Primer: T. Henry Moray and the "Sea of Energy"

Project folder: Morray-TH (Thomas Henry Moray, 1892-1974, Salt Lake City). Rewritten after quality engineering (QE, qe.md, verdict MAJOR) from the re-run arbiter.md (authoritative verdicts), the three re-run Step 2 reviews, and the six deep reads in deep/. The new material comes chiefly from the 1960 4th edition (Moray's own last edition) and the 1978 5th edition (edited by his son John E. Moray) of The Sea of Energy in Which the Earth Floats, which now have full OCR text. Revised again on 2026-10-03 to follow the arbiter's "Evidence-rule revision": single-source testimony is quoted and weighed rather than set aside by label, statements by parties with a stake are attributed to them, and gaps in the record are analysed (new subsection "Silences in the record" after the evidence table). New outside records come from the Investigator's notes in research/ and sites.md. Updated again on 2026-10-04 for the arbiter's "Investigator findings (2026-10-04)": the printed 1944 Senate hearing, an Army report of 1936, the family's 1946 radiation treatise for Cosray, Tesla's 1901 collector, early crystal amplifiers (Losev, Lilienfeld), and the FBI file's later "No Records" answer. Changed passages are marked (2026-10-04). Updated again on 2026-10-08 for the arbiter's "Investigator findings (2026-10-08)", which weighs the owner's follow-up questions (Q01, Q02, Q04, Q08, Q09, Q11, Q12): much of the physics printed in Moray's books turns out to be other authors' text (Le Bon 1905, Glasser 1944, Preston 1928) with the credits removed; new rows 5e (resonance transmutation) and 5f (Le Bon's observations); new parts for Claims 3 and 4 (photoelectric collection, neutrinos, radioisotope conversion, radon plating, a radium-bearing stone, the Foster 1990 pellet account); and a new Silence G. Changed passages are marked (2026-10-08); the new numbers are in example.py Step 7. QE remediation of 2026-10-08: row 5f now carries the "F. L." reply to Campbell (with its stake and second-hand status) and Le Bon's Strutt footnote with its answer, as in arbiter.md 5f; detector, experiment and isotope names are glossed at first use.

Source tags: [read] = from the documents in the folder (page numbers are PDF pages: "5th ed." = 1978 scan, "4th ed." = 1960 scan); [web] = Internet source with URL; [memory] = recalled, with confidence; [calc] = computed in example.py in this folder; [computed] = arithmetic recomputed in the evidence-rule recheck of 2026-10-03.

Source gap (rewritten after QE)

• The 4th (1960) and 5th (1978) editions now have OCR text and were digested in full. Deep reads cover 5th ed. pp.32-133 (the demonstrations, sworn statements and correspondence) and 4th ed. pp.116-197 (Moray's own history). Other pages of both books (litigation, patents, the Rural Electrification Administration (REA) episode, the 1940 shooting, the 1942-43 rebuild, theory chapters) were read only as page-by-page digests, so statements from them are less certain.
• Most witness letters and affidavits reach us through books edited by Moray or his sons, who were advocates. Several were written decades after the event (A. B. Jensen's 1938 events, statement dated 1956 or 1958, sworn 1963; Gardner 1963; Todd and Bringhurst 1971; Fletcher 1979 at age 94). Moray himself wrote that the 1932 booklet was "composed and edited too much like promotion" [read: 5th ed. p.62]. No original Securities Commission, REA, Bell Labs, utility or court document is in the folder. That this testimony comes through the family is a fact about its provenance, to be weighed; it does not make it non-evidence.
• Records outside the family's books (found by the Investigator, Step I; URLs from its notes, not re-fetched): 1944 Senate committee testimony on the REA, as reported in the press (Moray on the REA payroll Feb 1939-Feb 1941, Frazer's role, engineer Wood's report) [web https://newspapers.lib.utah.edu/details?id=28911552; web https://archive.org/details/per_chicago-daily-tribune_1944-12-11_103_296]; the 1932 Washoe County (Reno) court notice "In re Moray Products Company" [web https://newspapers.lib.utah.edu/details?id=28958019]; a 1942 Salt Lake judgment against Moray, Sanford v. Moray [web https://newspapers.lib.utah.edu/details?id=25602932]; a 1970 suit against Mark J. Hagmann [web https://newspapers.lib.utah.edu/details?id=29749860]; and an FBI FOIA log entry for a 74-page file on Moray [web https://archive.org/details/foia-log-2000-2002]. The folder itself holds a 1951 FBI interview memo [read: gov-FBI screenshot] and a letter from former REA deputy administrator Robert B. Craig [read: 5th ed. p.119].
• Added 2026-10-04 (Investigator's third run, weighed by the reviewers and the arbiter; URLs from the notes, not re-fetched): the printed 1944 Senate hearing itself (78th Congress), with the investigators' report at pp.1344-1345 and an interview with REA Administrator Slattery at pp.1370-1376 [web https://babel.hathitrust.org/cgi/pt?id=umn.31951d03587118y]; four National Archives (NARA) War Department index sheets for an Army report of 8 Dec 1936 by Brig. Gen. Shinkle, filed as "sound detector" (the report itself not found); a later FBI log entry answering "No Records" in 2013 for the same man whose 74-page file was released in 2000 [web https://archive.org/download/UnsortedFBIDocuments/foia-log-1-1-2013-to-3-31-2013-pdf_djvu.txt]; the family's 1946 Short Treatise on Alpha, Beta, Gamma Ray Therapy [web https://archive.org/details/thomas-henry-moray-radiant-energy]; and, unread, letters about Moray from 1928-57 kept by the Canadian government engineer Wilbert B. Smith in the Arthur Bray fonds at the University of Ottawa (finding aid only). Those letters were kept outside the family's editing and are the best way to check whether the family's books quote witnesses in full.
• Added 2026-10-08: the physics pages are largely borrowed. The Investigator matched, nearly word for word and in order, Radiant Energy printed pp.53-59 (alpha-particle speeds, "the experiments of Rutherford ... taken up again", the spinthariscope, phosphorescent sulphide screens and diamonds, doped phosphors, the folded-cylinder electroscope) and Isotopes in Ionic Recovery of U3O8 pp.4-6 (thorium "ten thousand times more active") to Gustave Le Bon, The Evolution of Matter (1905) [web https://www.rexresearch.com/lebonmat/lebonmat.htm]; Isotopes pp.20-25 follow Glasser, Medical Physics (1944) and Preston, The Theory of Light (1928) [web Google Books snippets]. Le Bon's "I think", his credits to Rutherford and Crookes and his footnotes are missing; a few sentences are added ("Herein be the door to greater discoveries", "and sometimes other properties", "an atomic change such as ... silver to cadmium"). The arbiter checked the key quotations against the saved pages. So these pages show what Moray read and chose to print, not what he measured, and nothing in the record says who prepared each printing or why the credits are missing. The match is against a web transcription and snippets, not the editions Moray's editor used. It does not bear on the lamp demonstrations (Claims 1-2). The folder's Radiant Energy printing labelled "1926" mentions the atomic pile and plutonium, so it was compiled after 1945; "1926" is not its date of composition.
• No document records a neutral, metered measurement of the device's input or output power. Every power figure is a sum of the rated ("nameplate") wattages of lamps and appliances. Moray's own 1960 text says instruments are "of no value" for his valves [read: 4th ed. pp.192-194].
• So: conclusions about what the box did rest on the testimony of named observers. Many were associates of Moray; some were state or REA officials (the Securities Commission inspection, Crim, Craig, Wood). Conclusions about physics rest on calculation plus measured constants ([web]/[memory]).

---

1. The idea in one paragraph

From 1925 to 1938 Moray showed many invited visitors (at least 30 named witnesses on about 25 dated occasions) a wooden box, or later several boxes, wired only to an outdoor antenna and to a ground (a water pipe or a driven rod). Lamps lit and a flat iron got hot. The early loads were one 100-watt lamp plus an iron. By 1936-38 witnesses were counting 30 to 50 large lamps, with rated loads adding up to 4-8 kilowatts. In 1938 one such demonstration was given for the Utah Securities Commissioner, who had the parts weighed and the box drilled. Moray said the box held no battery and did not steal power from nearby lines, and he took it to sites the witnesses chose, two of them about 26 miles from the nearest power line. He said the energy came from a "sea" of oscillating radiation reaching the Earth from space, caught by a tuned (resonant) circuit and turned into useful current by a secret mineral "valve". He also said the first stage had to be "excited" from outside each time before it ran on its own. He never let anyone examine the valve, and later claimed his radiation could change elements into gold and uranium and treat disease. (2026-10-08) Much of the radioactivity physics printed in his books is copied, without credit, from the French writer Gustave Le Bon's 1905 The Evolution of Matter and two later textbooks, so those pages tell us what he read rather than what he measured. Critics asked for tests he would not accept, he asked for backing they would not give, and a New York test that Harvey Fletcher of Bell Labs said he arranged never took place (Fletcher says Moray backed out; Moray called that "100% untrue"). From February 1939 to February 1941 the federal Rural Electrification Administration (REA) paid him, bought him laboratory equipment and sent engineers to watch the device; no test result from that engagement is on record, and why it ended is disputed. By his sons' account an REA investigator smashed the device in 1939 and Moray was shot at in 1940. The printed 1944 Senate hearing (now found) confirms the REA engagement but says nothing about a hammer or a shooting; the FBI and REA files that could check those events have not been read. The physics question is whether any known source could feed kilowatts to a box through an antenna. The answer is no. The historical question is what actually happened in those demonstrations. Nobody ever metered the power going in or coming out, so that question is still open.

