Primer: Dunning-Davies and Norman on scalar waves, the quaternion "seventh field" and the aether

Source tags used throughout: [read] from the documents in the project folder; [web] checked online (URL given); [memory] recalled from training, with confidence high/medium/low; [calc] a calculation done for this primer and reproduced in example.py. The final verdicts come from arbiter.md.

Source gap (updated after QE). All six documents now have readable text. The 2024 Hadronic Journal paper "Thoughts on the Existence of an Aether" (29 scanned pages) was read in full from its OCR text, with page images checked for equations. The 3-page 1987 CIA memo "Scalar Waves" was read as page images. The 69-page "Gravity v.2" has now been deep-read across pp.3-5 (the Van Flandern eclipse claim), pp.11-27 (the main paper's "evidence" section), pp.28-51 (the "Poynting vortex" and hadronic-fuel appendices) and pp.52-69 (Thomas Townsend Brown's 1928 British patent GB 300,311 and the Carstoiu / Dunning-Davies note on gravitational vortex fields) [arbiter.md, Round 2]. The end of "Final consequences of an aether" (pp.6-10) was deep-read. The works the authors build on (Jack 2003, Whittaker 1903/1904, Thornhill 1983/1984, Sweet and Bearden 1991, Santilli's own papers) were not available. The mathematics in sections 4-5 comes from the 2020 preprint, which was read in full, with its equations checked against page images [read: page images pp. 8-10]. The hadronic-fuel numbers in section 5b come from the Gravity v.2 appendix [read] and the AME2020 table of atomic masses [web]. The neutron-synthesis evidence comes from the authors' 2017 co-authored experimental paper, which the review fetched online [web]. The Sweet and Schauberger device facts and the lensing point still rest on [memory] or secondary web pages, and are marked that way.

---

1. The idea in one paragraph

Jeremy Dunning-Davies (a retired mathematics and physics lecturer at the University of Hull [read]) and Richard L. Norman argue that modern electromagnetism lost something when Maxwell's equations were rewritten in the vector notation of Heaviside and Gibbs. If you go back to Maxwell's quaternions, as reworked by the mathematician Peter Jack, an extra scalar quantity appears, which they call the "Temporal Field". They say this field supports "scalar waves" that carry no ordinary electric or magnetic field, travel faster than light, and are the real nature of gravity. They connect this to a physical aether (a space-filling medium), to reports of antigravity and over-unity devices, to a non-probabilistic reading of quantum mechanics, to Ruggero Santilli's "hadronic mechanics" (in which the neutron is a proton and an electron bound together), and to a claim that governments already use and hide this science. Later papers add proposals for clean nuclear "fuels" triggered by 1.294 MeV photons. In 2017 the two authors also co-signed an experimental paper, run by a company for which Norman is a scientific advisor, that claims neutrons were made from hydrogen gas in an electric arc. The heart of the formal case is a single short derivation in a 2020 preprint, and that derivation is what this primer teaches you to check for yourself.

---

2. What you need to know first

2.1 Electric and magnetic potentials. Instead of working with the electric field E and magnetic field B directly, you can write them in terms of a scalar potential phi (the "voltage") and a vector potential A. Because div B = 0 always, B can be written as the curl of something, and Faraday's law then lets E be written with phi and A. Most of Maxwell's equations then become wave equations for phi and A. Read: D. J. Griffiths, Introduction to Electrodynamics, chapter 10 ("Potentials and Fields") [memory: high].

2.2 Gauge freedom. Different pairs (phi, A) give exactly the same E and B. You can add the gradient of any function chi to A, provided you also subtract (1/c) times the time derivative of chi from phi. This is called a gauge transformation. Since only E and B (and quantities built from them, such as energy and force) are measured, anything that changes under a gauge transformation is bookkeeping, not physics. To remove the ambiguity one imposes a "gauge condition"; the common choice in radiation problems is the Lorenz condition. Read: Griffiths ch. 10; J. D. Jackson, Classical Electrodynamics, chapter 6 [memory: high]; J. D. Jackson and L. B. Okun, "Historical roots of gauge invariance", Reviews of Modern Physics 73, 663 (2001) [memory: high].

