longitudinal 1/r component; Liénard-Wiechert potentials and Lorentz covariance of the field tensor would fail; gauge-invariance structure of QED would need rebuilding.

Strongest honest case: closed-loop inductance only constrains the integral of the force around a loop, and the transverse and spherical forms differ by a gradient-like term that integrates to zero over closed loops. So closed-loop data cannot discriminate; the thesis fit is unsurprising. The only real discriminator is an open-circuit or radiated-field measurement of the longitudinal part. That test is cheap and undone. Case against: the force is ad hoc, no field equations or conservation laws are given for it, and its retardation structure is borrowed from the theory it rejects.

Verdict: Never properly tested (the discriminating measurement was never run; I found no record of a neutral test; not shown dismissed without testing, since no submission/review record is in the folder). The radial 1/r longitudinal radiation prediction, taken as a far-field claim, is Tested, not supported only if the general far-field transversality of antenna radiation is accepted as that test [memory medium; not checked on the web].

Claim 2. New Magnetism replaces Biot-Savart and Lorentz; B is obsolete; force without relative motion

a. Math consistent? The NM force on a target charge has v_S, v_T entering as F = K_M Q_S Q_T/r^2 [(v_T.r)v_S - (v_S.r)v_S - (v_S.v_T) r]. [read] ne.pdf, secrets_of_qvxb.pdf. Two problems. (i) It is not Galilean or Lorentz invariant as stated: it depends on absolute velocities v_S, v_T (the paper argues "no relative motion needed" and an ether frame). Physics that depends on absolute velocities contradicts the null results of Michelson-Morley, Kennedy-Thorndike and Ives-Stilwell [memory medium, unchecked on web]. (ii) The derivation in the_secrets_of_qvxb.pdf replaces B by the field of a point charge and then declares B "moves with its source", giving F = Q_T (v_T - v_S) x B. That step is a postulate, not a derivation; it contradicts the textbook transformation, where E and B mix under boosts. With a velocity-dependent term -(v_S.v_T) r that survives when v_S = v_T the force between two co-moving like charges differs from Coulomb. The BMP uses exactly this term, and NE then needs fields to travel faster than light (stated in ne.pdf p.26). b. Experiments: Lorentz force verified in cyclotrons, mass spectrometers, cathode-ray tubes, Hall effect, and the structure of motors [memory high]. Two parallel moving charges (e.g. co-moving beam) experience a reduction of net repulsion by 1/gamma^2 relative to rest-frame Coulomb in accelerator beams [memory high]; NE has to reproduce that and the docs show no such calculation. c. Tested? Distinti's own homopolar/Paradox 2 measurements (claim 3), the corner transformer never built, magnet-attraction modelling without data in the sampled pages. No independent test found. Dismissed without testing: possibly, but the folder cannot show it. d. If true: Lorentz force and special-relativity-compatible electrodynamics would need replacement; accelerator beam dynamics and particle physics fits would need re-deriving.

Strongest honest case: textbook F = Q(v - v_B) x B really is ambiguous about what v_B means when the source of B is a rotating magnet; Faraday's disk with co-rotating magnet is a legitimately subtle case that many textbooks treat badly. Raising that is a fair point. A learner should note that the correct answer comes from the lab-frame Lorentz force acting on all conducting parts including the magnet's own conduction electrons. Case against: that subtlety is already resolved in mainstream electrodynamics without discarding B; the NM replacement loses invariance and has no demonstrated advantage in measured data.

Verdict: Split. (1) Strong form (NM replaces Lorentz/Biot-Savart, B obsolete): Refuted by calculation only in the sense of conflicting with already-tested relativity and beam data; stated more carefully, Tested, not supported (by decades of accelerator and invariance experiments, though not tests aimed at NM) [memory high for Lorentz tests, medium for the specific ether null results]. (2) The narrow claim "force without relative motion between charge and the B-field source" in the homopolar case: Not testable as stated, because "relative to the B field" is not defined for a static magnetised body and mainstream theory gives the same measurable outputs in the lab frame.

Claim 3. Classical EM is anomalous (self-inductance, corners, antennas, homopolar, curl H = 2J, Paradox 2)

Treat as a bundle; items differ.