---

2. What you need to know first

Energy conservation and power. Energy (joules, J) is never created. Power (watts, W = J/s) is the rate at which energy flows. A device that delivers power P for time t must take in energy at least P·t from somewhere: fuel, a battery, a wire, or radiation. A kilowatt-hour (kWh) is 1000 W for one hour = 3.6×10⁶ J. Text: Halliday, Resnick and Walker, Fundamentals of Physics, chapters on work and energy.

Energy density and flux of radiation. Radiation that travels at speed c and stores energy density ρ (J/m³) carries a power per unit area (flux, W/m²) of at most ρc, which applies when all of it moves in one direction. For an electromagnetic (EM) wave the flux is given by the Poynting vector S = E × H (E the electric field, H the magnetic field). Text: Griffiths, Introduction to Electrodynamics, Ch. 8 (conservation laws, Poynting's theorem) and Ch. 9 (EM waves).

RLC circuits, resonance and reactance. RLC means a resistor (R), an inductor (L) and a capacitor (C) in one circuit. Charged and left alone, the circuit rings at a natural frequency and the ringing dies away as R turns the stored energy into heat. Driven by an alternating source, it responds most strongly at its resonant frequency. A capacitor's opposition to alternating current, its reactance, is X_C = 1/(2πfC) ohms at frequency f (hertz, Hz). It is huge at 60 Hz and small at radio frequencies, which makes a capacitor a simple frequency test. Text: Halliday/Resnick/Walker, chapter on electromagnetic oscillations and alternating current; Horowitz and Hill, The Art of Electronics, Ch. 1.

Antennas. A receiving antenna takes power out of a passing wave. The most it can take is the wave's flux times the antenna's "effective area". Text: Balanis, Antenna Theory (fundamental parameters chapter); or the ARRL Antenna Book for a practical view.

Atmospheric electricity (added 2026-10-04). In fair weather the air near the ground carries a steady vertical electric field of about 100 V/m, so a wire 27 m up sits about 2.7 kV above the ground. The air conducts very weakly, so only a tiny current flows down through it, about 2 picoamperes per square metre (pA/m²; 1 pA = 10⁻¹² A) [web https://en.wikipedia.org/wiki/Atmospheric_electricity]. A high voltage with a tiny current is a small power. Text: Feynman Lectures on Physics, Vol. II, Ch. 9 ("Electricity in the Atmosphere").

Rectifiers / crystal detectors. Some minerals, such as galena (lead sulfide), let current pass more easily one way than the other. The "cat's-whisker" crystal radio of the 1900s-1920s used this to turn radio-frequency alternating current (AC) into audible direct current (DC). A rectifier changes the form of electrical energy. It adds none. Text: Horowitz and Hill Ch. 1 (diodes); historical overview [web https://en.wikipedia.org/wiki/Crystal_detector].

Batteries as stored energy. A lead-acid battery, the only practical rechargeable kind in the 1920s-30s, stores about 35-40 watt-hours per kilogram (Wh/kg) [web https://en.wikipedia.org/wiki/Lead%E2%80%93acid_battery]. Modern lithium-ion stores about 160-300 Wh/kg [web https://en.wikipedia.org/wiki/Lithium-ion_battery]. Dividing a run's energy by these numbers gives the battery mass needed to fake it.

Mass-energy and nuclear change. E = mc² gives the energy locked in mass. Turning one element into another means changing the number of protons (Z) in the nucleus. That takes nuclear reactions at mega-electron-volt (MeV; 1 MeV = 1.602×10⁻¹³ J) energies. Chemistry and ordinary electricity work at about 1 eV, a million times less. Text: Krane, Introductory Nuclear Physics; or any intro modern-physics text (e.g. Serway, Moses and Moyer, Modern Physics).

Radioactivity and how far radiation goes (added 2026-10-04). Radioactive nuclei emit alpha particles (helium nuclei), beta particles (electrons) and gamma rays (high-energy photons). Their reach differs enormously: alpha particles of a few MeV stop in tens of micrometres of tissue or a sheet of paper; beta particles of about 1 MeV stop in millimetres of tissue; gamma rays thin out gradually, and the thickness of a material that halves them is its half-value layer [memory, high]. A half-life is the time for half of a radioactive sample to decay; radium-226's is about 1,600 years [web https://en.wikipedia.org/wiki/Radium]. Text: Krane, Introductory Nuclear Physics (alpha, beta and gamma decay; passage of radiation through matter). (Added 2026-10-08.) Activity is counted in curies: 1 curie (Ci) = 3.7×10¹⁰ decays per second, and the becquerel (Bq) is one decay per second. In beta decay the energy Q released per decay is shared between the electron and an unseen neutrino, so Q is the electron's end-point (maximum) energy and the average electron carries only about a third to two-fifths of it [memory, medium]. To change one element into another a charged particle must get over the Coulomb barrier, the electric repulsion at the nuclear surface: for a proton and silver (atomic number Z = 47) it is about Z × 1.44 MeV·fm / R ≈ 9.8 MeV, with R ≈ 6.9 fm (1 fm = 10⁻¹⁵ m) the sum of the two nuclear radii [calc, example.py Step 7a]. Isotopes are written by element and mass number A (protons plus neutrons): Pb-210, also written ²¹⁰Pb, is lead-210; Po-210 is polonium-210 and Ra-226 radium-226; a superscript "m", as in ¹⁰⁸ᵐAg (silver-108m), marks a long-lived excited (metastable) state of the nucleus.

The photoelectric effect (added 2026-10-08). Light falling on a metal ejects electrons only if each photon carries more than the metal's work function W (a few eV). A photon's energy is E = hf = hc/λ (h = Planck's constant, 6.626×10⁻³⁴ J·s; f frequency; λ wavelength), so the threshold is a wavelength, not a brightness: brighter light ejects more electrons, not faster ones. Sunlight at the ground reaches down to about 295 nm, i.e. 4.2 eV [calc, example.py Step 7d], enough for some metals and amalgams (mercury alloys). A charged metal plate that loses its charge in sunlight is the classic demonstration (Hertz 1887, Hallwachs 1888; Elster and Geitel used amalgams from 1889) [memory, high; Elster-Geitel via the arbiter]. Text: Serway, Moses and Moyer, Modern Physics (chapter on the particle properties of light).

Controls and blanks. In a measurement of something rare (a trace of gold, a faint glow), a blank is the same procedure run with the supposed cause left out, so you can see what the procedure itself adds. A control changes one thing at a time. A blind test hides from the analyst which sample is which. No statistics beyond averages are needed in this primer; the point is simply that a result means little until you know what the blank gives. Text: J. R. Taylor, An Introduction to Error Analysis (early chapters).

---

3. The textbook view

3.1 Power carried by radiation

Take a stream of radiation with energy density ρ (joules per cubic metre) moving at the speed of light c = 3.0×10⁸ m/s toward a flat collector of area A (m²). In one second a column of length c (metres) and cross-section A passes through the collector. So the power that arrives is

P = ρ c A (watts) … (1)

This is an upper limit. It assumes every bit of the energy is moving straight at the plate and is all absorbed. If the radiation arrives equally from all directions (isotropic), the flux onto one face is ρc/4 [memory, high; standard kinetic-theory result].

For an EM plane wave with electric field E (volts per metre, V/m, root-mean-square (rms) value) the flux is

S = E² / Z₀, Z₀ = √(μ₀/ε₀) ≈ 377 Ω … (2)

Here μ₀ and ε₀ are the magnetic and electric constants of vacuum, and Z₀ is the "impedance of free space". In a travelling wave the ratio E/H is always Z₀, where H is the magnetic field in amperes per metre. Static fields, such as the Earth's steady magnetic field or the fair-weather electric field, have no such ratio and carry no steady energy flow.

