Morray-TH (T. Henry Moray, 1892-1974): Arbiter (Step 3, re-run after QE "MAJOR")

Inputs: qe.md, facts.json, claims.md (revised), physicist.md, mathematician.md, contrarian.md (all three re-run after QE), the six deep reads in deep/, and the earlier arbiter.md. Revised 2026-10-03 under the SOURCE HONESTY evidence rules (no suppression of evidence, interested parties, absence of evidence), also using sites.md and research/2026-10-03-*.md; see "Evidence-rule revision (2026-10-03)" at the end. Updated 2026-10-04 for the Investigator's findings and the three reviewers' readings of them: see "Investigator findings (2026-10-04)" (placed before "Silences in the record"); the other updated sections are marked 2026-10-04. Updated 2026-10-08 for the follow-up investigations (Q01, Q02, Q04, Q08, Q09, Q11, Q12) and the three reviewers' readings of them: see "Investigator findings (2026-10-08)" (placed after the 2026-10-04 section); sections updated then are marked 2026-10-08. Source folder (read only): source folder: Morray-TH. Tags: [read] = in the documents, as extracted by the deep readers (page numbers are PDF pages: "5th ed." = 1978 scan, "4th ed." = 1960 scan); [web] = URL given; [memory] = recall, confidence given; [calc] = arithmetic done here; [computed] = arithmetic recomputed in the evidence-rule recheck of 2026-10-03.

Source gap (rewritten after QE)

• The 4th (1960) and 5th (1978) editions of The Sea of Energy now have OCR text and were digested in full. Deep reads cover 5th ed. pp.32-133 and 4th ed. pp.116-197 only. 5th ed. pp.134-279 (Moray Products litigation, patents, REA episode and the 1940 shooting, the 1942-43 rebuild, theory chapters) and 4th ed. pp.1-115 are digest only, so statements from those pages are less certain. The Rex Research, Nuenergy, Isotopes and Photo Actinia documents were deep-read for the pages named in claims.md.
• Every witness letter, affidavit and the 1932 "Brief History" reaches us only through books edited by Moray or his sons. Several were solicited decades later (A. B. Jensen: events 1938, letter dated 22 Dec 1958 in the 4th ed. and 1956 in the 5th, 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, Frazer or Reno-court document is in the folder. That the testimony reaches us through the family is a fact about provenance, to be weighed; it does not make the testimony non-evidence. Outside the family's books, the Investigator (Step I) has now found: the 1944 Senate REA testimony as reported in the press (Moray on the REA payroll Feb 1939-Feb 1941, Frazer's role, the Wood 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 court notice "In re Moray Products Company" [web https://newspapers.lib.utah.edu/details?id=28958019]; the 1942 Sanford v. Moray judgment [web https://newspapers.lib.utah.edu/details?id=25602932]; the 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 memo of an interview with Moray [read: gov-FBI screenshot] and a letter from former REA deputy administrator Robert B. Craig [read: 5th ed. p.119].
• No document records a neutral, metered measurement of the device's input or output power. Every power figure is a nameplate count of lamps and appliances. Moray's own 1960 text says instruments are "of no value" for the valves and are "not giving ABSOLUTE readings" [read: 4th ed. pp.192-194]. Conclusions about what the device did rest on the testimony of named observers: many were associated with Moray, and some were state or REA officials (the Securities Commission inspection, Crim, Craig, Wood). Conclusions about physics rest on [calc], [web] and [memory].
• Handwritten notes (Betz 1954, Wachs 1955, Noffsinger 1984), the AEC and Sen. Bennett fragments are partly illegible. The 1,039-page FDA volume was searched, not read (notice 8137 was read). The FBI, CIA and NRC files in the folder hold one Moray passage each; a separate 74-page FBI file indexed to Moray (released in 2000) is not in the folder and not on the FBI Vault (see "Silences in the record"). Audio, video and photographs were not interpreted. The Uppsala item rests on a 2011 enquiry to Uppsala University (a negative search under "Moray").
• Where a verdict below rests on [memory] alone, it says so. None does for a central verdict.

How the three reviews were reconciled

All three re-runs corrected the same first-round errors: nameplate loads of 1.5-8 kW were recorded 1930-38; instrumented checks existed (none measured power); a test offer and a non-neutral rebuild are on record; the examiner did state a "no natural source" ground; Moray's own per-person figure is 1,193,600 lamps; the gold work had a crude control; the USDA item was a build order. They agree on every refutation by calculation. The remaining differences, and my rulings:

Numbers recomputed here [calc]:

Test classification used (extended after QE)