2.3 The wave equation and its sign. The one-dimensional wave equation is ∂²u/∂x² − (1/v²) ∂²u/∂t² = 0. Any function u = f(x − vt) solves it, a shape moving at speed v. The minus sign between the space and time terms is essential. If the sign is plus, ∂²u/∂x² + (1/v²) ∂²u/∂t² = 0, you get Laplace's equation (as in electrostatics), whose solutions do not travel: trying u = exp(i(kx − ωt)) gives ω = ±i v k, so the "wave" grows or decays exponentially in time instead of moving. Equations with the plus sign are called elliptic; the wave equation is hyperbolic. Read: any ODE/PDE text, e.g. M. L. Boas, Mathematical Methods in the Physical Sciences, the chapter on partial differential equations [memory: high].

2.4 Quaternions (introduced from scratch). A quaternion is a number q = a + bi + cj + dk, where a, b, c, d are real and the three units obey i² = j² = k² = ijk = −1. Hamilton invented them in 1843. The useful fact: if you multiply two "pure" quaternions (a = 0), u = u₁i + u₂j + u₃k and w likewise, you get

u w = −(u · w) + (u × w),

a scalar part (minus the dot product) and a vector part (the cross product) in one go. Gibbs and Heaviside later split this product into separate dot and cross products, which became ordinary vector calculus [memory: high]. Quaternions have a "Euclidean" norm: |q|² = a² + b² + c² + d², all plus signs. Spacetime in relativity needs one minus sign (time versus space), which is why quaternion versions of relativistic physics use complex-number coefficients ("biquaternions") [memory: medium]. That point turns out to matter in section 4. Read: J. B. Kuipers, Quaternions and Rotation Sequences (Princeton, 1999), the opening chapters [memory: medium].

2.5 Energy in the electromagnetic field. In Gaussian units (the units the paper uses) the energy stored per unit volume is u = (E² + B²)/(8π), and the energy flow is the Poynting vector S_P = (c/4π) E × B. If E and B are both zero everywhere, there is no electromagnetic energy anywhere. Read: Griffiths ch. 8 [memory: high].

2.6 Special relativity in one sentence. Light in vacuum has the same speed c in every inertial frame, and no signal can be sent faster than c without letting some observer see effect before cause. Read: E. F. Taylor and J. A. Wheeler, Spacetime Physics [memory: high].

2.7 Beta decay and Q-values. In beta-minus (β⁻) decay a neutron inside a nucleus turns into a proton, emitting an electron and an antineutrino (a neutral, nearly massless particle that passes through almost all matter). The nucleus keeps its mass number A (protons plus neutrons) and its atomic number Z (protons) goes up by one. The energy released is the Q-value: Q = (M_parent − M_daughter) c², where M are atomic masses (the electron masses then cancel). Masses are tabulated in atomic mass units u, with 1 u × c² = 931.494 MeV (1 MeV = 1.602 × 10⁻¹³ J). If Q is negative the decay cannot happen on its own. The released energy is shared at random between the electron and the antineutrino, so the electron gets a spread of energies (a "spectrum"), on average roughly a third to a half of Q, and the antineutrino carries the rest away. In "double-beta decay" a nucleus for which a single β⁻ step is forbidden (Q < 0) jumps two steps at once, Z → Z+2, emitting two electrons; it is extremely slow (half-life of Mo-100 about 7 × 10¹⁸ years [memory: high]). Read: K. S. Krane, Introductory Nuclear Physics (Wiley, 1988), the chapter on beta decay [memory: high]; or the nuclear chapter of any modern-physics text such as Serway, Moses and Moyer, Modern Physics [memory: medium].

---

3. The textbook view

Gaussian units, as in the paper. Symbols: E electric field, B magnetic field, ρ charge density, J current density, c speed of light (2.998 × 10⁸ m/s), t time, ∇ the gradient operator (∂/∂x, ∂/∂y, ∂/∂z), ∇² = ∂²/∂x² + ∂²/∂y² + ∂²/∂z² (the Laplacian). A subscript t means ∂/∂t.

Maxwell's equations [memory: high]:

∇ · E = 4πρ ∇ × E = −(1/c) ∂B/∂t ∇ · B = 0 ∇ × B = (1/c) ∂E/∂t + (4π/c) J

Potentials: E = −∇phi − (1/c) ∂A/∂t, B = ∇ × A.

