Primer: Stephan Gift (ether, gravity, magnetism)

Study guide for the Gift-Stephan folder. It is based on the arbiter's final verdicts (arbiter.md, including its "QE revision" section and that section's "Round 2" subsection) and on the physicist's, mathematician's and contrarian's reviews. I redid the worked example in section 5 myself; example.py reproduces every number in it.

Page numbers in this primer are PDF page numbers unless marked "printed". The 2018 and 2009 PDFs start with a cover page, so PDF page = printed page + 1 (arbiter, QE revision).

Source tags: [read] = taken from Gift's papers in the folder; [read: page image] = read from a rendered page because the text layer was garbled; [web] = an Internet source, with its URL; [memory] = recalled, with confidence high/medium/low; [calc] = computed in example.py.

Source gap (carried over from Steps 1-3, updated after QE). The text layers of the 2001 gravity paper and the 2009 magnetism paper are damaged. After QE, the gauge constructions were re-read from page images (2001 PDF pp.3-10; 2009 PDF pp.3-5), as were the 2006 equations (PDF pp.4-12) and Shtyrkov's Figure 5 in the 2018 paper (PDF pp.18-20). In a second QE round the 2009 covalent-bond sections (PDF pp.5-9: Earnshaw, Greenspan's full list, HCl, "rule of two", meson future work, references) were also deep-read, from the text layer plus page images of pp.6 and 8, so the 2009 paper is now read in full (pp.3-9) (arbiter, Round 2). Several works Gift relies on are not in the folder, so everything said about them comes second-hand through Gift or from [memory]: Shtyrkov's geostationary-satellite aberration data, the Marmet and Kelly GPS reports, Greenspan's bond simulations, Luder's book on chemical bonding, Barut's e⁺e⁻ scattering work, Gillespie's bonding text, Aspden's 1980 force law, Hatch's "clock bias" papers, and the Miller/Demjanov/Galaev/Munera ether-drift experiments. The conclusions on Shtyrkov, Greenspan, Luder, Barut and Aspden below rest on Gift's own account plus [memory], not on the primary papers.

The folder holds four papers [read]: - 2001, Physics Essays: "A Negation of Einstein's General Theory of Relativity..." (gravity) - 2006, Journal of Scientific Exploration 20(2):201-214: "The Relative Motion of the Earth and the Ether Detected" (Roemer and Doppler effects) [read] - 2009, Progress in Physics (PP-16-03): magnetism as a separate force carried by "magnatons" - 2018, Applied Physics Research preprint: "On the Existence of a Cosmic Ether: Detection of the Rotational and Orbital Motions of the Earth Using GPS Technology"

Gift is an electrical engineer at the University of the West Indies (Trinidad) [read: 2018 title page].

---

1. The idea in one paragraph

Gift argues that light does not really travel at the same speed in every direction for everyone. On the spinning Earth, he says, a light signal sent east goes slightly slower than one sent west, because the ground moves east under it. He finds this difference in the clock-synchronisation rules used by the Global Positioning System (GPS) and in old astronomical effects such as Roemer's timing of Jupiter's moon Io. From this he concludes that there is a cosmic "ether", a medium or preferred frame that light moves through, and that Einstein's special relativity is therefore wrong. In separate papers he also argues that Einstein's general relativity fails, and offers a simpler Newton-style theory of gravity in its place. He further argues that magnetism is a force of its own, carried by a new particle he calls the "magnaton", and that this force explains chemical bonding. The measured numbers Gift starts from are real and correct. The question this primer studies is whether those numbers mean what he says they mean.

---

2. What you need to know first

2.1 Reference frames and Galilean velocity addition. A reference frame is a set of rulers and clocks used to label events with positions x and times t. An inertial frame is one in which a free object moves in a straight line at constant speed. In Newtonian (Galilean) physics, if a train moves at speed v and you throw a ball forward at speed u inside it, the ball moves at u + v relative to the ground. Gift's c - v and c + v are this rule applied to light. Read: J. R. Taylor, Classical Mechanics, the chapters on Newton's laws and on non-inertial frames (ch. 9 covers rotating frames) [memory, high].