For each claim I keep these apart: - Tested, failed / tested, passed: a controlled experiment was run and the result recorded. None exists in this folder for any Moray claim. - Never properly tested: testable, but no neutral, controlled, instrumented test on record. - Dismissed without testing: rejected by people who ran no experiment. - Witnessed by named observers, uninstrumented for power (new; renamed in the 2026-10-03 revision from "witnessed by friends"): observed repeatedly by invited observers, many associated with Moray and some state or REA officials, some with crude checks (Yates, Jensen), never metered. - Government investigation, outcome record missing (added 2026-10-03): the REA engagement of 1938-41 (Crim, Wood, Frazer). Crim's report, as remembered by Craig, called the tests "entirely satisfactory as far as performance" went [read: 5th ed. p.119]; Wood's report, as quoted in 1944 Senate testimony, called Moray "essentially honest, unable to originate or perpetuate a hoax or fraud" [web Chicago Tribune 11 Dec 1944]; by the family's account about two months of Frazer's tests ended with Frazer smashing the device [read: 5th ed. p.180]; the printed 1944 Senate record instead describes Frazer as spending "the next 2 years" on the project and does not mention a hammer (corrected 2026-10-04: the two spans are not shown to conflict; see "Investigator findings (2026-10-04)"). None of these reports is in hand. - Test offered, inventor declined or withheld (new): Moray withheld the detector in every recorded examination except possibly 1938 [read: 5th ed. pp.66-68, 117]; he refused Eyring's specific-wattage test ("the amperage would burn out his device"; documented in John Moray's account, no second source found yet) [read: 5th ed. p.74]; the Fletcher-arranged New York test is disputed (above). - Test offered by critics, not carried through (new): Bell Labs (1937) wrote that a small chance warranted a "complete investigation" and made none; Eyring defined a 3-4 mile criterion and never followed up; Fletcher's 1928 antenna-less run was never done by anyone; no critic insisted on a power meter at the leads. - Non-neutral rebuild failed (new): Hayes (Moray's patent agent, who had said in 1929 he saw everything) testified at Reno in 1932 that he and associates tried to build the machine from the patent papers and failed (editors' report, no transcript) [read: 4th ed. pp.169-170]. It fits both "key details withheld" and "not reproducible from the papers". It is not neutral and not a test of the original device. Like the witness accounts, it reaches us in one source (the 4th-ed. editors); it is weighed on the same terms as testimony in Moray's favour. The court file that would hold the testimony is now located (Second Judicial District Court, Washoe County, 1932) [web https://newspapers.lib.utah.edu/details?id=28958019]. - Not testable as stated.

Claim-by-claim

Claim 1: A box connected only to an antenna and a ground lit lamps and heated irons without internal source or heating

• Agree (all three): Never properly tested. Witnessed on about 25 dated occasions, 1925-38, by at least 30 named witnesses, many of them Moray's promoters or associates. Never metered, never replicated.
• What the record contains [read: 5th ed. pp.32-133; 4th ed. pp.116-197]. Nameplate loads: 655 W (Oct 1925); about 0.9 kW (Fletcher, Sep 1928); 0.11 kW for 157.9 h ending at 635 W (1-8 Oct 1928, Hayes sworn); 1.5 kW (Welling and the Mexican Consul, Oct 1930, house meter checked); 2.15 kW (Lovesy, second-hand); about 2.5 kW (Yates, Dec 1930); "over 4,000 W" (Moray's own 1931 memo, Knudsen visit); about 8 kW (Gardner, about 1936); about 4 kW (Todd, 1937); about 5 kW for Pyper (1937; house mains opened, house lit from the box; he did not see inside the larger box); about 5.5 kW for over an hour (A. B. Jensen for the Utah Securities Commissioner, spring 1938); about 5.25 kW desert test 25-30 miles from Johnson's Pass (A. B. Jensen, 1938, no duration, no meters, written 24 years later); about 5.7 kW (Bringhurst, July 1938). The 50 kW figure has no witness, date or measurement. The earlier arbiter's "no document records it ever being observed" is right for 50 kW only; for 4-8 kW the documents do record lamp counts.
• Disagree: the physicist labels the weak form "Established as testimony only"; the other two say Never properly tested. I use "Never properly tested" for the physical claim. The testimony is evidence about the device, not only about the documents: about 30 named people, on about 25 dated occasions, say they saw lamps lit and an iron heated with no visible source, and several say they could not explain it. That evidence is strong that the observed effects occurred as seen. What it cannot settle, because no one metered the leads, is where the energy came from (see "Strongest argument against").
• REA officials' evidence (added 2026-10-03). (a) Robert B. Craig, former REA deputy administrator, in a letter printed in the 5th ed.: "The most important report that I remember is by a Mr. Ben Crim, a senior regional engineer in REA, who went to Salt Lake City to observe certain tests of Dr. Moray at his laboratory. Mr. Crim reported that the tests were entirely satisfactory as far as performance [went]", and "I personally observed on several occasions the operation of Dr. Moray's device ... I can testify to the results, but not to [the explanation; OCR breaks off]" [read: 5th ed. p.119]. (b) REA engineer Franklin P. Wood's report, quoted from 1944 Senate committee testimony: "Moray is unfathomable. He is an unmitigated ass, crackpot, nitwit, religious fanatic, essentially honest, unable to originate or perpetuate a hoax or fraud. In some inexplicable way he has, I believe, stumbled on to something that he himself cannot understand" [web https://archive.org/details/per_chicago-daily-tribune_1944-12-11_103_296]. The REA kept Moray on the payroll after Wood's report (Halleck, same source). Stakes: the REA paid Moray $25 a day and in 1944 was under political attack; Rep. Halleck, who made the testimony public, was attacking the REA and quoted the report to ridicule it ("nut house schemes"). Weight: these are the nearest things in the record to assessments by engineers not employed by Moray. They weigh against deliberate fraud by Moray (Wood: "unable to ... perpetuate a hoax") but contain no measurement, so they do not separate a real source from priming, a ground path or observer error.
• Strongest argument for the proponents (contrarian, accepted): the 1928 run. 17-19 kWh in two small boxes rules out lead-acid if the load was real. Several trained observers (Fletcher, Hayes, Yates, Pyper, a state-commissioned "battery expert") said they could not explain it. Fletcher: "I am just as puzzled as ever" [read: 5th ed. p.75]. He named two hypotheses, an external source or battery-type action kept alive by atmospheric energy, and concluded neither.
• Strongest argument against (contrarian and mathematician, accepted): (i) every wattage is a nameplate sum, and bulbs "burned cold" with a dime-sized hot spot, so delivered power was probably far below nameplate; (ii) every disconnect test (antenna off, light off) is equally predicted by a hidden feed (Knudsen, 1931, suspected a wire in the 90-ft pole; this is Knudsen's suspicion, recorded by the Morays, with no inspection on record either way, and the Morays blame Knudsen's switching for burning out the valve); (iii) the 1928 ground was a basement water pipe in a mains-wired house, normally bonded to the service neutral [memory, high], so pulling the main switch does not remove that path; (iv) every start-up needed physical priming (magnet stroking 4-10 min, a shock at the terminals, rubbed discs, a hand electrostatic plate) and large condensers were present; Moray's own text says the first stage is "excited by an external energy source" [read: 4th ed. p.116]; (v) the one remote test with a measured distance (Strawberry, 26 miles from the last line) needed only 0.6 kg of battery; (vi) the long, high-energy tests were at Moray's house with Moray's antenna; (vii) the 1938 capacitor test shows a high-voltage, low-current, high-frequency antenna lead, which is what an oscillator driven from stored or hidden energy also produces.
• Hayes: one non-neutral rebuild by someone with access and a stake failed.
• Final verdict (split):
• Lamp-and-iron loads of 0.1 to about 8 kW nameplate were reported 1925-38 by many named witnesses, including state and REA officials (documented testimony, weighed above as evidence that the effects were seen): Never properly tested as a physical claim. Witnessed by named observers, uninstrumented for power; the neutral whole-device test was offered and not completed; one non-neutral rebuild failed.
• "Box ran from antenna and ground alone at those loads": Never properly tested. Battery explanations are disfavoured for the 1928 and 1938 events if nameplate is near delivered power; hidden feed, ground path, priming charge and observer error are not excluded. The Strawberry run excludes nothing about batteries.
• 4-50 kW from 60 lb as a standing performance: Never properly tested; 4-8 kW are nameplate counts, 50 kW is unwitnessed; if from any known radiation, see Claim 3.