a. Math: Self-inductance of an ideal zero-thickness wire does diverge (logarithmically); with finite wire radius it is finite and standard (Neumann with geometric mean distance). That is an idealisation issue, not an inconsistency. [memory high]. The "curl H = 2J" item: [read] anomolies.pdf item 18 asserts it with the derivation pushed to a paper not in the folder (displacement_dilemma.pdf). Unverifiable here; Maxwell's equations are well known to give the correct field of a conduction current plus displacement current, so I treat it as unsupported until the derivation is seen. Corner emf: the "classical" corner calculation applies only the acceleration term to a test charge and omits the other terms of the full field of a steady current (nm.pdf p.78 says so: NM "cancels" it using the extra terms; classical is given no cancelling terms). In standard theory a steady current produces no E-field on a stationary charge, and the cancellation comes from the velocity-field and charge terms [memory medium]. I regard the corner effect as a mis-stated classical prediction. b. Experiment: Maxwell/Neumann inductance formulas are used successfully in EE; antenna patterns of dipoles match cos(pi/2 cos th)/sin th to within measurement [memory high]. The thesis claim that classical methods are "singular or erroneous" is contradicted by routine PEEC/partial-inductance use [memory medium]. c. Tested? Paradox 2 is the only experiment: a home-built disk with two opposed disc magnets, Cu ring, brushes. [read] paradox2.pdf: no-magnet run gives no signal (4.1) and reversal with rotation direction (reported), 1000x chopper preamp, 1 Hz filter, 10 kOhm load. These controls are better than usual for an amateur. Weaknesses [read]: no calculation of the classical prediction for this geometry; magnets are conductive (ceramic/neodymium unspecified) and rotate with the rotor, so Lorentz force on the moving conductors and brush contact at different radii can give a nonzero emf by the standard rule; no uncertainty analysis; unreplicated; data not in folder. A few hundred microvolts at a rotating brush contact is within the range of contact and thermoelectric effects even with the 8 uV Seebeck estimate, which only covers temperature offsets of 20 F. Dismissed without testing? Unknown. Nothing in the folder shows an independent attempt. d. If true: the flux rule's exceptions would be fundamental rather than a bookkeeping issue; Faraday's law in the form emf = -dPhi/dt for moving circuits would need a new form.

Strongest honest case: the flux rule truly fails for some moving-contact geometries (Feynman's own discussion) [memory medium; web check failed], and homopolar machines are still sloppily taught. A tabletop measurement with a chopper amplifier and null controls deserves a calculation of the Lorentz-force prediction rather than a shrug. Equal rigour for Distinti: he never gives that prediction himself, and the item list mixes real subtleties with items that look like idealisation artefacts or a straw man (corners, thin-wire inductance).

Verdict: Split. Homopolar/Paradox 2: Never properly tested (one uncontrolled amateur run; no neutral replication; the mainstream Lorentz-force explanation was not computed for this geometry, by either side). Self-inductance of thin wire, corner emf: Internally inconsistent as the researcher presents them for the classical theory only in the sense of idealisation or an incomplete classical calculation; the corner emf has never been measured (no experiment built). Curl H = 2J: Not testable as stated here (derivation not in the folder; I cannot assess it).

Claim 4. Mass and energy from charge (BMP; electron mass; E = mc^2; proton 1.5e-18 m; H = 1.0034 g/mol)

a. Math: Force balance gives Vt = c for any r_P (confirmed; see correction 2). Effective mass M = K_M Q^2/r_P from equating -Ma with NI is fine as definition. Energy: KE = (1/2)MC^2 plus PE = K_E Q^2/(2 r_P) = (1/2) M C^2 so E = MC^2; consistent, but only because KE was computed with Newtonian (1/2)MV^2 at V = c, where relativity says momentum diverges, and NI-defined M is used as inertial mass for KE without proof. Electron "result" is an identity (correction 2). Proton radius is the same formula inverted (r_P = K_M e^2/m_p = r_e/1836), so it has no predictive content. The hydrogen model has charge-pair terms chosen so that the proton and electron "pairs" add; the match to 1.0034 is a numerical coincidence from a fitted structure, and the target 1.0000 is wrong (correction 3). Fields must propagate "much faster than light" for the orbit to be causal (ne.pdf p.26): contradicts the paper's own rule that field changes propagate at c (retarded). b. Experiment: electron structure is probed to below 1e-18 m with no size found [memory medium; web check failed]; proton charge radius is 8.4075e-16 m [web NIST]; the model's 1.53e-18 m proton radius is ~550 times smaller than measured (8.4075e-16 / 1.5347e-18 = 548). The measured electron has spin 1/2, magnetic moment g ~ 2, which the model omits (paper says it does so). c. Tested? Nothing to test: no new observable is predicted that differs from the input constants. d. If true: classical electron radius would be physical size; QED mass renormalisation, Standard Model Higgs-Yukawa mass origin, hadron structure would be replaced.

Strongest honest case: the old Abraham-Lorentz electromagnetic mass programme was serious physics in 1900-1910 and the "4/3 problem" is real; Distinti's NI inertia is an attempt to avoid it. The idea that part of inertia is electromagnetic self-energy is legitimate. Case against: circularity of the electron result, the wrong hydrogen target, c-speed orbital motion with Newtonian KE, and the proton radius conflict.

Verdict: Refuted by calculation for the proton radius claim (8.4075e-16 m measured vs 1.5347e-18 m model; the factor 548 follows from accepted laws and measurement, no new experiment needed) and for the hydrogen "match" (1.0034 vs true 1.00783 g/mol; the paper's own number is off, and the formula has a free structure). The electron-mass "derivation" is Internally inconsistent in the sense of circular: M = K_M e^2/r_e equals m_e by definition of r_e, so nothing is derived. Historically superseded but instructive: the broader idea of electromagnetic mass (Abraham-Lorentz).