2.2 Clock synchronisation and the one-way speed of light. Measuring a speed means dividing a distance by a time. For a one-way trip from A to B, the start time is read on clock A and the end time on clock B, so the result depends on how the two clocks were set to agree. Setting them requires sending a signal between them, and the signal's speed is the thing we are trying to measure. Einstein's convention sets them by assuming the trip out takes as long as the trip back. A single clock timing a round trip (A to B and back to A) needs no convention, and every such measurement gives c. The one-way speed is therefore partly a matter of convention [web: https://en.wikipedia.org/wiki/One-way_speed_of_light, cited by the physicist and contrarian]. This one idea settles most of the argument about claims 1-3. Read: E. F. Taylor and J. A. Wheeler, Spacetime Physics (2nd ed.), which builds relativity from clocks and light flashes [memory, high].

2.3 Special relativity (SR) basics. SR rests on two postulates: (1) the laws of physics are the same in every inertial frame; (2) light moves at c in every inertial frame, with clocks set by Einstein's convention. The consequences are time dilation (moving clocks run slow by the factor γ, defined in section 3), length contraction, and the relativity of simultaneity: clocks that are synchronised in one frame are out of step in another. The relativity of simultaneity is the key to section 5. Read: D. J. Griffiths, Introduction to Electrodynamics, ch. 12 "Electrodynamics and Relativity" [memory, high].

2.4 Rotating frames and the Sagnac effect. A rotating frame is not inertial. You can detect rotation from inside a closed box with a gyroscope or a Foucault pendulum, without looking outside. Light can detect it too. Send two beams in opposite directions around a closed loop on a rotating platform, and the beam going with the rotation arrives later. This is the Sagnac effect (1913). Michelson and Gale measured it for the rotating Earth in 1925 [web: https://en.wikipedia.org/wiki/Sagnac_effect, cited by the physicist]. Both Newtonian physics and SR predict it. Read: the Taylor and Wheeler book above, and N. Ashby, "Relativity in the Global Positioning System", Living Reviews in Relativity (2003), free online [memory, high].

2.5 Lorentz ether theory and test theories. Before Einstein, Lorentz and Poincaré built an ether theory in which rods moving through the ether shrink and clocks moving through it slow down by exactly the right amounts. In that theory every experiment gives the same result as in SR. The only difference is which frame's clocks you call "really" synchronised. The Robertson-Mansouri-Sexl "test theories" make this precise. A theory with exact relativistic time dilation and length contraction "is experimentally equivalent to special relativity, independent of the chosen synchronization" [web: https://en.wikipedia.org/wiki/Test_theories_of_special_relativity, fetched by the arbiter]. Gift calls this contraction FLL (Fitzgerald-Lorentz-Larmor) contraction.

2.6 The equivalence principle and accelerated charges. The equivalence principle (EP) says that inside a small closed box you cannot tell standing still in a gravitational field from accelerating in empty space. The Larmor formula gives the power radiated by an accelerating charge q, as seen from an inertial frame: P = q²a²/(6πε₀c³), where a is the acceleration and ε₀ is the permittivity of free space. A charge resting on a table in a gravitational field raises an old puzzle: does it radiate? Read: Griffiths, Introduction to Electrodynamics, ch. 11 "Radiation" [memory, high].

2.7 Gauge invariance, in plain language. In quantum mechanics a particle is described by a wavefunction ψ (psi), a complex number at each point in space. Multiplying ψ by a phase factor e^{iα}, where α is a constant, changes nothing you can measure. Now ask for the same freedom when α varies from point to point. The equations then only stay unchanged if you add a new field that shifts to cancel the change in α. In quantum electrodynamics (QED, the quantum theory of light and charges) that field is the electromagnetic potential, and its quantum is the photon. This "gauge argument" is how modern physics introduces force-carrying fields. Gift uses the same trick, with α allowed to vary only in time (for gravity) or only in space (for magnetism). One more piece matters for the verdicts. A gauge theory has two parts: the "matter" part (how ψ couples to the new field) and a "kinetic" part (the energy of the new field itself, which is what lets it carry waves). Both parts must stay unchanged under the shift. If the kinetic part is not, the theory is not gauge invariant, however neat the matter part looks. In electromagnetism the kinetic part is built from E and B, which do not change when the potentials shift; that is why it works. Read: D. J. Griffiths, Introduction to Elementary Particles, the chapter on gauge theories [memory, high for the book; chapter number not given].