3.2 What an antenna can collect

A lossless antenna matched to its load collects at most

P = S · A_eff, with A_eff = G λ² / (4π) … (3)

Here λ is the wavelength in metres, G is the antenna's gain (1 for an antenna that receives equally in all directions), and A_eff is the effective area [web https://en.wikipedia.org/wiki/Antenna_aperture]. A real antenna also loses power to wire and ground resistance. A resonant antenna (one tuned to the wave) has a far larger effective area than its wire suggests: a half-wave dipole has G = 1.64, so A_eff = 1.64 λ²/(4π) [web https://en.wikipedia.org/wiki/Antenna_aperture]. The effective area grows as λ², so at low frequencies it can be huge. What then limits the power is the field strength E, through S = E²/Z₀.

3.3 Resonance does not make energy

In a series RLC circuit driven by a source voltage V(t), the capacitor charge q(t) obeys

L q'' + R q' + q/C = V(t) … (4)

q' is dq/dt (the current I) and q'' is d²q/dt². - Undriven (V = 0): the charge oscillates at angular frequency α' = √(1/(LC) − R²/(4L²)). Its successive peaks shrink by the factor e^(−πR/(2Lα')) each half cycle. The energy goes into heat in R. - Driven at resonance (angular frequency ω₀ = 1/√(LC)): the inductor and capacitor cancel each other's opposition to current. The current amplitude is V₀/R, where V₀ is the source's peak voltage. The capacitor voltage rises to Q·V₀, where the quality factor is Q = √(L/C)/R. The average power is only V₀²/(2R).

So resonance makes voltages large. It does not raise power above what the source supplies. Over any long time, energy out ≤ energy in. A circuit given a charge at the start ("primed") and then left alone looks for a while like a sustained one, then dies away.

3.4 Measured "background" energy in our surroundings

• Cosmic rays (high-energy particles from space): about 1 eV/cm³ = 1.6×10⁻¹³ J/m³ [web https://en.wikipedia.org/wiki/Cosmic_ray].
• Starlight and the cosmic microwave background: same order, about 0.3 eV/cm³ [physicist.md; memory, high].
• Sunlight above the atmosphere: 1361 W/m² (the solar constant), i.e. ρ = 4.5×10⁻⁶ J/m³ [web https://en.wikipedia.org/wiki/Solar_constant]. The Earth's disc intercepts 1361 × π × (6.371×10⁶ m)² = 1.74×10¹⁷ W in total [calc]. The sun shines only by day.
• Fair-weather atmospheric electricity: a static field of about 100 V/m with an air-earth current of about 2 pA/m² [web https://en.wikipedia.org/wiki/Atmospheric_electricity]. The power available to a collector is at most current density times voltage, about 5×10⁻⁹ W per square metre for a 27 m antenna [calc, example.py Step 4b].

3.5 Lamps are not power meters

An incandescent lamp's rated wattage holds only at its rated voltage and frequency. A lamp run on high-frequency current, or under-driven, can glow while drawing much less. So "thirty 150 W lamps lit" is an upper label, not a measurement. To measure delivered power you need voltage and current (or a calorimeter that measures heat) [memory, high].

3.6 Making versus concentrating an element

Building gold (Z = 79) out of light elements (Z ≤ 19) means adding about 60 protons to each nucleus. Each added charged particle must climb the electric (Coulomb) repulsion of the nucleus, a barrier of order MeV to tens of MeV [memory, high]. Photons and electrons of a few MeV knock nucleons out of nuclei; they do not add them [memory, high]. By contrast, raising the percentage of an element in a sample needs no nuclear physics at all: throw away the rest. If the amount of uranium is fixed, going from concentration c₀ to c₁ means only the fraction c₀/c₁ of the mass survives. That is what ordinary ore milling does.

---

4. Where Moray departs

The exact point of departure is the sentence "the oscillations can continue forever because of the actions of the universe." Everything before it is a correct textbook derivation of a decaying circuit. The departure is a missing term. A real source would appear as V(t) in eq. (4). Once it is written down, its power has to be counted, and Section 5 counts it. Moray's own 1960 statement that the first stage is "excited externally" makes the gap sharper: a primed circuit with no V(t) runs down.

To be fair, Moray says outright that his device is not perpetual motion and that it conserves energy: it takes energy from an outside source [read]. That is a legitimate position. A solar cell or a crystal radio also takes free ambient energy. The problem is not the principle. It is the size of the source.

---

5. Worked example: can any known source feed the box, and can a hidden battery?

All numbers are reproduced by example.py [calc]. Run it with .venv/bin/python review/Morray-TH/example.py

Step 1: Moray's own Appendix IV

Put P = 50,000 W and A = 1 m² into eq. (1):

ρ = P/(cA) = 50,000 / (2.998×10⁸ × 1) = 1.67×10⁻⁴ J/m³ = 167 μJ/m³.

The book gets 166 μJ/m³ and notes that this equals ½CV² for a 3.3 μF capacitor at 10 V: ½ × 3.3×10⁻⁶ × 10² = 1.65×10⁻⁴ J. The arithmetic is correct. The appendix's mistake is the word "small". Small compared with what?

Step 2: Compare with what is actually there

• Cosmic rays: ρ = 1 eV/cm³ = 1.602×10⁻¹⁹ J / 10⁻⁶ m³ = 1.60×10⁻¹³ J/m³.
• Even all moving straight at the plate: flux ρc = 4.8×10⁻⁵ W/m². Shortfall 50,000 / 4.8×10⁻⁵ ≈ 1.0×10⁹.
• Arriving "from every direction", as the book itself says [read]: flux ρc/4 = 1.2×10⁻⁵ W/m². Shortfall ≈ 4.2×10⁹.
• Full sunlight: 1361 W/m², so 50 kW per m² would be 36.7 times full sunlight. The book also claims night operation.

So the one quantitative argument the book offers shows that its source would need to be about a billion times denser than the measured cosmic-ray background.

Step 3: Does resonance rescue it?

Take a 1 MHz tuned circuit: L = 100 μH, C = 253 pF, R = 5 Ω, driven by V₀ = 1 V. - f₀ = 1/(2π√(LC)) = 1.00 MHz. - Q = √(L/C)/R = √(100×10⁻⁶ / 253×10⁻¹²)/5 = 628.7/5 ≈ 126. - Capacitor voltage at resonance: Q·V₀ ≈ 126 V. That is impressive, and an observer with a spark gap or neon lamp would see it. - Average power: V₀²/(2R) = 1/(2×5) = 0.1 W. It depends only on the source and R.

Without the source, the book's own formula applies. Successive peaks shrink by e^(−πR/(2Lα')) = 0.9876 per half cycle, so after 1000 half cycles (half a millisecond) only 4×10⁻⁶ of the charge is left. "Continue forever" needs a V(t), and Step 2 limits V(t).

Step 4: The best case with a real radio field

Moray's box had a real antenna, so take an ordinary EM wave at its most favourable: - At f = 1 MHz (the AM broadcast band), λ = c/f = 300 m. The ideal aperture for gain 1 is A_eff = λ²/(4π) = 300²/12.57 ≈ 7,150 m². That is generous, because a short wire antenna with ground losses gets far less in practice. - A strong field near a broadcast station, E = 1 V/m: S = 1²/377 = 2.65 mW/m². - Perfect collection: 2.65×10⁻³ × 7,150 ≈ 19 W. - To light the load Fletcher recalled in his 1979 affidavit, about 1.4 kW (twelve 75 W lamps plus a 500 W iron [read]; the 1928 letters give three 100 W lamps and a 575 W iron, about 0.9 kW [read: 5th ed. pp.43-46]), the field would have to be E = √(1400 × 377 / 7150) ≈ 8.6 V/m everywhere around the box. - Correction accepted by the arbiter (2026-10-04). A resonant antenna does better than the gain-1 figure. A half-wave dipole at 1 MHz has A_eff = 1.64 × 300²/(4π) ≈ 1.2×10⁴ m², so 1.4 kW needs about 6.7 V/m. Go lower in frequency and the aperture grows as λ²: at E = 1 V/m, 1.4 kW needs A_eff = 1400/(2.65×10⁻³) ≈ 5.3×10⁵ m², which a dipole reaches only at λ ≥ about 2,000 m, i.e. below about 150 kHz [calc]. So the best case is about 1 V/m, a field found only near a transmitter. In an ordinary distant-station field of 1 mV/m the 1 MHz dipole gets about 31 μW [calc]. Fields of 1 V/m are easy to measure, and none was reported at the canyon, Strawberry or desert sites. The record does not say how far the 1928 house or the Dugway site were from a transmitter, which is the one gap in this argument.