Gauge transformation (chi is any function of x, y, z, t):

A → A + ∇chi, phi → phi − (1/c) ∂chi/∂t. E and B are unchanged.

Lorenz gauge. Choose the potentials so that L ≡ ∇ · A + (1/c) ∂phi/∂t = 0. Then in empty space

∇²phi − (1/c²) ∂²phi/∂t² = 0, ∇²A − (1/c²) ∂²A/∂t² = 0,

wave equations at speed c. Under a gauge change, L shifts by ∇²chi − (1/c²) chi_tt, the d'Alembertian (the relativistic wave operator, written □). So L can always be set to zero; it is a choice, not a field.

Pure gauge. Start from zero potentials (no field) and apply a gauge transformation with chi = S. You get A = ∇S, phi = −(1/c) S_t. Then E = −∇(−S_t/c) − (1/c)∇S_t = 0 and B = ∇ × ∇S = 0, for any S. Such potentials look busy but describe nothing: no force on any charge, no energy. (In quantum mechanics, the Aharonov-Bohm effect shows potentials can matter, but only through gauge-invariant loop integrals; pure-gauge potentials still have no effect [memory: high].)

Two genuinely true points the authors build on. (i) The Poynting vector is defined only up to adding the curl of some vector, because only its divergence enters the energy-conservation law [memory: high; also stated in the CIA memo, read p.1]. That ambiguity does not change the energy density (E² + B²)/8π. (ii) Whittaker (1903) showed that a static 1/r potential can be written mathematically as a superposition of travelling spherical waves [read: preprint, quoting Whittaker]. This is true, and it works for any wave speed V; the integral ∫₀^∞ (2/πμ) sin(μ(Vt − r))/r dμ equals 1/r whatever V is [read; calc]. So it shows a mathematical decomposition, not a physical speed.

The aether after 1905. Special relativity (1905) did not need an aether, and no experiment found one (section 6, claim 3). In a 1920 address at Leiden, "Ether and the Theory of Relativity", Einstein said that in general relativity "space without ether is unthinkable". The 2024 paper quotes this as support for a material aether [read: 2024 paper p.9]. But Einstein meant the gravitational field itself (the geometry of spacetime), and in the same address he said that the idea of motion may not be applied to this ether [memory: high; the paper does not quote that sentence]. A medium with no state of motion has no velocity u, so it cannot drag light or carry a D/Dt term. Thornhill's gas aether, which has a velocity and a temperature, is a different thing. The address was also given after both relativity theories, not before.

Neutrinos have been detected. Cowan and Reines detected reactor antineutrinos in 1956 by inverse beta decay: an antineutrino hits a proton and makes a neutron and a positron. The positron annihilates at once (two 0.511 MeV gammas), and a few microseconds later the neutron is captured by cadmium (a further gamma burst). Requiring this delayed pair of signals is what separates the event from background [memory: high]. Today detectors such as Super-Kamiokande and IceCube record neutrinos routinely [memory: high].

Longitudinal waves do exist in plasmas (oscillations of the charge density along the direction of travel; Langmuir waves) [memory: high; the CIA memo says this too, read p.3]. These have a non-zero E field. They are not what the authors mean.

---

4. Where this researcher departs

The departure happens in three steps, all in the 2020 preprint [read: page images pp. 8-10].

Step 1: the quaternion scalar part becomes a "field". Jack writes the potential as a quaternion A = phi + A₁i + A₂j + A₃k and the derivative as d/dr = (1/c)∂/∂t + i∂/∂x + j∂/∂y + k∂/∂z. Multiplying these with the quaternion rule from 2.4, the product has a vector part (giving the usual E and B) and a scalar part. The paper defines (their eq. 1)

T = −(1/c) ∂phi/∂t + ∇ · A.

Note the sign. The plain quaternion product gives a scalar part with the opposite relative sign to the Lorenz combination [calc]. This is not a typo by the authors; it comes from the quaternion norm being Euclidean (all plus signs). The textbook version needs the minus sign that relativity puts between time and space.

Step 2: "slightly modified Maxwell equations". The paper writes (eqs. 3-6), for empty space:

T acts like a fictitious charge density and current. Take the divergence of (3): the left side is zero (div of a curl), so 0 = (1/c) ∂(∇·E)/∂t + ∇²T. Insert (5):

(1/c²) ∂²T/∂t² + ∇²T = 0.