Claim 2: It was not induction, hidden wires or batteries

• Agree: Never properly tested. Induction is unlikely for Strawberry (26 miles) and Dugway; not excluded in town. Batteries are disfavoured by the energy arithmetic for 1928 and 1938. Hidden lead, ground path and stored charge are not excluded.
• Disagree: the earlier arbiter said critics "did not insist on instruments" and that nothing was instrumented. That was wrong. Instrumented or crude checks exist: Yates 1929-30 (main switch opened, terminals shorted with no spark, condensers vs battery, crystal set on the leads gave hum-free reception); Welling 1930 (house meter checked); Pyper (aerial shorted to ground, "just turned off the power"); the 1938 Securities Commission inspection (parts weighed and meter-tested for battery action, box bored, 30 x 150 W + iron + motor over an hour) [read: 4th ed. p.132; 5th ed. p.117]. None measured power. Physicist and contrarian rank the Securities inspection highest; the limits are Moray's laboratory, Moray's wiring and antenna, no trace of the building wiring, a statement dated 1956-58 and sworn 1963, the Commissioner's own report absent, and no statement that the detector was opened.
• Stronger argument: the contrarian's ranking of evidence by how well it separates hypotheses, combined with the mathematician's point that a crystal set free of hum, a shorted terminal without spark, and a series-capacitor result are all consistent with an oscillator fed by a source or by stored charge.
• Tested / never / dismissed: induction was tested once, in the form Eyring defined, by Moray's side in Eyring's absence (Emigration Canyon, Dec 1925). Batteries were tested by weighing and metering parts (1938, Yates). A hidden lead was tested only by disconnect experiments that cannot separate the hypotheses. Millikan (1929) and Bell Labs (1937) declined or did not act, for the reason they gave (a concealed part), which I judge defensible; it is their stated reason, not an independent finding. Frazer's 1939 "glass test" (twelve sheets of government-supplied window glass in series with the antenna; "the device operated as if the glass had not been in series") [read: 5th ed. p.180, family account] is the same kind of series-capacitor check as 1938: it shows a high-voltage, high-frequency lead and does not separate an external source from a stored or hidden one. Test offered, inventor withheld the detector; tests offered by critics not carried through.
• Final verdict: Never properly tested. Mains induction is not supported for the remote runs; battery explanations are strained for 1928 and 1938; hidden feed, ground path, priming and observer error remain open.

Claim 3: The source is oscillating cosmic radiation captured by resonance; energy is conserved