2.8 Identical particles and the Pauli principle. Electrons are identical fermions: a state of two electrons must change sign when you swap them. This antisymmetry forbids two electrons from sharing the same quantum state (the Pauli exclusion principle). In atoms and molecules it also produces "exchange" energies of a few electron-volts (eV), with no extra force required. Read: D. J. Griffiths, Introduction to Quantum Mechanics (2nd ed.), ch. 5 "Identical Particles" [memory, high].

2.9 Earnshaw's theorem and why matter is stable. Earnshaw's theorem (1842) says that no arrangement of charges held only by static electric forces can be in stable equilibrium: the electrostatic potential obeys Laplace's equation ∇²V = 0 in empty space, and such a function has no local minimum, only saddle points [memory, high]. Taken classically, that means atoms built from charges should collapse or fly apart. Gift's use of it is a single sentence: the magnetic interaction "in the repulsive core mechanism is supported by Earnshaw's theorem", citing Gottfried and Weisskopf [read: 2009 PDF p.6, via the Round 2 deep read]. He applies it only to the repulsive core, not to the stability of atoms as a whole, states none of the theorem's assumptions, and never mentions the uncertainty principle or antisymmetry [read: 2009 pp.5-9]. The reply below is therefore the reviewers' own answer, not something the paper responds to. The textbook answer is that the theorem is classical and static. In quantum mechanics an electron squeezed into a small region must have a large momentum (the uncertainty principle, Δx Δp ≳ ħ/2, where ħ = 1.055 × 10⁻³⁴ J s is the reduced Planck constant), so its kinetic energy grows as the region shrinks; this stops the collapse and sets the size of atoms. Pauli antisymmetry (2.8) then stops matter in bulk from collapsing [memory, high]. Read: Griffiths, Introduction to Electrodynamics, ch. 3 (Laplace's equation, where Earnshaw's theorem is stated) [memory, high for the topic; chapter medium]; Griffiths, Introduction to Quantum Mechanics, ch. 1 on the uncertainty principle [memory, high].

---

3. The textbook view

3.1 Light timing between two clocks (claims 1-3)

Symbols: - c = 299 792 458 m/s, the speed of light in vacuum. - ECI (Earth-centred inertial) frame: a frame centred on the Earth whose axes do not rotate. GPS computes all its light paths in this frame [read: 2018 p.4, citing the IS-GPS-200E specification]. - ω (omega) = 7.292 × 10⁻⁵ rad/s, the Earth's rotation rate. - R = 6.378 × 10⁶ m, the Earth's equatorial radius. At latitude λ the ground moves east at v = ωR cos λ (465 m/s at the equator). - l = the distance between two clocks, A (west) and B (east). - γ = 1/√(1 - v²/c²), the Lorentz factor (equal to 1 to within about 10⁻¹² here).

Textbook statement: in the ECI frame light moves at c in every direction. Clocks on the ground are set to a shared "coordinate time" defined in the ECI frame. Using these clocks, a signal sent from A east to B takes

Δt_east = l/c + lv/c², Δt_west = l/c - lv/c² (to first order in v/c).

The extra term lv/c² is called the Sagnac correction. GPS and the international time-transfer rules (the CCIR, now ITU, formula) apply it as a matter of routine [read: 2018 eqs 1-7; memory, high for routine use]. Around a closed loop the corrections add up to

Δt_Sagnac = 4ωA/c²,

where A is the area the loop encloses, projected onto the equatorial plane. A single clock can measure this (a ring laser does), so it does not depend on any synchronisation convention.

The SR reading, which section 5 checks with numbers: the ground clocks are not inertial and are not synchronised by Einstein's convention in their own rest frame. Their readings differ from Einstein-synchronised ones by the relativity-of-simultaneity offset γvl/c². SR's postulate 2 says nothing about this coordinate speed, because the postulate applies to inertial frames with Einstein-synchronised clocks.