That is the plus-sign (elliptic) equation of 2.3. The paper's own equations allow no travelling T-waves at any speed [calc, example.py Part C]. In the textbook, the same manipulation gives charge conservation instead.

Step 3: the c√3 speed. The paper imposes (7) ∇·A = −(1/c) ∂phi/∂t (this is the Lorenz condition), so T = 2∇·A. It sets B = 0, so A = ∇S, gets from (3) the relation ∇(∇²S − S_tt/c² + T) = 0, sets an integration constant to zero ("this assumption, because assumption it is" [read]), substitutes T = 2∇²S and obtains

∇²S − (1/(3c²)) ∂²S/∂t² = 0, "a wave speed of c√3" [read: p.10-11].

The paper's derivation uses only equation (3). It never checks equation (5). Section 5 does that check. Then the paper adds Thornhill's aether, in which the total derivative D/Dt = ∂/∂t + u·∇ (u is the aether's velocity) replaces ∂/∂t near matter and reverts to ∂/∂t far away, and concludes that far from mass the waves are "near instantaneous" [read]. If D/Dt becomes ∂/∂t, you simply recover the ordinary equations; no step in the documents produces an infinite speed [arbiter.md].

A nuclear departure: neutrons, "aetherinos" and photon-triggered fuels. The later documents add claims outside electromagnetism.

The number 1.294 MeV is m_n − m_p (1.293 MeV) [calc, example.py Part E], not the threshold of either reaction, which section 5b computes. A lithium reaction chain on Gravity p.36, written Li(3,7) + ñ + a → Li(3,8) → 2 He(2,4) + 2γ, starts with three protons (Z = 3) and ends with four (two helium nuclei), so it does not conserve electric charge [read; arbiter.md].

Where the starting points are fair. The quaternion rewriting of Maxwell is legitimate notation (Maxwell used quaternion expressions in the Treatise [read: as cited by the paper; memory: medium]). The Poynting ambiguity and Whittaker's decomposition are real mathematics. A 19th-century aether was a reasonable mainstream position. The problem is the inference from these to a new physical field and a superluminal wave.

---

5. Worked example: does the c√3 wave satisfy the paper's own equations?

We test the paper's solution against its own equations (3) and (5). Everything is reproduced symbolically in example.py, Part D.

Setup. Take a plane wave moving along x at an unknown speed v:

S = f(x − vt), where f is any smooth function. Write f′, f″, f‴ for its derivatives.

Step 1: potentials. A = ∇S, so A_x = f′ (other components zero). Then B = ∇ × A = 0, as the paper wants.

Step 2: impose the paper's condition (7), ∇·A = −(1/c) ∂phi/∂t. Here ∇·A = f″. Try phi = g(x − vt); then ∂phi/∂t = −v g′. So f″ = (v/c) g′, giving

phi = (c/v) f′.

Step 3: T. T = −(1/c) phi_t + ∇·A = −(1/c)(−v)(c/v) f″ + f″ = 2f″. This matches the paper's T = 2∇·A = 2∇²S. Good.

Step 4: E. E_x = −∂phi/∂x − (1/c) ∂A_x/∂t = −(c/v) f″ + (v/c) f″ = ((v² − c²)/(cv)) f″.

Step 5: equation (3), x-component. ∇ × B = 0, so we need 0 = (1/c) ∂E_x/∂t + ∂T/∂x. - (1/c) ∂E_x/∂t = (1/c) · ((v² − c²)/(cv)) · (−v) f‴ = −((v² − c²)/c²) f‴ - ∂T/∂x = 2f‴ - Sum: (3 − v²/c²) f‴ = 0, so v² = 3c², v = c√3. This is the paper's result.

Step 6: equation (5). We need ∇·E = (1/c) ∂T/∂t. - ∇·E = ∂E_x/∂x = ((v² − c²)/(cv)) f‴ - (1/c) ∂T/∂t = (1/c) · 2 · (−v) f‴ = −(2v/c) f‴ - Setting them equal: (v² − c²)/(cv) = −2v/c → v² − c² = −2v² → v² = c²/3, v = c/√3.

Step 7: put in numbers (c = 2.998 × 10⁸ m/s): - Eq. (3) needs v = 5.193 × 10⁸ m/s. - Eq. (5) needs v = 1.731 × 10⁸ m/s.