• Agree: quantitative form Refuted by calculation (4.8e-5 W/m2 against 5e4 W: 1e9 short, 4e9 if isotropic; or 37 times sunlight; or 1.4-3.9 times all sunlight for the per-person figure). Resonance as a source has no source term. The RLC algebra in Ch.10 is textbook-correct with typos (period without 2 pi, sign, mu) [read: page image pp.113-115].
• New (all three): Moray's own 1960 text says the first stage is excited by an external energy source and then sustained by "harmonic coupling to the cosmic wave frequencies" [read: 4th ed. p.116]. A primed decaying oscillator looks like a sustained one for a short time. The supply is the entire question; the 158 h run is the nearest thing to an energy audit and it is unmetered.
• Nuenergy (Perreault: local spark-gap stations until 1928, then radioactive dopants): Perreault's own reconstruction, published in one source with no second source found yet; he is an interested party (he proposes to build a commercial generator on it). It is weighed on its merits. Physicist's web check: General Order 40 (30 Aug 1928, effective 11 Nov 1928) reassigned the broadcast band and says nothing about spark-gap power [web https://en.wikipedia.org/wiki/Federal_Radio_Commission]. It fails the same arithmetic (an ideal 7,200 m2 aperture at 1 V/m gives under 20 W; real antennas far less [calc; web https://en.wikipedia.org/wiki/Antenna_aperture]) and contradicts its own p.19 ("radio signals ... much too weak") [read]. The tube inspection claim has no instrument, method or count.
• Appendix III (Maxwell): static fields (E = 100 V/m, H = 39.79 A/m = B/mu0 for B = 5e-5 T) carry no power; E/H = 2.5 ohm against 377 ohm for a wave; a real E = 100 V/m wave carries 26.5 W/m2 [calc].
• Tested / never / dismissed: no one measured a field at the device. Astrophysical backgrounds are measured; they do not contain such a flux.
• Final verdict (split): (i) kilowatts from known radiation (cosmic rays, starlight, microwave background, atmospheric electricity, broadcast RF): Refuted by calculation. (ii) "Resonance sustains or creates energy" as written: Internally inconsistent (no source term). (iii) Unknown oscillating radiation: Not testable as stated until a frequency and flux are given. (iv) A primed or stored-energy or hidden-feed mechanism: Never properly tested (see Claim 1). (v) Ordinary antenna reception of microwatts to milliwatts: Established.

Claim 4: The Moray valve (a stone; preceded the transistor; 1968 radioisotope cascade)

• Agree: crystal detectors rectify and predate 1910; a rectifier does not supply energy; no sample, schematic or independent examination. Never examined by an outsider in a recorded test (detector withheld in 1925, 1927, 1928, 1931, 1933, 1963). First transistor action Nov-Dec 1947 [web https://en.wikipedia.org/wiki/Transistor]; "bi-polar germanium diode" is a diode; no gain was ever shown.
• Performance figures conflict (all three): stone 1 oz / less than 1 oz / 8 oz / 2 oz; 50 kW in 60 lb (about 833 W/lb) in the 1930s against about 1 lb per kW (1962) against "below 1 lb per watt" (1968), a factor of about 600-1,000; 10-20 good tubes in 100 (1960) against 2 in 100 (1962); 29 tubes per unit at $500; $5.38 M to finish; Hayes 1931: 4,000 W with 50 kW "potential". Moray's 1960 status describes a delicate, balance-critical system, with failure modes (felled tree, jarred bench, hammer blow, Knudsen's switching) arriving when the detector was disturbed.
• Hayes's failed rebuild is the one recorded copy attempt (non-neutral, negative, ambiguous).
• Perreault's tube inspection (Nuenergy, rewritten 2026-10-03). Perreault reports first-hand that he bought tubes said to be Moray's: "the widow of T. H. Moray gave these tubes and his notebook to a man named Stan Mahurin"; Mahurin passed the tubes to a friend, where they "remained in storage for over twenty-five years"; "Upon inspection of these tubes there does not appear to be any radioactive material in them. However, this does not mean that it was never in the tubes" [read: Nuenergy p.12]. This is evidence, from one published source with no second source found yet: (i) that some Moray tubes physically survived his death, which bears on the silence about the device (below), and (ii) that a visual or casual check found no apparent radioactive material. Its weight on (ii) is low because no instrument, count or method is given, and the provenance chain is relayed to him ("It was communicated to me"). Perreault is an interested party: he proposes his own mechanism and plans a commercial generator [read: Nuenergy deep read]. His Pb-210 "burned out" argument fails on its own numbers (half-life 22.2 years [web https://en.wikipedia.org/wiki/Isotopes_of_lead]); his Bi to Po-210 route has no nuclear mechanism [memory, high]. The tubes still exist, by his account, and a gamma-spectroscopy check of them would test both his claim and Moray's radioactive-dopant account.
• 1968 proposal: a different, honest concept (radioisotope excitation). About 1 W per lb is roughly what a radioisotope thermoelectric generator gives (MMRTG about 2.8 W/kg [web https://en.wikipedia.org/wiki/Multi-mission_radioisotope_thermoelectric_generator]); betavoltaics give microwatts [web https://en.wikipedia.org/wiki/Betavoltaic_device].
• Disagreement: physicist and mathematician apply "Refuted by calculation" to "the detector alone supplies net power"; the contrarian does not. I keep the earlier ruling: that is energy conservation applied to the detector alone, so Refuted by calculation; whether the box had another source is Claims 1 and 3.
• Final verdict (split): (i) a mineral or semiconductor rectifier-type detector in the box: Plausible, untested. (ii) "Preceded the transistor": Never properly tested and Not testable as stated without the stone (the evidence is Moray's and his sons' statements and a redacted 1927 certificate; no physical stone has been examined; see "Silences in the record"). (iii) Detector alone as a net power source: Refuted by calculation (energy conservation). (iv) 1968 radioisotope cascade at about 1 lb per W: Plausible, untested as a modest-power concept; no result recorded.

Claim 5: Matter dissociates into energy; the technology transmutes elements and heals

5a. Dissociation of matter into energy. Supplying 50 kW needs only 17.5 mg per year of mass converted [calc], so a weighing could never test it. (Added 2026-10-08: the thesis that matter dissociates is Gustave Le Bon's of 1903-05, which the books reproduce without his experiments and footnotes; see "Investigator findings (2026-10-08)", section 2, for the split.) Bulk matter decaying to energy at relevant rates is excluded by the proton lifetime lower limit (about 2.4e34 years [web https://en.wikipedia.org/wiki/Proton_decay]) and the absence of heat; Moray gives no rate or mechanism. Verdict: Not testable as stated (no rate, no reaction); the bulk-decay form is Refuted by calculation (stability limit).