And Appendix III's 3,979 W/m² came from static fields, which carry no flow. A real wave with E = 100 V/m carries 100²/377 = 26.5 W/m².

Step 4b: Tesla's collector on atmospheric electricity (added 2026-10-04)

Tesla's 1901 patent uses the same layout as Moray's box: a raised plate or wire, a ground, and a condenser between them. Here the source is the fair-weather field, not a wave, so the power is limited by the tiny current the air conducts (Section 2). - A 27 m antenna in a 100 V/m field sits at V ≈ 2.7 kV. - Current density J ≈ 2×10⁻¹² A/m². Power per square metre of collecting area: J × V = 2×10⁻¹² × 2,700 = 5.4×10⁻⁹ W/m² [calc]. - The three reviewers made different assumptions: from 2×10⁻¹⁰ W/m² (the contrarian, using only 100 V) to 8.4×10⁻⁹ W/m² (the mathematician, from a global-current figure) [arbiter recomputation]. - So 1 kW needs between 1.2×10¹¹ and 5×10¹² m² of collector (the larger is about half the area of the United States), and 1,000 m² gives about 5×10⁻⁶ W. Every version is short of kilowatts by a factor of 10⁷ or more.

The arbiter adds atmospheric-electric collection to the known sources that are Refuted by calculation for kilowatts. Tesla's own mechanism (charged particles from the sun) is weaker still at the ground: the mathematician bounds the solar wind at about 3 W for 10,000 m² even in space, and it does not reach the ground. Bearden's later reading of the "sea of energy" as vacuum (zero-point) energy gives no frequency, flux or detector, so the arbiter rules it Not testable as stated; running 29 stages in parallel adds no energy, because each stage is limited by the flux through its own aperture.

Step 5: The proponents' best facts (be fair)

The record now read in full gives three runs worth checking against a hidden battery. All wattages are nameplate sums [read, as reproduced in the sons' and Moray's books].

So if the lamps drew their rated power and if nothing came in on the leads, a 1920s-30s battery in the box does not explain runs (b) and (c). This is the strongest case in the file for Moray, and you should feel its force.

Now the limits, each from the documents: - Nameplate is not delivered power (Section 3.5). In 1937 the bulbs "burned cold except each had a hot spot about the size of a dime" [read: 5th ed. p.112]; other witnesses saw light "softer … more like daylight" [read: p.133]. Delivered power was probably well below the labels, which weakens both the "kilowatts" claim and the battery arithmetic for the short large-load tests. - The leads were never metered while running. Jensen's 1938 meters on the antenna and ground read "electrically dead", but the 4th edition places that reading before start-up [read: 4th ed. p.132], so it shows only that the idle leads were not live. - The ground path. In 1928 the ground was the basement water pipe of a mains-wired house [read: 5th ed. p.50]. Such pipes are normally bonded to the electrical service neutral [memory, high], so pulling the main switch did not remove that path. - Priming. Every start needed outside excitation (Section 4). A store of charge can explain minutes; it cannot explain 158 hours. Numbers (contrarian, re-done by the arbiter, 2026-10-04 [calc, example.py Step 5c]): a capacitor C at voltage V stores ½CV², so it needs C = 2E/V². The 1928 run (115 W × 157.9 h = 6.5×10⁷ J) needs 130 F at 1 kV; a 4 kW load for 10 minutes (2.4×10⁶ J) needs 0.2 F at 5 kV; the Strawberry run (9×10⁴ J) needs 1.8 mF at 10 kV, which is tens of kilograms of oil-filled capacitors [memory, medium]. So priming can cover start-up and the shortest runs at most. That narrows what is left to a feed through the leads (including the water-pipe ground), a hidden source, or an unknown one; it excludes none of these. - A hidden lead. The physicist Vern Knudsen, at a 1931 demonstration, suspected a wire in the 90-ft antenna pole [read: 5th ed. pp.123-125]. That is Knudsen's suspicion as the Morays record it; no inspection of the pole is on record either way, and the Morays say Knudsen's switching burned out the valve. Every disconnect test (antenna off, lights off) is predicted equally by a hidden feed.

Step 5b: what the antenna lead carried. On 20 Mar 1938 E. G. Jensen put a 60 W lamp and/or a 0.025 μF capacitor in series with the antenna lead. The output lamp still lit; the 60 W lamp never did [read: 5th ed. p.127]. - X_C = 1/(2πfC) for 0.025 μF is 106,000 Ω at 60 Hz, 64 Ω at 100 kHz and 6.4 Ω at 1 MHz [calc]. So the lead current was high-frequency, not mains 60 Hz. - A 60 W, 120 V lamp needs 0.5 A to light. It stayed dark, so the lead current was well below that, say under 0.1 A. Carrying 100 W at 0.1 A needs P/I = 1,000 V or more on the lead [calc]. - A high-voltage, low-current, high-frequency lead is what an oscillator produces whatever feeds it, a cosmic source, a primed store or a hidden supply. The test is real information, but it does not pick between the hypotheses. - The REA's Felix Frazer ran a test of the same kind in 1939: twelve sheets of government-supplied window glass in series with the antenna, after which "the device operated as if the glass had not been in series" [read: 5th ed. p.180, the sons' account]. A glass sheet is a capacitor, so the result says the same thing as Jensen's and has the same limit.

Step 6: Could a nuclear process be the hidden source? (added 2026-10-08)

The follow-up questions asked whether radioactivity, a silver-to-cadmium change or neutrinos could be the unseen input. None of these is a Moray claim about the 1925-38 box, but each is a candidate "hidden source", so count it [calc, example.py Step 7]. - Silver to cadmium by resonance. A proton must climb about 9.8 MeV to enter a silver nucleus. A 1 GHz radio photon carries 4×10⁻⁶ eV, visible light about 2 eV, and the hottest collapsing bubble in sonochemistry has kT ≈ 1.7 eV (k is Boltzmann's constant, T the temperature, here 20,000 K). The best of these is about 5.7 million times too small, and a stronger drive only adds more of the same small quanta. The real routes need a neutron (¹⁰⁹Ag + n, then the ¹¹⁰Ag beta decays in 24.6 s to ¹¹⁰Cd) or an MeV particle source, which the booklet never describes. The mathematician adds that the Preston passage the booklet borrows is a free damped discharge (Section 3.3), which cannot grow without a drive. Converting 1 g of silver at 1 MeV per atom is 8.9×10⁸ J, about 250 kWh. - Beta decay as a power supply. At the full 2.89 MeV per ¹¹⁰Ag decay, 1 kW needs 2.16×10¹⁵ decays/s = 58,000 Ci. That is a floor: at the mean beta energy of about 1 MeV it is about 1.7×10⁵ Ci (the arbiter states both). The 157.9 h, 110 W run of 1928 would need at least 1.35×10²⁰ decays, only about 25 mg of silver (unweighable; the arbiter gives 24 mg), but at least 6,400 Ci while it ran, a radiation field comparable, by the physicist's estimate, to a sizeable radiotherapy source [memory, medium]. Electrical output cannot exceed the nuclear power released, and none of the witnesses or later sources reports a radiation effect at the device. - The real thing at small scale. Moseley's 1913 radium cell reached 150 kV at 0.01 nA (both figures from a secondary source, Wikipedia's "Atomic battery"; his paper has not been read [web https://en.wikipedia.org/wiki/Atomic_battery]): 150×10³ × 10⁻¹¹ = 1.5 μW, about 7×10⁸ times short of a kilowatt. So direct conversion is real at microwatts and refuted at kilowatts. - Neutrinos. The Sun's neutrinos carry about 30 W/m² through everything [memory, high], but a metre of lead stops only about 10⁻¹⁹ to 2×10⁻¹⁷ of them (two estimates), so capture is short by 10¹⁷ or more. The arbiter adopts the range; the conclusion does not depend on which number. - The Foster pellet flash (1990). A phosphor can store about 10¹⁹ filled traps per cm³ at 3 eV each, about 4.8 J/cm³. A 2,000 A press-heater supply at an assumed 10 V is 20 kW, so one second of it is about 4,000 times that store. Breakers and a transformer blowing therefore fit a short circuit through a hot conductive pellet better than a release of stored energy; but the 10 V is an assumption and no one measured, so the arithmetic favours the short without deciding it.

Conclusion of the example

Known radiation falls short by factors from about 10⁷ (fair-weather atmospheric electricity) to billions (cosmic rays, against Appendix IV's 50 kW). Ordinary radio needs a field of about 1 V/m even for a resonant antenna at low frequency, which is found only near a transmitter. Resonance raises voltage, not power, so it cannot close the gap. (2026-10-08) Nuclear candidates do no better: resonance cannot reach the MeV an element change needs, and beta decay at a kilowatt needs tens of thousands of curies and a radiation field nobody reported. A battery hidden in the box is strongly disfavoured for 1928 and 1938, if the lamp labels are near the real load, and a charged condenser can cover only start-up and the shortest runs. What is left is an unmetered input path (hidden wire, ground path), a much smaller real load than the labels suggest, or something unknown. Only a power measurement on the leads could tell these apart, and none was made.