3.2 First-order and second-order tests (claim 3)

If the apparatus moves at v through an ether and nothing contracts, light down and back along an arm of length l takes l/(c-v) + l/(c+v) = 2l/c × 1/(1 - v²/c²). The v/c terms cancel and only v²/c² is left. Round-trip experiments such as Michelson-Morley are therefore called "second order". With FLL contraction this v²/c² term also disappears. First-order effects (proportional to v/c) only show up in one-way timing, and one-way timing depends on the synchronisation convention (section 2.2).

3.3 Gravity (claim 4)

Newton: force F = GMm/r², where G = 6.674 × 10⁻¹¹ N m² kg⁻² is the gravitational constant, M and m are the two masses, and r is their separation. The potential Φ obeys Poisson's equation ∇²Φ = 4πGρ, where ρ is the mass density. General relativity (GR) reduces to this for weak fields and adds corrections. Writing u = 1/r and h = angular momentum per unit mass, the GR orbit equation is

d²u/dφ² + u = GM/h² + 3GMu²/c².

The extra 3GMu²/c² term gives Mercury's extra perihelion advance of 43 arcseconds per century. For light it gives a deflection of 4GM/(c²R) at closest distance R, which is 1.75 arcsec at the Sun's edge. Newtonian physics with a photon "mass" E/c² gives half that, 0.876 arcsec [calc]. A pure scalar theory of gravity (Nordström's) gives no light bending at all, and a perihelion shift of about -7 arcsec per century, with the wrong sign [web: https://en.wikipedia.org/wiki/Nordström's_theory_of_gravitation, fetched by the physicist and mathematician].

3.4 Magnetism and chemical bonds (claim 5)

Magnetism is part of electromagnetism. The magnetic field is B = ∇ × A, where A is the vector potential, and in magnetostatics ∇²A = -μ₀J, where μ₀ is the permeability of free space and J is the current density. Two magnetic dipoles μ₁ and μ₂ a distance r apart have an interaction energy of order (μ₀/4π)μ₁μ₂/r³. The electron's dipole moment is about one Bohr magneton, μ_B = 9.274 × 10⁻²⁴ J/T. Chemical bonds come from the Coulomb force plus Pauli antisymmetry (section 2.8).

---

4. Where Gift departs

---

5. Worked example: the GPS "one-way speed of light"

This is Gift's central argument (2018 paper, eqs 1-8) [read]. We redo it with real numbers and then check it against SR. Code: example.py, steps 1-8.

Setup. Clock A sits at latitude 40° N. Clock B is l = 1000 km east of A on the same latitude circle. Both clocks are set by the GPS/CCIR rule, which keeps them on ECI coordinate time.

Step 1: how fast is the ground moving? The radius of the latitude circle is r = R cos 40° = 6.378 × 10⁶ × 0.7660 = 4.886 × 10⁶ m. The ground speed is v = ωr = 7.292 × 10⁻⁵ × 4.886 × 10⁶ = 356.3 m/s, so v/c = 1.19 × 10⁻⁶.

Step 2: the ECI picture. In the ECI frame light moves at c. While the pulse travels, B moves away from it at v. The pulse catches B when ct = l + vt, so t = l/(c - v). - l/c = 3.335 641 0 ms - eastward, l/(c - v) = 3.335 644 9 ms - westward, l/(c + v) = 3.335 637 0 ms

Step 3: the first-order offset. l/(c - v) ≈ l/c + lv/c², and

lv/c² = (10⁶ m)(356.3 m/s)/(8.988 × 10¹⁶ m²/s²) = 3.964 ns.

Gift's own form of the correction, 2ωA′/c² with A′ = (l/2πr)·πr² (his eq 3), gives the same 3.964 ns [calc]. So eq 4 in his paper is right.

Step 4: Gift's division. l/Δt_east = 299 792 101.7 m/s, which equals c - v to every digit shown, and l/Δt_west = c + v. This was bound to happen. We put "light moves at c in ECI" in at step 2, so dividing by the ground distance has to give back c ∓ v. This is where the inference breaks. The "measurement" repeats the assumption it started from. Gift himself states that the GPS specification assumes c in ECI [read: 2018 p.4].