5b. Gold and silver (Research Institute, 1963-70; J. E. Moray, Hooper et al., not T. H. Moray). - What changed. The earlier arbiter said "no blanks" and "assays by the Institute's own people" and refuted the claim by an energy budget. The first two corrections stand: the residue went to "an independent assayer company" (unnamed, no certificate, not stated blind) [read: Rex p.2-3]; two irradiated no-solution controls exist (0.12 and 0.26 oz Au/ton against 17.2-100 with solution, 66- to 833-fold) [read: Rex pp.13-14]. The third, "the energy budget does not refute it", is itself withdrawn (2026-10-03): it rested on an arithmetic slip, 125 J per gram, for the correct 1.25e8 J per gram of residue [computed; see the reconciliation table], against reported doses of 1.6-200 J per gram irradiated [read: Rex pp.7, 13-14]. - Why the controls are not enough: n = 2; the controls are nonzero (0.12-0.26 oz/ton is 4-9 g/tonne, comparable to ordinary ore grade); no unirradiated blank, no environment-only blank, no assay or flux blank, no split or blind assay, the reagent-grade sample may have had no control at all (a reading of the labels, flagged by the deep reader); outside accelerator operators hosted the Institute's own technician and solutions. Fire-assay flux carries gold and silver and is always run with reagent blanks [memory, high; not web-checked]. John Moray later says the "environment" contains arsenopyrite, pyrite and antimony, classic gold hosts [read: Rex p.21; memory, high]. The Institute's own conclusion needs pre-existing "isomers of the isotopes" in the ore, which contradicts the gold-free reagent test [read: Rex p.14]. The Institute's own records show failed duplication (six runs at 4-7 oz/ton), 3 of 9 pilot runs below 5 oz/ton, "very few" worthwhile linac results afterwards, and yields that scale inversely with sample mass (30 g: 65-67; 60 g: 33-35), which a fixed added amount divided by sample mass would produce [read: Rex pp.6-7, 17-19; calc]. The statistics are internally inconsistent: Rudolph mean 15.16 with 10.51 at 97.5%; Olsen mean 11.79, 95% interval 9.22-14.36 (14.66 misprinted); 10.5 reused at 99% on an edited distribution; an interval read as a probability on the mean [read]. - Two independent refutations. The energy budget (1.25e8 J per gram of residue needed, at most about 200 J per gram delivered [computed]) and the nuclear argument each rule out a nuclear origin alone. The nuclear argument: building Z = 79 from elements of Z of 19 or less needs about 60 protons added per nucleus. Photons and electrons of 0.75-8 MeV remove nucleons and do not build them; charged-particle capture faces Coulomb barriers of order MeV to tens of MeV per step [memory, high; the mathematician's "about 100 MeV" is [memory, medium]]; a yield near 1e20 atoms per gram would leave strong activation products and an isotopically mixed gold; none is reported. Atom-scale gold has been made by accelerators at a net energy loss (Seaborg 1980, Bi to Au) [web https://en.wikipedia.org/wiki/Nuclear_transmutation]. - 1940-41 assays (Boaz Mine, Norris MT; Union Assay Office SLC, serial 88193) appear in the 1945 pamphlet and Fate 1956 with no certificates in the folder; not evaluated. - Final verdict (split): nuclear origin of the gold at the reported yields: Refuted by calculation (energy budget and nuclear argument, each sufficient alone). That the assays are real gold from the process rather than flux, container, reagent, sampling or drying artefacts: Never properly tested (thin internal control; no outside replication; no neutral party ran a test, so I do not use "Tested, not supported", though the Institute's own later runs mostly did not reproduce the high numbers).

5c. Uranium ore "ionic recovery" (booklet Isotopes in Ionic Recovery of U3O8). New in the claim list. - Reported: 0.04-2.6% ore raised to 21-67% U3O8; a US Bureau of Mines radiometric report (28 Sep 1956, sample UP 6234-4: 18-28% "U3O8 equivalent") with no "before" value and a note of an unidentified fibrous, isotropic substance that "might be an artificial product"; an AEC Grand Junction estimate on a 30% sample that it "appears to be fused or reduced uranium dioxide ... probably 75 and 85%"; commercial assays (Crisman & Nichols; Black & Deason); the same ore lot at 19.6-51.4% on repeated runs [read: Isotopes deep read]. - If no uranium is created, 96-99.8% of the mass is discarded [calc], which is ordinary milling. No masses, mass balance, controls, splits or isotopic analysis. Radiometric equivalents rise when radium daughters are concentrated [memory, high]. "Aging" runs the wrong way (uranium decays toward lead). The Bureau and AEC statements are third-party but are radiometric or visual evaluations of treated product and say nothing about upgrading. "Dr. Marven said no such ore exists except by aging" is a remark relayed in the booklet with no source document; it is recorded here, but it says nothing about whether the ore was upgraded, so it does not move the verdict. - Final verdict (split): creating uranium: Refuted by calculation (mass balance; no low-energy route to build Z = 92). That the data reflect an upgrade rather than concentration, contamination, sampling scatter or radium daughters: Never properly tested; the numbers presented fit the ordinary explanations.