They cannot both hold unless f‴ = 0 everywhere, which means no wave at all. At the paper's speed v = c√3, the ratio of the two sides of (5) is ∇·E ÷ ((1/c) T_t) = (2/√3) ÷ (−2√3) = −1/3, not +1 [calc; agrees with arbiter.md].

Step 8: two side-results. - At v = c√3, E_x = (2/√3) f″, not zero. So the c√3 wave is not the "E = B = 0" scalar wave of the CIA memo that the paper says it generalises. - The CIA memo's wave, A = ∇S, phi = −(1/c) S_t, gives E = B = 0 for any S (example.py Part A). Its energy density (E² + B²)/8π is exactly 0 J/m³ at every point, so it carries no electromagnetic energy. The memo's "wave equation" for S is just the Lorenz condition written for these potentials.

What the example teaches. Equation (3) and equation (5) are two different constraints. The paper used one and stopped. When a system has more equations than unknown functions (an "overdetermined" system), you must check every equation, because a solution of one may violate another. Here the deeper reason is Step 2 of section 4: the two equations together force the elliptic equation for T, so no real speed can satisfy both.

Run it yourself:

.venv/bin/python review/Dunning-Davis-J/example.py

Part E of the script also does quick checks used below: the Sweet-Bearden antigravity experiment's 175 μW in and 1 kW out (a power gain of 5.7 × 10⁶) [read; calc], the energy needed to confine an electron inside a neutron (p c ≈ ħc/r ≈ 197-235 MeV for r = 1.0-0.84 fm, against 0.782 MeV available [calc; standard masses, memory: high]), and m_n − m_p = 1.293 MeV.

5b. Second example: do the hadronic "fuels" give net energy?

The Gravity v.2 appendix proposes Zn-70 and Mo-100 as fuels [read: pp.44-48]. A 1.294 MeV photon forces step 1, N(A, Z) → N(A, Z+1); step 2 is an ordinary β⁻ decay N(A, Z+1) → N(A, Z+2). The two electrons are caught on a conducting screen. example.py, Part F, reproduces every number below from AME2020 atomic masses [web: https://www-nds.iaea.org/amdc/ame2020/mass_1.mas20.txt].

Step 1: Q-values (Q = ΔM × 931.494 MeV/u, using the masses in the script):

Step 1 has Q < 0, so it cannot happen on its own; that is why these nuclei are stable in practice. The total is positive. In pure mass-energy bookkeeping the authors are right that energy is released: it is the energy of ordinary double-beta decay.

Step 2: what can actually be collected. In each step the antineutrino takes a large share and escapes. Averaging a standard β spectrum gives the mean electron energy per event [calc, approximate]: - Zn-70: about 0.87 MeV in electrons for each 1.294 MeV photon spent. Ratio 0.67. - Mo-100: about 1.80 MeV. Ratio 1.39.

Step 3: the verdict in numbers. Zn-70 loses energy even if every photon is made with 100% efficiency and every photon triggers the reaction. Mo-100 breaks even only if (efficiency of making 1.294 MeV photons) × (probability a photon triggers step 1) ≥ 1/1.39 = 0.72. In reality a 1.3 MeV photon in a metal almost always loses its energy by Compton scattering off electrons, and wall-plug efficiency into a narrow MeV gamma line is far below 1% [memory: medium]. No measured "resonance" is given in the documents [read]. Hence the arbiter's split verdict: Zn-70 fuel Refuted by calculation; Mo-100 fuel Never properly tested, with the required conditions implausible.

What this example teaches. "Energy is released" (Q > 0) and "net energy can be recovered" are different claims. An energy budget must track where every share goes, including particles you cannot catch, and the energy spent to make the trigger.

---

6. Evidence table

"Tested?" means a neutral, controlled test is on record. None of the six folder documents reports a new measurement. But Gravity v.2 cites a 2017 experimental paper on neutron synthesis that Dunning-Davies and Norman co-signed. It was run by Thunder Energies Corporation, which sells the neutron source and hosts the detector data, and for which Norman is "Scientific Advisor" [read: Gravity p.2; web: AJMP 2017, see section 9]. That paper is a claimant-run test, not a neutral one (row 5e). The second QE round found that the main paper's "evidence" section (Gravity v.2 pp.14-25) contains no measurement by the authors at all: its five items (Laithwaite, Schauberger, Sweet-Bearden, Brown, hadronic fuels) are quotations or anecdotes relayed mainly through T. Valone's Electrogravitics Systems and N. Cook's The Hunt for Zero Point, plus one abstract [read: deep read pp.14-25; arbiter.md Round 2].