Step 5: what does SR say about the same clocks? Take the inertial frame S′ that moves along with the clocks at that moment. In S′, with Einstein-synchronised clocks, the trip takes l′/c, where l′ is the separation measured in S′. The Lorentz transformation turns S′ times into ECI times: t = γ(t′ + vx′/c²). The pulse arrives at x′ = l′, t′ = l′/c, so its ECI arrival time is γ(l′/c + vl′/c²). SymPy shows this equals l/(c - v) exactly, once you use l = l′/γ (step 5 of the script prints 0) [calc].

So special relativity predicts Gift's number exactly. The lv/c² term is the relativity-of-simultaneity offset γvl/c² = 3.964 ns between two ways of setting the same pair of clocks. Both theories agree on every clock reading. They disagree only on which clock setting to call "true".

Step 6: remove the convention. Put a mirror at B and time the round trip on clock A alone:

l/(c - v) + l/(c + v) - 2l/c = 2lv²/c³ = 9.4 × 10⁻¹⁵ s.

The ±3.964 ns terms cancel exactly. What remains is second order, 10⁻¹⁵ s, and in Lorentz-type theories it is removed by time dilation and contraction. Once the convention is taken out, the first-order "ether signal" is gone.

Step 7: the part that is real. Now send light all the way around the equator (v = 465.1 m/s, enclosed area A = πR² = 1.278 × 10¹⁴ m²). The eastward minus westward time is

4ωA/c² = 4 × 7.292 × 10⁻⁵ × 1.278 × 10¹⁴ / 8.988 × 10¹⁶ = 414.8 ns.

This matches Kelly's 414 ns, quoted by Gift [read: 2018 p.4]. It does not depend on any convention, because a pulse that goes all the way round comes back to the same clock. But it measures rotation (ω), not speed through a medium, and SR in the ECI frame predicts it. Moving in a straight line at constant velocity (for example the Earth's 29.8 km/s orbital motion) gives no such loop effect.

Step 8: scale check. If there were a Galilean ether with no contraction, a two-way (resonator) test would show a fractional effect of about v²/2c². That is 1.2 × 10⁻¹² for the Earth's spin and 4.9 × 10⁻⁹ for its orbit. The cavity tests that Gift cites set limits near 10⁻¹⁷ [read as cited]. So a medium without contraction is ruled out. A medium with contraction makes the same predictions as SR (section 2.5). Either way, the GPS numbers cannot pick out an ether. (The arbiter's practical check put the expected size at v/c; for a two-way test the correct scale is v²/c², which is still more than 10⁵ times the limit.)

Lesson. Every number in Gift's derivation is right. The mistake is in what the numbers are taken to mean: a speed that depends on a convention is reported as a measurement.

---

6. Evidence table

Verdicts are the arbiter's (arbiter.md), on the project scale.

---

7. What a learner should take away

Solid. - The Sagnac effect is real, and rotation can be detected without looking outside (414.8 ns around the equator). - GPS clocks really do apply a ±lv/c² correction between east and west. - Gift is right that the one-way speed of light is a convention, not a measured fact, and that the textbook line "c is the same in every direction for everyone" mixes two-way measurements with a convention. - He is right that "Michelson-Morley disproved the ether" overstates things: a Lorentz-type ether with contraction survives Michelson-Morley by design. - The charge-in-a-gravitational-field puzzle is a real and subtle problem. - The "Pauli force" is an effective description, not a fundamental force.