5d. Cosray therapy (US 2,460,707, filed 30 Apr 1943, granted 1 Feb 1949; continuation-in-part of a 15 Nov 1940 filing). - Corrections to the earlier arbiter. (i) "No clinical data on either side" was wrong: an unsigned, undated tally of 785 treated exists [read: Photo Actinia pp.49-53]. It has no control group, no follow-up, "relieved" undefined, all patients also on conventional treatment, and its arithmetic fails (row sums: 783 treated, 706 relieved, 75 no effect against a stated 42; the sinus row has 309 treated but 304 + 3 = 307; sinus is 39% of patients; cancer 54/78 = 69% printed 75%; blood pressure 91% printed 99%; "metabolistic general" 44/44 is not a disease) [calc by mathematician and contrarian, from the deep read's transcription; I did not re-add the rows]. (ii) The June-July 1940 USDA item is a "written request and order" to build three devices, not a request for clinical testing; the "extended clinical tests by various M.D.'s" are unnamed and without results [read]. (iii) The FDA misbranding action (NJ 8137): default entered 3 Apr 1963; the company contested; its motion to vacate was denied 20 Aug 1963; article destroyed 19 Mar 1964. The FDA's charge, as the prosecuting agency, was that the labelling was false and misleading; the court entered a default and did not rule on efficacy. The FDA's description of the device is the agency's characterization, not a finding of fact about what it does. "Finding of no therapeutic value" is withdrawn. (iv) The patent's tube carries uranium or carnotite coating while the book condemns radium therapy [read]. - Evidence on outcomes, each from one source with no second source found yet: (a) a lessee, Gilbert, testified to a 1964 Senate committee on health frauds that she leased a Cosray machine for $3,210 on a satisfied user's word, that John Moray warned her never to let anyone open the cabinet, and that her mother showed no improvement; the FDA and the Florida State Board of Health then seized it [read: health-frauds p.85]. The hearing was convened on quackery, so its framing is the committee's. (b) Betz's 1954 handwritten note reports that a Geiger counter rate rose near the device and that lead did not shield it [read: digest]; partly illegible, no instrument details. (c) In 1949 visitors were told the "cos ray machine" was "reportedly quite successful in treating cancer" [web https://newspapers.lib.utah.edu/details?id=30889738], a claim in circulation, not a result. (d) The 785-patient tally above. These are evidence about what patients and users reported; none has a control, so none can show the device caused improvement or failed to. - Final verdict: Never properly tested in any controlled trial. An uncontrolled tally and a labelling default are not tests; it was neither "tested and failed" nor cleanly "dismissed without testing" (regulators acted on labelling). The tally itself is Internally inconsistent (local finding about the table, not the device).