---

7. What a learner should take away

Solid. - Maxwell's equations can be written in quaternion form; this is real mathematics and real history. - Potentials have gauge freedom, and the Poynting vector is ambiguous up to a curl. Both are standard results the authors state correctly. - Pure-gauge potentials A = ∇S, phi = −(1/c)S_t exist and give E = B = 0. The CIA memo's algebra is correct. - Whittaker's decomposition of 1/r into travelling waves is valid mathematics.

Wrong, and why the mistake is instructive. - Treating a gauge quantity as a field. T changes when you change gauge while E and B do not, so it cannot be measured. The general lesson: before calling a quantity physical, check whether it is gauge-invariant. - A sign inherited from the mathematics. The quaternion product has Euclidean signature, so its scalar part has the wrong relative sign for relativity. The result is an elliptic equation for T, with no waves at all. The lesson: the choice of algebra is not neutral; it carries a geometry with it. - Checking one equation of an overdetermined system. The c√3 speed satisfies equation (3) and violates equation (5) of the same paper (section 5). Always substitute a proposed solution into every equation. - "Zero field but carries energy." With E = B = 0, the electromagnetic energy is zero. An ambiguity in the energy flux does not create energy density. - Whittaker proves too much. His construction works at any speed V, so it cannot show gravity is superluminal; the GW170817 measurement shows it is not. - A closed loop with no source. The Sweet-Bearden abstract claims "closed battery-less systems with damped positive feedback", and Gravity v.2 quotes a B-2 that coasts on "essentially zero fuel". A gain measured with an outside power supply (175 μW in, 1 kW out) is a claim about an amplifier, and the energy might come from somewhere unmeasured. A self-running loop has no outside supply, so energy conservation alone rules it out. Ask where every joule comes from. - "Energy released" is not "energy recovered". The hadronic fuels do release mass-energy (Q > 0), but the antineutrinos carry much of it away, and the trigger photons cost energy. For Zn-70 the books do not balance even in the ideal case (section 5b). - Relayed claims are not evidence. Every item in Gravity v.2's "evidence for coupling" section (pp.14-25) is a quotation of a quotation; the authors measured nothing there. When a document lists evidence, ask for each item who measured what, with which instrument. - A number that depends on the unit system is not physics. The Brillouin "ε = −1/G = −1.5 × 10⁷" is 1/G in cgs units and 1.5 × 10¹⁰ in SI [calc]. A genuine dielectric constant is a pure number; check units before reading meaning into a match. - Read the whole document, including the appendix. The Carstoiu note reprinted in Gravity v.2 gives vacuum gravitational waves at speed c, which contradicts the superluminal gravity of the main text; a p.13 footnote contradicts pp.18 and 22 the same way. The cited note also has its own error: it says Einstein's equations give gravitational waves at half the speed of light, but general relativity gives c [read: p.63; memory: high]. Even the note's remark that a spin-up of the Earth "may well violate" its stability condition Ω² ≤ 4πγρ (Ω an angular velocity, γ the gravitational constant, ρ a density) is far off: with γ = 6.674 × 10⁻¹¹ m³ kg⁻¹ s⁻² and Earth's mean density ρ ≈ 5515 kg/m³, 4πγρ ≈ 4.6 × 10⁻⁶ s⁻², while Earth's Ω² = (7.29 × 10⁻⁵ s⁻¹)² ≈ 5.3 × 10⁻⁹ s⁻², about 870 times smaller [calc; the deep read and arbiter print 4.6 × 10⁻³ s⁻² and "six orders", an arithmetic slip; the conclusion, no danger, stands]. - Renaming is not discovery. The "aetherino" has exactly the neutrino's properties. A new name with no new measurable difference is not a new particle. - Who ran the test matters, and so do the controls. The one experiment the authors signed was run and published by the company that sells the device, with no background run and readings in step with the electrical discharges. That does not prove it wrong, but it is not a neutral test. A cheap, independent neutron-counter run with arc-off and dummy-load controls would settle it [arbiter.md]. - Read a quotation in context. Einstein's 1920 "ether" has no state of motion; quoting only the first half makes it look like support for a gas aether. In the same way, the earlier review misread the Sweet-Bearden abstract by assuming two figures described one device. They describe two different experiments, so the "inconsistency" the review first reported does not exist.