Wrong, and why the mistake is instructive. 1. Circular inference. Section 5 shows how a convention fed into a formula comes back out looking like a measurement. Before you call something "measured", ask what was assumed in order to compute it. 2. Misreading SR. SR's postulate is about inertial frames with Einstein-synchronised clocks. A coordinate speed of c ∓ v on a rotating surface, with ECI-synchronised clocks, is what SR predicts. 3. A non sequitur. "My ether theory gives the same results as SR, therefore SR is refuted" does not follow. Two theories that predict the same results cannot be told apart by those results. 4. Borrowed coefficients. The gravity paper gets GR's numbers by importing GR's correction term. Kept consistently, its own equation gives the wrong light bending. 5. Orders of magnitude. A force 10⁻⁵ times too weak cannot hold molecules together. Estimating the size of an effect early (side check B in the script) catches this kind of mistake in a few lines. 6. Half a gauge theory. Both of Gift's gauge constructions show that the matter part is invariant and stop there. The field's own kinetic term is never written, and for gravity the one he needs is not invariant under his own shift (section 4). Checking only the convenient half of a symmetry is a common way to fool yourself. The same pattern (relax Lorentz covariance, then announce a particle) recurs: 2001 Table II claims pions and a strong-force unification, and the 2009 paper says isospin rotations may give rho-mesons and pions unified with gravitons, all asserted without calculation, as hoped-for results backed only by self-citations (Gift refs 22 and 23), with no masses or couplings [read: 2001 Table II; read: page image 2009 p.8]. 7. What is the source? Gift's gravity couples to the sum of particle masses, so binding energy would not gravitate and different materials would fall at different rates by about 10⁻³, some ten billion (10¹⁰) times above the measured limits. Always ask what a new field couples to, then test the consequence against the best existing experiment. 8. A classical theorem used outside its domain. Earnshaw's theorem is true for static classical charges; atoms are quantum objects, and the uncertainty principle stabilises them (section 2.9). Gift cites it in one sentence without stating its assumptions. Before using a theorem, check its assumptions. 9. Speed without direction. Gift calls Marinov's 303 ± 20 km/s "comparable" to the CMB's 390 ± 30 km/s. They differ by 87 km/s, about 2.4 combined standard deviations (√(20² + 30²) = 36 km/s), no direction is compared, and Gift's Roemer analysis gives yet another speed (30 km/s) for the same Earth through the same ether [calc, arbiter]. 10. Second-hand confirmation. Gift calls Greenspan's simulations "an extremely strong indication" of his model, but quotes no diameters, no force law and no attraction strength, and a classical simulation has no spin to tell a magnetic force from any other attraction. He also cites Luder's electron-repulsion book without saying it argues the opposite. When a paper leans on someone else's result, ask what numbers it actually reports and whether the cited source supports the specific claim.

Genuinely open (small, but real). - Shtyrkov's geostationary aberration measurement: open only in the sense that nobody is on record as having tried to replicate it. The figure Gift reproduces (2018 Fig. 5) is too thin to count as evidence: no error bars, no fit, no stated method of turning the antenna data into a speed, and no stated subtraction of the satellite's own small orbital motion, pointing or refraction. The annual change Gift calls "evident" (29.3 to 30.3 km/s) is about as large as the scatter [read: page image 2018 p.18; arbiter]. A year-long replication with a dish, a geostationary beacon and a standard reduction would settle it. SR predicts no seasonal 20″ term. - Marinov's coupled-mirror "absolute speed" of 303 ± 20 km/s: no independent replication and no neutral negative test found [web: https://en.wikipedia.org/wiki/Stefan_Marinov]. To settle it, read Marinov's 1980 paper and search for independent coupled-shutter repeats. Modern one-way and resonator tests already bound anisotropies far below what a 300 km/s ether wind would give without contraction (section 5, step 8). - Was Miller's drift data fairly reanalysed? The mainstream account (Shankland 1955; Joos found about 1/50 of Miller's effect) is persuasive, but DeMeo's counter-reading, which Gift cites, has not been evaluated here. - The H₂ benchmark: the 2009 paper prints the experimental frequency as 1.38 × 10¹⁴ Hz (confirmed on the page image), while the standard value is about 1.32 × 10¹⁴ Hz, about 4.5% lower [read: page image 2009 p.6; memory, high]. Whether this is a misprint or Greenspan's own figure needs Greenspan's primary papers. It does not change the argument, which compares 2.20 × 10¹⁴ Hz with about 1.3-1.4 × 10¹⁴ Hz. Barut's "sharp resonances and novel asymmetries" in e⁺e⁻ scattering, which Gift cites second-hand with no energies or data [read: 2009 p.5], are likewise unchecked; no verdict depends on them. (The other small puzzle, the "3 × 10⁻⁴²" gravity strength in Gift's 2001 Table II, is now explained: G m_p m_e/(ħc) = 3.2 × 10⁻⁴², where m_p and m_e are the proton and electron masses, not the proton mass squared [calc, arbiter QE revision].)