---

6. Evidence table

The gold energy budget (corrected 2026-10-03). contrarian.md once gave "125 J per gram of residue", and an earlier version of arbiter.md used that figure to withdraw the energy-budget argument; arbiter.md now carries the corrected figure, and the contrarian's 2026-10-04 recheck restores it too, so all the files agree. The 125 J figure is wrong by a factor of a million: 8 MeV is 8×10⁶ eV, and the 10⁶ was dropped. Step by step [computed]: 929.76 troy oz per short ton × 31.1035 g per troy oz ÷ 907,185 g per short ton = 0.0319, so the residue is 3.19% gold; 0.0319 g ÷ 196.97 g/mol × 6.022×10²³ atoms/mol = 9.75×10¹⁹ gold atoms per gram; × 8 MeV × 1.602×10⁻¹³ J/MeV = 1.25×10⁸ J per gram (35 kWh per gram), the same as physicist.md's 1.25×10¹¹ J/kg [calc, example.py Step 6]. The Institute's own reported doses were 1/2 megarad in the Aug 1963 control test, 0.16-4 megarad in the Varian series and up to 20 megarad at Electronized Chemicals [read: Rex pp.7, 13-14]; the dose of the 929.76 oz/ton run itself is not stated ("about one minute" of photons [read: Rex p.3]). At 1 rad = 0.01 J/kg [memory, high], 1 megarad is 10 J per gram, so those doses are 1.6-200 J per gram irradiated: short of 1.25×10⁸ J/g by a factor of about 6×10⁵ to 8×10⁷ [computed]. So the energy budget and the nuclear argument both rule out a nuclear origin at that yield. The verdict, Refuted by calculation, is unchanged; it has two independent supports, not one.

Silences in the record (what the gaps might mean)

A gap where the record ought to continue is itself something to explain. For each gap below: what the record says, the competing explanations with the evidence for and against, and the record that would decide. None of these gaps changes a physics verdict; they bear on Claims 1, 4, 5 and 6. Silences B and C and the replications note were updated, and Silence F added, on 2026-10-04. Silences A and B were extended, and Silence G added, on 2026-10-08.

A. Where did the device go? The 1939 device was smashed (by Frazer in the sons' account; by Moray himself in Frazer's counter-claim, which also reaches us through the sons) [read: 5th ed. pp.180-181]. In 1942 Moray rebuilt a smaller unit with "the remaining bit of what was known as the 'Swedish Stone'" and drove it to Idaho, Wyoming and Denver in 1943 [read: 5th ed. pp.187-188]. Then it "burned out during a test", and he had already dismantled the detector "out of fear of compromising its secret" [read: 5th ed. p.189]. Yet demonstrations are recorded later: about a dozen witnesses in 1957 (digest only) and Eyring in 1963 (the AEC wrote in 1970 that he "has seen Mr. Moray's demonstration") [read: AEC fragment]. That conflict is unexplained. After Moray died in 1974, Perreault says, tubes and a notebook passed from his widow to Stan Mahurin, and Perreault later bought the tubes [read: Nuenergy p.12]. No inventory of what remained exists in the record.

Would decide: the 1974 Salt Lake County probate file (estate property); the family archive and the Mahurin notebook; a laboratory test (gamma spectroscopy, X-ray, composition) of Perreault's tubes; the 1970 Hagmann case file (Salt Lake, case 195569); statements from the 1957 and 1963 witnesses; the 1957 clearance-hearing record; (2026-10-04) the Ottawa Bray fonds letters of 1928-57 (Schiess's file is described as holding "witness accounts of the machine in use" and drawings), a located record not yet read. Added 2026-10-08. (1) A second handling of detector material is reported: Foster's 1990 account of a contractor pressing radium-doped "Swedish Stone" pellets (one participant's account in an advocate's compilation; Claim 4). If it happened, detector material was still being handled 16 years after Moray's death; it bears on what remained, not on the demonstrations. (2) The deciding test for the tubes is sharper: gamma and alpha spectroscopy of a surviving cathode (the 46.5 keV gamma line of Pb-210, the alpha of its granddaughter Po-210, the gamma lines of the Ra-226 daughters) plus a ZnS:Ag scintillation screen (zinc sulphide with a trace of silver, which flashes where an alpha strikes it). One measurement tests the radon-plating route, Perreault's "no radioactive material" and any silver-neutron route (¹⁰⁸ᵐAg, 439-year half-life, which decays mainly by electron capture to palladium, not cadmium [web https://en.wikipedia.org/wiki/Silver-108m]).

B. The 74-page FBI file. The FBI's FOIA log for 2000 lists request 917273, "MORAY, T HENRY", 74 pages reviewed, 72 released, "Granted in Part" [web https://archive.org/details/foia-log-2000-2002]. It is not on the FBI's public Vault, and the folder holds only the 1951 Yakovlev interview. Explanations: (1) routine: the Vault posts only a selection of released files [memory, medium]; (2) ordinary contents: the 1951 interview, the FBI's 1942 questions about why he wanted Gustave LeBon's The Evolution of Matter [read: 5th ed. digest p.191], and the 1957 clearance would fill a file of this size; (3) a record of the 1940 shooting, which Moray promised to report to the FBI. Because the family says one attacker was an FBI man, this is an accused party's own file, but it would still show whether a report was made. Would decide: the file itself, requested as previously released records under 917273. Sharpened 2026-10-04. The Investigator found more of the FBI's logs: request 917273 was received 25 Feb 2000 and closed 17 Aug 2000 (74 pages reviewed, 72 released); a second 2000 request was closed for lack of a notarised consent; and a 2011 request for "MORAY, THOMAS HENRY DR." (1178194-002), received 1 Dec 2011, was reopened on 8 Feb 2013 and closed three days later, on 11 Feb 2013, with 0 pages, "No Records" [read: research/files/FBI-FOIA-log-2013Q1_Moray-1178194-002_excerpt.txt] [web https://archive.org/download/UnsortedFBIDocuments/foia-log-1-1-2013-to-3-31-2013-pdf_djvu.txt]. A CIA request of 22 Jun 2000 (F-2000-01342) has no outcome found. The 72 pages are not on the FBI Vault, MuckRock, governmentattic or theblackvault. So the gap is now sharper: a file released in 2000 and "No Records" in 2013. The arbiter lists five explanations: (1) destroyed or transferred under a retention schedule after 2000; (2) filed under a different name form or index; (3) the 2011 request ("-002") had a narrower or follow-on scope (for: the three-day reopen-to-close interval fits a quick re-search better than a fresh review of a 74-page file; against: the log has no scope field, so what was searched is not stated); (4) the file was transferred to NARA; (5) reclassification or withholding. The physicist argues against (1) and (2) from FBI practice, but that is [memory, medium] and unchecked, so it weighs little. Nothing supports (5) and nothing excludes it. No physics verdict depends on this. Would decide: the 2000 release itself, citing 917273; NARA for any transfer. (2026-10-08) If the family's statement that the FBI asked about Le Bon's book in 1942 is right, that enquiry should appear in these pages; until then it is family-only.

C. Missing reports and hearings. - REA reports (Crim, Wood, Frazer). Craig wrote that they went into the Administrator's files and are probably "lost or destroyed" because the REA moved to St. Louis in 1941 [read: 5th ed. p.119]. Explanations: wartime loss (Craig's explanation); deliberate removal (fits the family's account; no evidence found); they survive at the National Archives (NARA, Record Group 221) but nobody has searched (the catalogue was blocked to the Investigator). Would decide: RG 221, 1938-41. - The 1944 Senate REA hearing. Located 2026-10-04 and partly read (investigators' report pp.1344-1345; Slattery interview pp.1370-1376) [web https://babel.hathitrust.org/cgi/pt?id=umn.31951d03587118y]. It dates the REA episode 1939-41, with testimony in 1943-44, not the sons' "1942"; the sons say Moray was never called [read: 5th ed. p.184]. In it Slattery, who says he ended Moray's employment after Senator King's remark, also agreed that Moray's claims and those of the convicted "Count" Tomadelli were "pretty similar" and dates his Tomadelli memo to 23 Nov 1942, when he thought Moray "still with us"; that conflicts with the payroll end of 28 Feb 1941, and it is an interested party's opinion, not a test [read: Slattery interview pp.1372-1373, saved copy]. The investigators say fuller REA project files are in "a separate report" not seen. The finding aid for REA Administrator Carmody's papers (FDR Library) lists REA chronological files (boxes 102-103), card indexes (91-94) and "Senate Hearings [1944]" (box 99), without naming Moray. So Craig's "lost or destroyed in the 1941 move" is no longer the only explanation: a Senate staff had "voluminous information", and the Carmody boxes exist. Would decide: the separate report, Carmody boxes 91-94, 99, 102-103, NARA RG 221 and RG 46. - The 1938 Securities Commissioner's report, Hayes's 1932 Reno testimony, the 1940 USDA order: none in hand. Not yet requested, lost, or left out by the family's editors (who printed favourable statements but not the Commissioner's report). Would decide: Utah State Archives; Washoe County Clerk; NARA. - The patent applications. In 1954 the Patent Office told Moray they "had been destroyed because the office was under the impression that the applications were abandoned"; the sons say this was wrong and that they "disappeared" [read: 5th ed. p.170]. Routine disposal (the Office's explanation), misfiling or removal. Would decide: NARA RG 241 and the file index for Serial 550,611.