Claim 6: The work was blocked, ignored or suppressed

• Agree: the patent rejections, FDA action, NASA and Bennett letters, AEC and Bell Labs responses are documented events.
• Sabotage, violence and foreign interest: the evidence (rewritten 2026-10-03). These accounts are evidence, quoted and weighed here; the reviews' "family account only" is not a reason to set them aside.
• Destruction of the device, 1939. John Moray, relaying his father: after about two months of tests Frazer said "I have tested and tested until I am fully satisfied and feel there are no tests left to make. As long as I have the device I will want to run tests, and I wish you would destroy the damn thing so I cannot ask for any more"; after the glass test, "one morning in Henry's laboratory, totally unexpectedly, Frazer took a hammer and smashed the [device and] said, 'Now I cannot ask for any more tests!'" [read: 5th ed. p.180]. The same source records Frazer's counter-account: "Later when my father made protests, Felix claimed that Dad had destroyed his own device" [read: 5th ed. p.181]. So there are two competing accounts of who destroyed it, both reaching us through the sons. A third published version, Tom Bearden's 1978 foreword, says the device was smashed by an assistant who "in my personal opinion was a trained Soviet agent" [read: 5th ed. p.18]; Bearden was a promoter of the book, and his version conflicts with the sons' naming of Frazer, an REA man. Outside corroboration: 1944 Senate committee testimony, as reported, says Frazer, "an investigator for the works projects administration, spent two years 'assisting, begging, cajoling and threatening Moray to develop his radiant energy'" and that Moray was paid by the REA from February 1939 to February 1941 [web https://newspapers.lib.utah.edu/details?id=28911552]. That confirms from a government record Frazer's role, the dates, and that his conduct included "threatening"; it does not mention the hammer. Weight: the destruction is documented in one detailed published account with a named actor, date range and setting; no source contradicts that the device was destroyed in 1939; who destroyed it is disputed within that same source.
• The shooting, 2 Mar 1940, and later attacks. John Moray (then 13, not an eyewitness to the struggle, relaying his father): his father was struck in the dark at the laboratory door; in the struggle a gun discharged, "grazing the side of Dad's leg"; he drove off the attackers at gunpoint, exchanged fire with a third man, saw "a fourth man ... Henry recognized this man as Felix Frazer", and recognized "the voice of an FBI man he had known at one time as a security guard"; "the doctor knew Dad was doing government work, and Dad promised to report the incident to the FBI himself. Consequently, the doctor made no direct report. As far as I know, no police report was ever made" [read: 5th ed. p.186]. The book adds two later attacks, a 1943 threat by Robert B. Perish, "Now it appeared that the U.S. Government was saying that he was to be liquidated", and that Moray's response was that his notes had been sent to friends in England and Canada [read: 5th ed. p.189]. The witnesses named (the unnamed doctor, S. E. Bringhurst, the assistant Julius Noyes) left no statement in the folder. Weight: a specific, dated, detailed account from one published source (Moray as told to his son, printed 1978); no second source found yet; nothing found contradicts it. The Utah newspaper search found no coverage, but it was a word search over poor OCR, and by this account no report was made, so newspaper silence is expected and is not contradiction [web research/2026-10-03-utah-digital-newspapers.md].
• Russian interest. This is not only the family's account. The FBI's own Salt Lake memo of 21 May 1951, filed in a Soviet espionage case ("Semen Markovich Semenov ... Espionage-R"), records Moray saying that John and Daniel Magdiel "would like to see Russia use this new invention in their second Five Year Plan", that John Magdiel contacted the Russian embassy, and that a Russian "whom Dr. Moray knew as A. A. Yakovlev" visited his laboratory in October 1929 [read: gov-FBI screenshot; FBI file p.14]. Moray told the FBI that this was not the Anatoli Yakovlev in their photograph. So the FBI took the Soviet contact seriously enough to interview him. Nothing in the memo bears on the 1939-40 violence; its silence there is not contradiction (it was an interview about Yakovlev).
• Interested parties on this subject: the family (heirs to the claim, selling the book and seeking funds); Bearden (promoter); Frazer (accused; his counter-claim survives only through the family); the FBI (one assailant is alleged to have been an FBI man, so its own files are those of an accused party); Halleck (a political opponent of the REA). Frazer's identity is a lead, not a finding: the Investigator found that a Felix J. Frazer of that era was a Technocracy economist and later an occultist [web https://archive.org/details/report00ofnationalnatirich], which fits the 1944 "stupid profit system" phrase and the sons' "investigating spiritual seances" detail; identity not proven.
• Business and court record (added 2026-10-03; the rule cuts both ways). Washoe County, Nevada, 1932: Moray petitioned the court to appoint directors of Moray Products Company [web https://newspapers.lib.utah.edu/details?id=28958019], confirming the Reno litigation the family describes. Third District Court, Salt Lake, 1942: judgment of $1,290 against Moray "for defendant's failure to deliver stock for money invested in corporation to use radiant energy in aircraft" [web https://newspapers.lib.utah.edu/details?id=25602932]; this is a court finding against Moray, not mentioned in the family's books as far as the digests show. 1970: the Research Institute and Moray sued Mark J. Hagmann for return of papers [web https://newspapers.lib.utah.edu/details?id=29749860], while the 5th ed. prints Hagmann's endorsement ("convinced of the validity of his work") [read: 5th ed. p.279]; the endorsement and the suit come from parties in a dispute.
• Patent grounds, corrected. The 23 Dec 1932 action gives inoperability (no means for the unheated cathodes to emit electrons; radioactive ionisation gives no space current without plate or grid potential). A separate Patent Office action states "No natural source of electric wave energy is known to the Examiner and proof of the existence of such a source is required" [read: gov-Commerce image, qe.md spot check 2; 4th ed. pp.163-164]. The earlier arbiter's "appears only in the sons' summary" is withdrawn. Contrarian: the cathode wording was weak (cold-cathode tubes were known [memory, high]) but the demand for a source is physically correct; physicist and mathematician: both grounds are the same two calculations as Claims 3 and 4. I agree with the contrarian: grounds partly weak, partly sound. These are the examiner's characterizations of the invention (the Patent Office was the deciding party, not a neutral tester), and they sit beside Moray's own description of a cold-cathode detector driven by an external wave source. Demanding proof of a source matches routine utility practice [memory, medium], so the rejection is not evidence of suppression by itself. The family's competing account: patent examiners were said to be "honeycombed" with employees of large companies (a 1931 letter from John Y. Smith relaying a former GE man) [read: 5th ed. digest pp.171-172]; that is a relayed allegation, recorded here, with no document about these applications. A related therapeutic tube did get a patent in 1949. On the later loss of the application files, see "Silences in the record".
• Both sides of the "blocked" story [read: 5th ed. pp.74-102]. Each official statement below is the characterization of a party with an institutional stake (an agency deciding whether to fund, a senator answering a constituent), reported as theirs: NASA agreed a $5 M programme per John Moray, then on 21 Jul 1961 judged the probability too low (the sources give the phone-call date as 24 Jul; the order is unclear; I do not resolve it); Sen. Bennett's 30 Oct 1961 letter says Moray asked for $50,000 without disclosing the box's location or the process (John Moray disputes one point); AEC (Larson, 1970): Eyring "convinced there was no demonstrable evidence ... Since Mr. Moray has not permitted examination of the apparatus, there is no way that one can rationally deal with the problem"; Wilkinson (BYU, 1959) offered help if Moray published or accepted an impartial jury. Against suppression: Moray declined Eyring's wattage test, ran the Knudsen test only on a written-statement condition, withheld the detector throughout, stopped demonstrating in 1932 on patent advice, and by his own 1960 account only 10-20% of tubes work.
• Researcher check (contrarian, accepted): John Moray gives Fletcher a sentence that is Judd's own; any refusal is read as suppression; witnesses include promoters and business associates (Judd, Hayes, Bringhurst, Lovesy); numbers shift (Strawberry 26 / 50+ miles; four or five lamps; 560 / 635 / 675 W; Hayes's "0.878 hp" should be 0.851). Critics' side: Fletcher's letters mis-date his visit, say Moray "refuses to let anybody see it" against his own 1928 account that he handled all but one part, and he asked for fees for reports; Eyring's accounts conflict across 1925, 1963, 1965 and 1970; Millikan declined without looking; Bell Labs wrote that a small chance warranted investigation and did nothing; the 2005 NASA survey calls the claim "not scientifically defensible" but is a literature survey, not a test [read].
• Stronger argument: the contrarian's "standoff" reading: Moray would show but not disclose; scientists would not endorse without disclosure; both stances are defensible alone and together guaranteed no test. There is no "tested and failed" and no sign of data being reanalysed, omitted or blocked from publication; Moray did not publish.
• Final verdict (split): (i) the patent rejections, FDA action and NASA, REA and AEC responses as events, with the grounds each stated: Established as documented history (an unusual use of the scale; the underlying documents reach us through the sons' books, and Frazer's, Crim's and the Securities Commissioner's reports are not in the folder). Added 2026-10-03: Moray's REA engagement (Feb 1939 to Feb 1941, $25 a day), Frazer's role and his "threatening" of Moray, the Wood report, and Soviet contact in 1929 as recorded by the FBI in 1951 are now also Established as documented history, from records outside the family's books. (ii) "Dismissed without testing because of orthodoxy": Never properly tested. The main critics looked, set conditions and withheld approval because the detector was withheld; tests were offered and not completed by both sides; a non-neutral rebuild failed; the Fletcher New York test is unresolved; the REA's own investigation (Crim, Wood, Frazer) ran tests whose reports are missing. Millikan, NASA, the AEC and the Patent Office did dismiss without a test. (iii) Changed 2026-10-03. The specific events (Frazer's destruction of the device in 1939; the 2 Mar 1940 shooting and later attacks; the 1943 threat): Never properly tested, was "Not testable as stated". They are specific, dated claims with named people, documented in one detailed published account; they could be checked against records that exist or should exist (the 74-page FBI file, REA files in NARA RG 221, the printed 1944 Senate hearing, Salt Lake police and medical records, Bringhurst's papers), and none of those has been examined. Nothing found contradicts them; the one competing account (Frazer's claim that Moray destroyed his own device) comes through the same source. Only the broad form, an organised conspiracy by unnamed parties to suppress the technology, stays Not testable as stated; but the specific allegation that an FBI man was among the assailants is testable against the FBI file.