Claim 5. Gravity as induction in a flowing ether

a. Math: ng.pdf (abstract, secs 2-6) gets a time-dilation factor sqrt(1 - V^2/c^2) from the BMP with axial velocity V (V_t' = sqrt(c^2 - V^2)); then replaces V by sqrt(2GM/r) to get 1 - 2GM/(r c^2) (no square root: not even equal to the earlier special-relativistic form it just derived) and a Schwarzschild radius. The result is obtained by substitution of the free-fall speed, which is the standard weak-field heuristic and not a derivation of GR; the paper itself says the ether is a "mathematical abstraction" and the BMP "may not represent an actual physical system". Inconsistent mixing of factors (square root dropped). b. Experiment: Light bending, Shapiro delay, perihelion of Mercury, GPS clock rates, gravitational waves are accepted and are matched by GR [memory high]; the paper provides no calculation for any of them. Ether with absolute motion constrained by Michelson-Morley, Kennedy-Thorndike [memory medium; unchecked]. A medium "consumed" by matter predicting decay and molten core: no quantitative content. c. Tested? No. Not a testable-as-stated model, since no prediction differs from GR with numbers. d. If true: GR's geometric picture and the equivalence principle in current form; the null results on ether drift would be reinterpreted; faster-than-light "work-around" would require rewriting causality.

Strongest honest case: ether-flow analogues of gravity (Painleve-Gullstrand coordinates, "river model" of black holes) are legitimate mathematics reproducing Schwarzschild exactly, so the infall speed sqrt(2GM/r) is not silly. A lesson for the learner: Distinti recovers it as a coordinate heuristic. Case against: he gives no field equation, so nothing beyond the Schwarzschild radius number emerges, and that number is input via the infall speed.

Verdict: Not testable as stated (as presented), with the Schwarzschild/time-dilation "derivation" being Internally inconsistent in its factor (sqrt dropped between the special-relativistic and gravitational versions). As an interpretation of GR (river model) it would be Plausible, untested only if reformulated; Distinti does not do that.

Claim 6. Longitudinal EM waves and engineering payoff (PCB/IC inductance within 5-10%)

a. Math: The thesis integrates NI along conductors with images (97% inductive reflectivity for copper) to give kinetic voltage; for closed loops the NI integral reduces to the Neumann formula, so the same numbers would be obtained from classical theory. Therefore an agreement with an LCR meter cannot distinguish NE from CE (observation from Step 1 stands). Longitudinal wave claim: nia1.pdf derives a dipole pattern with a longitudinal on-axis component and says classical needs "black magic"; the standard half-wave pattern cos(pi/2 cos th)/sin th is not "black magic" and is verified in practice. b. Experiments: thesis data are real engineering measurements: two sets, mean absolute error 1.9% (std 1.3%) and 4.1% (std 3.3%); ground-plane accuracy 10%; the author notes LCR meter residuals of a few pH to <100 pH up to 2 MHz [read: neThesis.pdf]. Near-field longitudinal electric field components do exist in Maxwell theory (near zone of a dipole) [memory high], so a "longitudinal" effect near antennas is not an anomaly; far-field radiation is transverse. c. Tested? Yes, as engineering validation by the author under thesis supervision (Fairfield University). Independent replication: none found in the folder; web replication check failed. No head-to-head against a commercial field solver or PEEC code on accuracy. Speed claim (10x vs partial inductance) is the only comparison. d. If true: for the engineering claim nothing changes (the method may simply be a valid alternative form of Neumann integration). For far-field longitudinal radiation: Maxwell far-field transversality and the Poynting-flux picture would change.

Strongest honest case: this is the best-measured part of the project, was done to a supervised thesis standard, and a fast singularity-free inductance code is useful whatever its theoretical story. Credit where due. Case against: the method does not distinguish NE from CE; the rhetoric ("all classical derivations erroneous") is far beyond the evidence; the longitudinal-radiation claim has no measurement of the far-field polarisation.

Verdict: Split. Engineering inductance accuracy (5-10% on traces): Plausible, untested in the sense of independent replication (measured by the author only; consistent with known physics because it reproduces Neumann-type numbers; not Established). Longitudinal far-field radiation: Never properly tested (no polarisation measurement is reported; mainstream far-field theory and routine antenna practice say transverse [memory high]).

Overall assessment (contrarian)

The part with the best claim to a fair hearing is the unperformed or under-calculated pair: the Paradox 2 disc with its null controls and the never-built corner-effect transformer. Both are cheap and discriminating, and neither has been done or properly calculated by a neutral party. This is "never properly tested", not "tested and failed" and not, on the evidence in this folder, "dismissed without testing".
Applying the same rigour to the researcher: a wrong comparison value (1.0000 for hydrogen), an identity presented as a derivation (electron mass), a proton radius 550 times off measured, a dropped square root, a claim of a corner effect that was never measured, and the bold generalisation from one home experiment to the end of the B field. The mass-from-charge and gravity claims are not rescued by calling them "hypothetical".
Single experiments that would settle most of the testable disagreement: (1) compute and measure the Lorentz-force emf of the Paradox 2 geometry with conductivity of each part (calculation plus a rerun with a conductive-segment control); (2) build the corner transformer in nm.pdf App. A; (3) measure far-field polarisation of a small dipole for any longitudinal component.