D. No police or newspaper record of the 1940 shooting. The account itself predicts this: the doctor made no report because Moray promised to tell the FBI, and "as far as I know, no police report was ever made" [read: 5th ed. p.186]. The Utah newspaper search found nothing, but it was a word search over poor OCR [web research/2026-10-03-utah-digital-newspapers.md]. Since the silence is what the account predicts, it does not count against it. Would decide: the FBI file (B); Salt Lake police records for March 1940; the papers of S. E. Bringhurst.

E. What happened to the gold work and the 1968 Air Force proposal? The Research Institute, Inc. (John E. Moray and colleagues) rented outside accelerators, ran a 1964 pilot plant on about 800 lb of ore, spoke at a mining conference in 1964 and 1965, and in 1979 had its consultant project a plant making about 238,000 oz of gold a year, needing about $6.4 million to break even [read: Rex pp.5-7, 12-13, 17-20]. The last test data in the folder are from Aug-Oct 1970 [read: Rex p.19]. The 1968 proposal to Air Force Systems Command says working models were "demonstrated to competent observers" [read: gov-Airforce]; no reply is in the folder. No journal paper, no patent assigned to the Institute [web research/2026-10-03-google-patents.md], no named assayer or certificate, no plant, no gold sale and no outside replication has been found. A process worth millions a year should leave such records, so the silence needs explaining.

Would decide: the Utah corporate file for Research Institute, Inc.; the assayer's certificates; accelerator logs or invoices; the 1964-65 mining-conference proceedings; Hooper's papers; the Air Force file and reply for the 1968 proposal (NARA).

F. The Army's 1936 report (added 2026-10-04). Four NARA War Department index sheets (Record Group 407) show a report by Brig. Gen. Shinkle, Army Ordnance, dated 8 Dec 1936, filed as "sound detector"; a Feb 1939 enquiry through Sen. Thomas calls it a "Radiant Energy device" (the enquirer's wording). The report itself has not been found. Explanations: filed under decimal file 413.68 or the 095 Steffensen name file and not yet requested (most likely); misfiled; destroyed under a schedule; withheld. Nothing points to the last two. Would decide: the file itself.

G. Moray's own experiments with the materials his books describe (added 2026-10-08). The books print the physics of sulphide screens, thorium deposits, amalgam photocurrents and resonance transmutation, yet the record holds no Moray experiment with any of them and no silver or cadmium assay; most of those pages are Le Bon, Glasser and Preston (Source gap).

Silence in other sources is not contradiction. Would decide: Moray's notebooks and the Mahurin material; the Ottawa Bray fonds letters; the 1943 and 1956 printings of Radiant Energy.

Replications (added 2026-10-04). None with metered input and output is on record after 90 years; what exists is hobby videos (2016-2025), an online "blueprint" and a 2025 "specification". Explanations: the detector was never published (Moray's own account); the claim is false; no capable builder has tried with the right part. The record cannot choose between them. It is a fact to weigh, not a verdict. (2026-10-08) The 1906 repetitions of Le Bon's experiments (Claim 5f) are not a replication of anything of Moray's; no Moray result has been repeated.

---

7. What a learner should take away

Solid. - Ambient radiation is real and can be harvested. Crystal radios ran on nothing but the incoming signal, and Moray's antenna would have picked up some microwatts to milliwatts. - Mineral rectifiers were real 1920s technology. - Moray framed his idea correctly: an open system drawing on an outside source, not perpetual motion. - Appendix IV's formula P = ρcA is correct and is exactly the right tool. - The witness record is larger and better than a quick reading suggests, and it is evidence about the device, not only about the books. About 30 named people, including Harvey Fletcher of Bell Telephone Laboratories, a state Securities Commissioner, REA engineers (Crim, as recalled by former REA deputy administrator Craig: tests "entirely satisfactory as far as performance [went]" [read: 5th ed. p.119]) and a physicist (Knudsen), saw lamps light. REA engineer Wood, not employed by Moray (though the REA, then under political attack, was paying Moray), reported him "essentially honest, unable to originate or perpetuate a hoax or fraud" (quoted in 1944 Senate testimony, made public by Rep. Halleck, an REA opponent who quoted it to ridicule the agency) [web https://archive.org/details/per_chicago-daily-tribune_1944-12-11_103_296]. That weighs against deliberate fraud; it contains no measurement, so it does not rule out priming, a ground path or observer error. Several ran simple checks. The 1928 and 1938 energy totals, if the loads were real, rule out a battery in the box. Fletcher wrote afterwards, "I am just as puzzled as ever" [read: 5th ed. p.75]. - (2026-10-04) Others did reach for the same ideas, and some of them were real. Tesla patented the antenna-ground-condenser layout in 1901, and Losev and Lilienfeld built crystal amplifiers in the 1920s (with a bias battery). That makes Moray's architecture part of a real lineage. It also means his priority claim is not unique, and a valve that gives gain without a supply is still untested. - (2026-10-08) Several effects the books describe are real: the photoelectric effect (Le Bon's amalgam discharge), Rutherford's concentration of thorium's active deposit on a charged wire, radon daughters plating onto a negative electrode, phosphors that store charge and release it under infrared, and radioisotope cells that give microwatts at high voltage (Moseley 1913). A radium-doped phosphor-type stone in the valve is consistent with known physics and untested; it would be a slow, microwatt-scale source. - (2026-10-08) Le Bon's case is a model of how an anomalous claim can be handled: within a year a leading chemist (Ramsay) repeated it, a competing group (Millikan and Winchester) reinterpreted it, and the answer was printed. Weber's neutrino claim likewise got four neutral tests. Moray's claims got no such tests, largely because no outsider was given the device (Claim 6). That is a comparison, not evidence about Moray.