Genuinely open (or at least unexamined). - The Sweet-Bearden and Schauberger device reports were never put through a neutral, instrumented power measurement; no data are on record either way [arbiter.md]. - Santilli's neutron synthesis has been tested only by interested parties, with weak controls; no independent replication, positive or negative, was found [web checks not done]. Of everything here, this is the cheapest claim to settle. - The Mo-100 fuel is not ruled out by energy bookkeeping alone; it fails only on the (well-founded) expectation that photons almost never trigger the step. No one has measured that probability for this nucleus. - Gravity v.2 is now deep-read throughout its argument (Round 2), but the sources it leans on remain unchecked: Van Flandern's eclipse essay, the NATO AGARD 1990 report (the one quoted sentence, via Cook p.83, is conditional: "if such a counteracting force was of a sufficient magnitude"), and Valone's and Cook's books [read: Gravity; arbiter.md Round 2]. - Quantum foundations are not closed: non-local hidden-variable theories (de Broglie-Bohm) survive Bell tests [memory: high]. The authors' version would still have to explain how superluminal links avoid letting anyone send signals faster than light.

---

8. Check yourself

1. Show that A = ∇S, phi = −(1/c) ∂S/∂t gives E = 0 and B = 0 for any function S. What is the energy density?
2. Under A → A + ∇chi, phi → phi − (1/c) chi_t, how does T = −(1/c) phi_t + ∇·A change? How does the Lorenz combination L = ∇·A + (1/c) phi_t change? Which one has the relativistic wave operator?
3. From the paper's equations (3) and (5), derive the equation T must satisfy. Try T = exp(i(kx − ωt)). What is ω, and why does that mean no waves?
4. Using atomic masses 69.925319 u (Zn-70), 69.926022 u (Ga-70) and 69.924249 u (Ge-70), find the Q-value of each β⁻ step and of the whole chain Zn-70 → Ge-70. If the two electrons together collect on average 0.87 MeV per 1.294 MeV trigger photon, is this a fuel?
5. The Sweet-Bearden antigravity test is quoted as 175 μW in and 1 kW out. (a) What is the power gain? (b) Why can a device that takes input from a power supply have a large gain without breaking any law, while a "closed battery-less system with positive feedback" cannot?

---

9. Further reading

Mainstream - D. J. Griffiths, Introduction to Electrodynamics (any edition from the 3rd), ch. 8 (conservation laws, Poynting vector) and ch. 10 (potentials and gauge) [memory: high]. - J. D. Jackson, Classical Electrodynamics, 3rd ed., ch. 6 [memory: high]. - J. D. Jackson and L. B. Okun, "Historical roots of gauge invariance", Rev. Mod. Phys. 73, 663 (2001) [memory: high]. - J. B. Kuipers, Quaternions and Rotation Sequences (1999), for quaternion algebra [memory: medium]. - E. F. Taylor and J. A. Wheeler, Spacetime Physics, for special relativity [memory: high]. - B. P. Abbott et al., "Gravitational waves and gamma-rays from a binary neutron star merger: GW170817 and GRB 170817A" (2017) [web: https://arxiv.org/abs/1710.05834]. - S. Carlip, "Aberration and the speed of gravity" (2000), the answer to Van Flandern's orbit argument [web: https://arxiv.org/abs/gr-qc/9909087]. - B. Hensen et al., "Loophole-free Bell inequality violation using electron spins separated by 1.3 kilometres", Nature 526, 682 (2015) [web: https://www.nature.com/articles/nature15759]. - E. T. Whittaker, A History of the Theories of Aether and Electricity, for the aether as a historical position [memory: high]. - A. Einstein, "Ether and the Theory of Relativity" (Leiden address, 1920), in Sidelights on Relativity; read the whole address, not only the passage the 2024 paper quotes [memory: high]. - C. L. Cowan, F. Reines et al., "Detection of the Free Neutrino: a Confirmation", Science 124, 103 (1956) [memory: high]. - K. S. Krane, Introductory Nuclear Physics (1988), for beta decay, Q-values and double-beta decay [memory: high]. - AME2020 atomic mass evaluation, table of masses [web: https://www-nds.iaea.org/amdc/ame2020/mass_1.mas20.txt].