---

8. Check yourself

1. Two ground clocks 500 km apart on the equator are set to ECI coordinate time. What east-west offset lv/c² do they show? Does that offset prove that light is slower going east?
2. Why does a single clock timing a round trip to a mirror and back show no first-order (v/c) effect, while the Sagnac loop around the equator does show a first-order effect in ω?
3. Gift's 2006 paper derives Roemer and Doppler results equal to SR's v/c, then says SR is invalid. State in one sentence why that does not follow.
4. Gift's 2001 orbit equation is u″ + u = GM/h² + 3GMu²/c². For a photon, h = cR. Which terms give 2GM/(c²R), 4GM/(c²R) and 6GM/(c²R)? Which one did Gift report, and how?
5. Estimate the magnetic dipole energy of two electrons 1 Å apart, using (μ₀/4π)μ_B²/r³. Compare it with the 4.5 eV bond energy of H₂.

---

9. Further reading

Mainstream - E. F. Taylor and J. A. Wheeler, Spacetime Physics, 2nd ed. (Freeman, 1992): clocks, simultaneity, light [memory, high]. - D. J. Griffiths, Introduction to Electrodynamics, ch. 11 (radiation, Larmor) and ch. 12 (relativity) [memory, high]. - N. Ashby, "Relativity in the Global Positioning System", Living Reviews in Relativity 6 (2003): the Sagnac correction in GPS, explained by a GPS expert [memory, high]. - R. Anderson, I. Vetharaniam, G. E. Stedman, "Conventionality of synchronisation, gauge dependence and test theories of relativity", Physics Reports 295 (1998) [memory, high]: the technical version of section 2.2. - R. Mansouri and R. U. Sexl, "A test theory of special relativity" I-III, General Relativity and Gravitation 8 (1977) [memory, high]; overview at https://en.wikipedia.org/wiki/Test_theories_of_special_relativity. - C. M. Will, "The Confrontation between General Relativity and Experiment", Living Reviews in Relativity (updated editions) [memory, high]: equivalence-principle and light-bending tests. - D. G. Boulware, "Radiation from a uniformly accelerated charge", Annals of Physics 124 (1980) [memory, high]; overview at https://en.wikipedia.org/wiki/Paradox_of_a_charge_in_a_gravitational_field. - D. J. Griffiths, Introduction to Quantum Mechanics, ch. 5 (identical particles, exchange) [memory, high].

Gift's own works (in the folder) - S. J. G. Gift, "On the Existence of a Cosmic Ether: Detection of the Rotational and Orbital Motions of the Earth Using GPS Technology", preprint/Applied Physics Research, 2018 (APRExistenceoftheEther2018.pdf). Start here; read §2.1 alongside section 5 above. - S. J. G. Gift, "The Relative Motion of the Earth and the Ether Detected", Journal of Scientific Exploration 20(2):201-214, 2006 (Ether-Detected-Gift-Stephen.pdf) [read]. PDF pp.11-12 hold the Marinov and Weisskopf passages. - S. J. G. Gift, "A Negation of Einstein's General Theory of Relativity...", Physics Essays 14(4), 2001 (A_Negation_of_Einsteins_General_Theory_of_Relativ.pdf); pp. 6-7 hold the field equation, the imported force law and the light-bending step. - S. J. G. Gift, magnetism/magnaton paper, Progress in Physics, 2009 (PP-16-03.PDF).