Wrong, and why it is instructive. - Confusing "small number" with "small compared with what exists". 166 μJ/m³ sounds tiny, but it is about a billion times the cosmic-ray density. Whenever a source is proposed, look up its measured density and do eq. (1). - Thinking resonance makes energy. A tuned circuit at Q = 126 shows 126 V from a 1 V drive. That looks like gain, but the power is unchanged. Always write the source term V(t) and compute the average power. - Taking static-field "power". E × H from static fields gives a number with units of W/m², but nothing flows (Appendix III). - Many witnessed lamp counts are still not a metered test. Thirty named witnesses watching lamps is strong evidence that lamps lit. It is weak evidence about how many watts flowed, and no evidence about where they came from. Every disconnect test is equally explained by a hidden feed, and bulbs that "burned cold" were not drawing their label. One recording wattmeter on each lead would have outweighed all the testimony. - Read the quote, not the summary of it (2026-10-04). Wood's sentence, "he has ... stumbled on to something that he himself cannot understand", is often read as an engineer saying he saw something he could not explain. Grammatically it is Moray who "cannot understand". It is an opinion that Moray was honest, not a report of an observation. Likewise "two months" (the family: tests before the smashing) and "2 years" (Senate staff: the whole engagement) look like a conflict but measure different spans. - Check retold numbers against the primary record. Bearden's "over fifty kilowatts" for the 157-h run is about 7,900 kWh against the record's 17-19 kWh. The 1934 "100 lights for 154 hours" could be either drift or the same test with small lamps; without the lamp wattage you cannot tell [calc]. - Wrong physics is not the same as self-contradiction. The 1946 treatise says alpha, beta and gamma rays all pass through the body and half an inch of lead, and that no shielding is needed. Alpha and beta cannot (refuted by known ranges); gamma can, which fits a radium source and makes "no shielding" a hazard the treatise never quantifies. The arbiter labels the alpha/beta part Refuted by calculation and the safety part Not testable as stated, not "Internally inconsistent". - Who wrote it? (2026-10-08) The radioactivity pages of Radiant Energy and Isotopes read like Moray's research, but they are Le Bon (1905), Glasser (1944) and Preston (1928) with credits and footnotes removed. Before citing a passage as an inventor's experiment, check whether it is quoted from someone else; the owner's question about "the researches of Rutherford ... taken up again" (Q01) turned out to be Le Bon's call to repeat Rutherford's alpha-velocity measurements, not a Moray research programme. Look at the direction of the edits too: what was removed ("I think", credits, footnotes, one rate "1 to 2 per minute" replaced by "a definite discharge") made checking harder, and what was added carries no data. - Resonance cannot buy MeV (2026-10-08). A quantum's energy is hf. Making a tuned drive "violent enough" adds more quanta, not bigger ones, so no drive at eV energies can push a proton over a 9.8 MeV barrier. The Isotopes booklet's own p.20 states the premise (energy proportional to frequency) that defeats its p.23 conclusion. - A current is not a particle count (2026-10-08). Le Bon compared ionisation currents and inferred equal numbers of particles, but an alpha makes about 10⁵ ion pairs and a photoelectron one charge. Always ask what each counted event carries. - Concentrating is not creating. Raising uranium from 0.05% to 22% says nothing about making uranium until you know the masses before and after (Section 3.6). The same goes for gold assays without blanks. - The decisive tests were offered and not completed, by both sides. Moray withheld the detector every time, declined Eyring's wattage test, and (by Fletcher's account, which Moray denied) did not come to the New York test. (Eyring's offer and Moray's refusal, "the amperage would burn out his device", are documented in John Moray's account, no second source found yet [read: 5th ed. p.74].) Critics in turn did not follow up: Bell Labs wrote in 1937 that even a small chance warranted a "complete investigation" and made none; Eyring set a test condition and never came back; no critic insisted on a meter at the leads. The REA did investigate for two years, but its reports are missing. (2026-10-04) The printed Senate record adds both sides: physicist Orin Tugman wrote in May 1940 that in 12 years of asking he was never shown a demonstration (his account, in one printed source; he had stakes of his own, and it reports refusals, not a failed test); and the reason the REA stopped paying is disputed between two interested parties, with no test result either way. This standoff, not a failed test and not a proven suppression, is the documented reason the physics question stayed open. - Silence is a question, not an answer. A device that vanishes, a 74-page FBI file nobody has read and lost agency reports are anomalies to explain, not proof of anything either way. List the explanations, say what evidence favours each, and find the record that decides (see "Silences in the record"). - Weigh evidence on both sides on the same terms. The violence account is one detailed source with nothing contradicting it; Hayes's failed rebuild, which counts against Moray, is also one source. Court records cut both ways too: a 1942 judgment against Moray for not delivering stock to an investor, and a 1970 suit by Moray's institute against Hagmann, whose praise the 5th ed. prints [web https://newspapers.lib.utah.edu/details?id=25602932; web https://newspapers.lib.utah.edu/details?id=29749860; read: 5th ed. p.279].

Genuinely open (historical, not physical). - What powered the 1928 and 1938 runs? Did power come in on the antenna, the pole (Knudsen's suspicion) or the water-pipe ground? How much power did the lamps really draw? - Did the priming store enough energy to matter for the short demonstrations? (2026-10-04: by the condenser arithmetic in Step 5, at most for start-up and the shortest runs.) - Who cancelled the New York test, and is there a telegram? What exactly did Hayes say at Reno? - Why did the claimed performance fall about 600-1,000 times between the 1930s and 1968? - What happened in the digest-only later episodes: the 1943 rebuild driven through Idaho, Wyoming and Colorado with a failed Denver demonstration, and demonstrations in June 1957 before about a dozen witnesses [read: 5th ed. digests pp.191-193; Polly file]? What was demonstrated in 1957 and 1963, after the sons say the detector material burned out? - Did the 1939 destruction and the 1940 shooting happen as the sons tell it? They are now classed Never properly tested rather than untestable, because the records that would check them can be named: the 74-page FBI file (request 917273), the REA files at NARA, Salt Lake police records. The printed 1944 Senate hearing, now found, is silent on both, but it is a hostile staff's summary, so that silence is weak. Does the FBI file name an agent present on 2 Mar 1940? Is the Felix J. Frazer of the REA the Technocracy economist of that name [web https://archive.org/details/report00ofnationalnatirich]? - (2026-10-04) What did Shinkle see in Dec 1936: a receiver or the lamp unit? What drove Craig's "small motor and fans", and at what frequency? Why did the REA end the engagement (Slattery says a senator's opinion, and agreed Moray's claims were "pretty similar" to the convicted Tomadelli's; the investigators say negative conclusions; neither cites a test)? Do the Ottawa letters quote the witnesses as the family's books do? Why does a 74-page FBI file released in 2000 return "No Records" in 2013? Are the Cosray applicators and the valve tubes the same material, and is it radium? - What remained of the apparatus in 1974, and are Perreault's tubes Moray's? A gamma-spectroscopy test of them would check both the radioactive-dopant account and Perreault's "no radioactive material". - (2026-10-08) Who prepared each printing of Radiant Energy, Sea of Energy and Isotopes, and why are Le Bon's, Rutherford's and Crookes's credits missing? Did Moray run any screen, amalgam, thorium-plate or silver/cadmium experiment (Silence G)? What did his insertion "and sometimes other properties" mean to him? Is Foster's 1990 pressing documented (letter, press log, contractor), and was the flash stored energy or a short? Do surviving tube cathodes carry Pb-210 or Ra-226 daughters? What does Moseley's full 1913 paper say? What did Le Bon's spark-strength ultraviolet do that Ramsay and Spencer's lamps could not? Which printings of Le Bon, Glasser and Preston did Moray's editor use? Does the 74-page FBI file hold a 1942 Le Bon enquiry? - The decisive experiment was never done: a rebuild in a shielded room cut off from mains and water pipes, with recording meters on every lead, calorimetry on the box, an antenna-less run, a dummy-detector blank, a metered priming step, and a neutral examination of the detector. (2026-10-08) Two cheaper additions: gamma and alpha spectroscopy plus a ZnS:Ag screen on a surviving tube cathode and any Cosray applicator, and a radiation survey beside a running device with a blank-controlled assay of electrode and process metals for cadmium in silver (a kilowatt from nuclear conversion needs radiation of at least that power). A blank-controlled repeat of one Le Bon amalgam experiment with spark-strength ultraviolet would settle the part Ramsay and Spencer could not test.

---

8. Check yourself

1. A proposed "background energy" has density 10⁻⁸ J/m³ and arrives from all directions. What is the most power a perfect 2 m² flat collector (one face) could take from it?
2. A series RLC has L = 1 mH, C = 1 nF, R = 10 Ω and is driven at resonance by a 0.5 V (peak) source. Find the resonant frequency, Q, the capacitor voltage and the average power.
3. The 1938 Securities Commission demonstration ran 30 × 150 W lamps and a 1000 W iron for an hour. How much lead-acid battery would that need? Give two reasons why this calculation still does not prove the energy came from the antenna.
4. Moray says his device "conserves energy" and that its first stage is "excited externally". Which single term is missing from the book's Chapter 10 equation, and what would you need to measure to fill it in?
5. An ore at 0.09% U₃O₈ is treated and comes out at 53.4%. Assuming no uranium was created, what fraction of the original mass is left? What one measurement would test the claim that uranium was "upgraded" rather than concentrated?

---

9. Further reading

Mainstream - D. J. Griffiths, Introduction to Electrodynamics, Ch. 8 (Poynting's theorem) and Ch. 9 (EM waves). - Halliday, Resnick and Walker, Fundamentals of Physics: electromagnetic oscillations and AC circuits. - P. Horowitz and W. Hill, The Art of Electronics, Ch. 1 (RLC, reactance, diodes, resonance). - C. A. Balanis, Antenna Theory: effective aperture; summary at [web https://en.wikipedia.org/wiki/Antenna_aperture]. - K. S. Krane, Introductory Nuclear Physics: binding energy, Coulomb barrier, why transmutation needs MeV. - J. R. Taylor, An Introduction to Error Analysis: why a measurement needs a blank and an uncertainty. - R. P. Feynman, The Feynman Lectures on Physics, Vol. II, Ch. 9: atmospheric electricity; summary at [web https://en.wikipedia.org/wiki/Atmospheric_electricity]. - Related work (added 2026-10-04): N. Tesla, US 685,957 (1901), "Apparatus for the utilization of radiant energy" [web https://patents.google.com/patent/US685957A/en]; O. Losev's crystal amplifiers [web https://en.wikipedia.org/wiki/Oleg_Losev]; J. E. Lilienfeld, US 1,745,175 [web https://en.wikipedia.org/wiki/Julius_Edgar_Lilienfeld]; radium and its half-life [web https://en.wikipedia.org/wiki/Radium]; half-value layer [web https://en.wikipedia.org/wiki/Half-value_layer]. - Cosmic-ray energy density: [web https://en.wikipedia.org/wiki/Cosmic_ray]; crystal detectors: [web https://en.wikipedia.org/wiki/Crystal_detector]; Seaborg's 1980 bismuth-to-gold (atom quantities, net energy loss): [web https://en.wikipedia.org/wiki/Nuclear_transmutation].