The researchers' own works (in the project folder) - J. Dunning-Davies and R. L. Norman, "Deductions from the Quaternion Form of Maxwell's Electromagnetic Equations", ResearchGate preprint, July 2020, DOI 10.13140/RG.2.2.28045.77283 [read]. The derivation checked in section 5. - R. L. Norman and J. Dunning-Davies, "Consequences of the Existence of an Aether" (preprint, 2018) [read]. - R. L. Norman and J. Dunning-Davies, "Gravity" (v.2, 2020, with appendices) [read: front part; deep reads of pp.3-5, 11-27, 28-51 and 52-69]. - J. Dunning-Davies and R. L. Norman, "Thoughts on the Existence of an Aether", Hadronic Journal 47, 153-181 (2024) [read: full OCR text]. - R. Norman, S. Bhalekar, S. Beghella Bartoli, B. Buckley, J. Dunning-Davies, J. Rak and R. M. Santilli, "Experimental Confirmation of the Synthesis of Neutrons and Neutroids from a Hydrogen Gas", American Journal of Modern Physics 6(4-1), 85-104 (2017) [web: http://www.santilli-foundation.org/docs/confirmation-neutron-synthesis-2017.pdf]. The authors' only experimental paper; read it alongside section 6, row 5e. - T. T. Brown, British patent GB 300,311, "A Method of and an Apparatus or Machine for Producing Force or Motion" (applied 1927, accepted 1928), reprinted in Gravity v.2 pp.52-59 [read; title from memory: medium]. - J. Dunning-Davies and R. L. Norman, "The Enigma that is Light" (undated manuscript) [read]. - "Scalar Waves", CIA memo dated 22 December 1987, CIA-RDP96-00792R000500240001-6 [read: page images]. - P. M. Jack, "Physical space as a quaternion structure I: Maxwell equations" (2003) [web: https://arxiv.org/abs/math-ph/0307038; title from memory: medium], the source of the quaternion treatment.

---

Answers

1. E = −∇phi − (1/c)∂A/∂t = −∇(−S_t/c) − (1/c)∇S_t = (1/c)∇S_t − (1/c)∇S_t = 0. B = ∇ × ∇S = 0 (curl of a gradient). Energy density (E² + B²)/8π = 0. These potentials are a gauge transformation of "no field at all".
2. T → T + ∇²chi + (1/c²) chi_tt. L → L + ∇²chi − (1/c²) chi_tt. Only L shifts by the relativistic wave operator (d'Alembertian). Because T can be changed freely by choosing chi, it is not a measurable field.
3. Divergence of (3): 0 = (1/c) ∂(∇·E)/∂t + ∇²T. With (5): (1/c²) T_tt + ∇²T = 0. The trial gives −k² − ω²/c² = 0, so ω = ±ick. Imaginary ω means exp(±ckt) growth or decay in time, not a shape moving at a speed; the equation is elliptic (Laplace-type), not a wave equation.
4. Step 1: (69.925319 − 69.926022) × 931.494 = −0.655 MeV (forbidden on its own). Step 2: (69.926022 − 69.924249) × 931.494 = +1.652 MeV. Chain: +0.997 MeV, positive, so mass-energy is released. But each event costs a 1.294 MeV photon and returns only about 0.87 MeV in electrons, because the antineutrinos escape with the rest. Even at 100% efficiency the output is 67% of the input, so it is a net loss, not a fuel (example.py Part F).
5. (a) 1000 / (175 × 10⁻⁶) ≈ 5.7 × 10⁶. (The abstract's separate range of 5 × 10⁴ to 1.5 × 10⁶ refers to different, open-loop power models, so there is no contradiction between the two figures.) (b) An amplifier's output power comes from its power supply; the small input only controls it, so gain alone says nothing about where the energy came from. You must measure every input, including the supply. A closed loop with no supply and no battery that keeps delivering power would be creating energy from nothing, which conservation of energy forbids. That is why the battery-less claim, not the gain, is the real over-unity claim.