Gift's earlier papers on the same GPS and light-speed claims (not in the folder; cited in the 2018 paper's reference list [read]) - "One-Way Light Speed Measurement Using the Synchronized Clocks of the Global Positioning System (GPS)", Physics Essays 23(2):271-275, 2010 (ref 28). - "Light Speed Invariance is a Remarkable Illusion", Physics Essays 23, 1, 2010 (ref 54); "Doppler Shift Reveals Light Speed Variation", Apeiron 17, 13, 2010 (ref 55). - "One-Way Light Speed Determination Using the Range Measurement Equation of the GPS", Applied Physics Research 3, 110, 2011 (ref 32); "Successful Search for Ether Drift in a Modified Michelson-Morley Experiment Using the GPS", Applied Physics Research 4(1):185-192, 2012 (ref 30); "Light Transmission and the Sagnac Effect on the Rotating Earth", Applied Physics Research 5, 93, 2013 (ref 29); "Time Transfer and the Sagnac Correction in the GPS", Applied Physics Research 6, 1, 2014 (ref 34). - "One-Way Speed of Light Using Interplanetary Tracking Technology", Physical Science International Journal 4, 780, 2014 (ref 49).

Others Gift relies on (not in the folder; not read) - S. Marinov, General Relativity and Gravitation 12, 57 (1980): the coupled-mirror "absolute motion" experiment [read as cited in 2006, ref 44]; background at https://en.wikipedia.org/wiki/Stefan_Marinov. - R. Wang (GPS and SR, ION 2000 proceedings, ref 33) and C.-C. Su, "A Local-Ether Model of Propagation of Electromagnetic Wave", European Physical Journal C 21, 701-715 (2001) (ref 56) [read as cited in 2018]. - D. Greenspan, Physics Essays 5, 250 (1992) and 5, 554 (1992): the classical bond simulations Gift cites as confirmation (refs 15, 16) [read as cited in 2009]. - Luder, The Electron-Repulsion Theory of the Chemical Bond (Reinhold, 1967) (ref 18) [read as cited in 2009]: argues for electron repulsion, the opposite of Greenspan's thesis [memory, medium]. - S. Albrecht, S. Reffert, I. Snellen et al., 2009 work on DI Herculis spin-orbit misalignment, which resolved the apsidal-motion puzzle Gift cites against GR [web: https://en.wikipedia.org/wiki/DI_Herculis].

---

Answers

1. v = 465.1 m/s and l = 5 × 10⁵ m, so lv/c² = 5 × 10⁵ × 465.1 / 8.988 × 10¹⁶ = 2.59 ns. It does not prove that light is slower going east. The offset is what you get by setting the clocks to ECI time, and SR predicts the same offset as the relativity of simultaneity γvl/c² (section 5, step 5).
2. On the round trip, l/(c-v) + l/(c+v) has the +lv/c² and -lv/c² terms cancel, so only v²/c² is left. Around a closed loop on a rotating body, the path the light covers is longer in one direction for the whole trip, because the loop itself turns. The effect depends on the enclosed area times ω, which is a property of rotation, not of the straight-line velocity v.
3. If two theories give identical predictions for an observation, that observation cannot favour either one, so results equal to SR's cannot refute SR.
4. With h = cR, the right-hand side is (GM/c²R²)(1 + 3cos²φ). The "1" (the Newtonian GM/h² term) alone gives 2GM/(c²R). The 3cos²φ term alone gives 4GM/(c²R). Both together give 6GM/(c²R) = 2.63″ (side check A in example.py). Gift reported 4GM/(c²R) = 1.75″ by dropping the GM/h² term, on the grounds that h is infinite for a photon. In fact h = cR is finite [read: page image 2001 p.7, confirmed by QE and the 2001 deep read; algebra checked by the arbiter and by example.py]. Right after quoting 1.75″, Gift also cites Marmet and Couture as arguing that deflection data carry large systematic errors [read: page image p.7], a hedge that sits oddly next to claiming the 1.75″ result.
5. U = 10⁻⁷ × (9.274 × 10⁻²⁴)² / (10⁻¹⁰)³ = 8.6 × 10⁻²⁴ J = 5.4 × 10⁻⁵ eV. That is about 10⁻⁵ of the 4.5 eV bond energy, and about 4 × 10⁻⁶ of the 14.4 eV Coulomb energy at the same distance. Magnetic dipole forces cannot account for the bond.