Moray and his circle (in the source folder) - T. H. Moray, The Sea of Energy in Which the Earth Floats, 4th ed. 1960 (Moray's own last edition): pp.116-197, his history of the demonstrations, the Hayes testimony and the "excited externally" passage. - T. H. Moray, ed. J. E. Moray, The Sea of Energy, 5th ed. 1978: pp.32-133 for the dated demonstrations, sworn statements and the Fletcher/Eyring/Millikan/Foster/Bennett correspondence; Ch. 1, Ch. 10 and Appendices III-IV for the theory. Read the son's commentary as advocacy and the reproduced letters as the evidence. - Moray, 1931 booklet and 1945 pamphlet on Radiant Energy; the folder's Radiant Energy printing labelled "1926", which by its content (atomic pile, plutonium) was compiled after 1945 [read, per qe.md and the arbiter] (2026-10-08: its pp.53-59 and Isotopes pp.4-6 reproduce Le Bon, and Isotopes pp.20-25 follow Glasser and Preston; read them as what Moray read, not as his experiments); 1962 Valley State College speech; 1968 Air Force proposal. - Isotopes in Ionic Recovery of U3O8 (uranium claims); Photo Actinia (Cosray, with the 785-patient tally); US Patent 2,460,707 (1949, Cosray tube) [web https://patents.google.com/patent/US2460707A/en]. - Rex Research compilation on the Research Institute gold work (1963-70), with the Aug 1963 controls. - Perreault, Nuenergy: an alternative explanation (spark-gap stations, radioactive dopants); useful as an exercise in checking an explanation with the same arithmetic. - Patent Office rejection, Serial 550,611, 23 Dec 1932, and the separate "no natural source" action (folder: Patents, gov-Commerce) [read]. - FDA Notices of Judgment, NJ 8137 (Cosray, 1963-64) [read]. - Robert B. Craig's letter on the REA tests (Crim) and the loss of REA files [read: 5th ed. p.119]; the 1951 FBI interview memo in the Yakovlev espionage file [read: gov-FBI screenshot]. - Records outside the family's books, found by the Investigator (see research/ and sites.md): the 11 Dec 1944 Salt Lake Tribune and Chicago Tribune reports of Senate REA testimony (Frazer, Wood) [web https://newspapers.lib.utah.edu/details?id=28911552; web https://archive.org/details/per_chicago-daily-tribune_1944-12-11_103_296]; court notices for Washoe County 1932, Sanford v. Moray 1942 and the 1970 Hagmann suit [web https://newspapers.lib.utah.edu/details?id=28958019; web https://newspapers.lib.utah.edu/details?id=25602932; web https://newspapers.lib.utah.edu/details?id=29749860]; the FBI FOIA log entry for file 917273 [web https://archive.org/details/foia-log-2000-2002]. Still to obtain: the FBI file itself, REA files in NARA RG 221. - Added 2026-10-04: the printed 1944 Senate hearing, pp.1344-1345 and 1370-1376 (REA payroll, Wood, Tugman, Slattery) [web https://babel.hathitrust.org/cgi/pt?id=umn.31951d03587118y]; J. E. and R. R. Moray, A Short Treatise on Alpha, Beta, Gamma Ray Therapy (1946) [web https://archive.org/details/thomas-henry-moray-radiant-energy]; the 2013 FBI log entry ("No Records") [web https://archive.org/download/UnsortedFBIDocuments/foia-log-1-1-2013-to-3-31-2013-pdf_djvu.txt]; T. E. Bearden's vacuum-energy reading (cheniere.org, his own theory; see research/2026-10-04-bearden-cheniere.md). Still to obtain: Shinkle's 1936 report (NARA RG 407), the Carmody Papers boxes at the FDR Library, the University of Ottawa Arthur Bray fonds 30-003 (Smith, Schiess, Betz, Buchli, Hooper files), CIA F-2000-01342, the 1970 Hagmann file, J. Cecil Alter, Utah, the Storied Domain (1932) pp.439-440. - Added 2026-10-08, the sources of the books' physics pages and their reception: G. Le Bon, The Evolution of Matter (1905; Legge transl.), Book IV ch.III and Book VI ch.VII [web https://www.rexresearch.com/lebonmat/lebonmat.htm]; E. Rutherford, Radio-activity (1904), sec.166 [web https://archive.org/details/radioactivity00ruth]; Rutherford's Nobel lecture (1908) on the alpha particle [web https://www.nobelprize.org/prizes/chemistry/1908/rutherford/lecture/]; W. Ramsay and J. F. Spencer, Phil. Mag. 12 (1906), the repetition of Le Bon's experiments; M. J. Nye, Hist. Stud. Phys. Sci. 4:163-195 (1974), on Le Bon's reception (not yet read); H. Kragh on Le Bon's Nobel nomination [web https://arxiv.org/pdf/1208.3114]; O. Glasser (ed.), Medical Physics vol.1 (1944) pp.1354-1355, and T. Preston, The Theory of Light (1928 ed.) p.588 (matched on snippets only). - Added 2026-10-08, for the nuclear alternatives: H. G. J. Moseley, Proc. R. Soc. A 88, 471 (1913) [web https://royalsocietypublishing.org/doi/10.1098/rspa.1913.0045] (abstract only; the 150 kV and 0.01 nA figures are from [web https://en.wikipedia.org/wiki/Atomic_battery]); A. Franklin (2010) on Weber's neutrino claim and its four null tests [web https://www.colorado.edu/physics/sites/default/files/attached-files/gravitywavesandneutrinos.pdf]; Borexino, Nature 562, 505 (2018), the Sun's neutrino luminosity; silver-108m isotope data [web https://en.wikipedia.org/wiki/Silver-108m]; zinc sulphide phosphors [web https://en.wikipedia.org/wiki/Zinc_sulfide]; H. Plauson, US 1,540,998 (1925), photoelectric amalgam collector (in the Nuenergy compilation, pdf p.83); Perreault, Nuenergy (2001) p.30, A. G. Foster's 1990 account. - NASA/CR-2005-213749 s.4.4.5: a literature survey that calls the claim "not scientifically defensible" and says replication is needed [read, per arbiter.md].

---

Answers

1. Isotropic arrival onto one face gives flux ρc/4 = 10⁻⁸ × 3×10⁸ / 4 = 0.75 W/m². Times 2 m² gives 1.5 W at most.
2. f₀ = 1/(2π√(10⁻³ × 10⁻⁹)) = 1/(2π × 10⁻⁶) ≈ 159 kHz. Q = √(L/C)/R = √(10⁶)/10 = 1000/10 = 100. V_C = Q·V₀ = 50 V. P = V₀²/(2R) = 0.25/20 = 12.5 mW.
3. (30 × 150 + 1000) W × 1 h = 5.5 kWh; at 35 Wh/kg that is about 157 kg of lead-acid. Reasons it does not prove an antenna source: (i) the wattages are lamp labels, and bulbs in these demonstrations often "burned cold", so the real load may have been far smaller; (ii) a wire coming in (a hidden lead in the pole, or the ground connection to a mains-bonded water pipe) has no mass limit, and the leads were metered only before start-up. A third: priming charge, though it cannot cover a full hour of kilowatts: 5.5 kWh = 2×10⁷ J at 5 kV needs C = 2E/V² ≈ 1.6 F [calc].
4. The source term V(t) on the right-hand side of L q'' + R q' + q/C = V(t). "Excited externally" supplies an initial charge, not a continuing V(t), so the circuit would run down. You would measure the voltage and current (hence power) entering through the antenna and ground leads, and the energy put in during priming: over a run, the energy in must equal or exceed the energy delivered to the lamps.
5. The uranium mass is fixed, so the surviving fraction is 0.09/53.4 = 0.17% of the original mass (99.8% discarded). The test: weigh the ore before and after and do a chemical (not radiometric) assay of both, so you can check whether the total mass of uranium went up. Splitting the sample for a blind assay and measuring the U-235/U-238 ratio would make it stronger.