Claim 7: "Dr."/D.Sc., Uppsala doctorate

Unchanged. The 5th ed. says Moray studied under the alias "James Cain Larson", passed orals by 1914, and received no certificate; the thesis "is not available" [read: contrarian, from the book]. The Uppsala library (7 Oct 2011) searched for "Moray", 1910-22 and the pre-1962 title registry; a search under "Moray" cannot exclude an alias; the 2018 enquiry has no recorded answer. The evidence for the doctorate is Moray's own account as printed by his sons; the Uppsala search found nothing, but it searched the wrong name, so its silence is not contradiction. The record that would decide is an Uppsala (or Swedish church or police registration) search under "James Cain Larson", 1910-1915. Never properly tested; no document yet supports or contradicts it; irrelevant to the physics.

Investigator findings (2026-10-04)

Added 2026-10-04 after the Investigator's three runs (sites.md leads 1-28; research/ notes dated 2026-10-03 and 2026-10-04; research/files/) and the "Investigator findings (2026-10-04)" sections of physicist.md, mathematician.md and contrarian.md. I read all three sections in full and re-checked the arithmetic that carries a ruling (venv-free Python, [calc]). I did not re-read the Senate hearing pages or the NARA images myself; I rely on the reviewers' reading of the saved copies and flag where a reviewer's reading goes beyond the quoted text. Numbering note: the physicist writes the Cosray claim as 5d (as here), the mathematician and contrarian as "5C"; the arbiter's 5c is the uranium claim. The first version of this section dropped one unfavourable item, Slattery's Tomadelli comparison (recorded by the physicist); QE caught it and it is restored below (Why REA stopped paying; Claim 6). With that correction, nothing found is left out because a reviewer judged it negatively; the weighing below sits beside the material, and Section 4 lists what each source is and who stands behind it.

1. Where the three reviewers agree, and where they do not

Agreed by all three: no new item is a metered measurement, so Claims 1, 2 and 7 keep their labels; Tesla US 685,957 (1901) is the same antenna-ground-condenser architecture as Moray's and does not change Claim 3; Losev and Lilienfeld show early solid-state gain devices existed (with a bias supply); the REA employment, the Wood report and Tugman's letter are documented in a printed Senate record that is the product of a hostile staff; the 1946 Cosray treatise describes a radium-like, penetrating, "harmless" applicator; the energy budget for gold (1.25e8 J per gram of residue) stands.

2. Claim-by-claim rulings on the new findings

Claim 1 (lamps and iron). Labels unchanged: Never properly tested (physical claim, weak and strong forms). New entries for the witness table: Dec 1936 Brig. Gen. Shinkle, Army Ordnance: four index sheets show a report filed as "sound detector"; the load and the report are not seen; the Feb 1939 enquiry through Sen. Thomas calls it a "Radiant Energy device", which is the enquirer's wording. A loudspeaker needs micro- to milliwatts, which an ordinary antenna-detector-amplifier supplies, so the Army record is consistent with both readings and settles neither. Craig (second-hand): "a small aluminum disk" ran "a small motor and fans by current, unseen current" (see table above). Wood (Aug 1939): opinion only. The record of a funded, equipped, two-year engagement that left no test result is now the strongest documented fact about "never properly tested": the opportunity existed and no neutral metered test is on record. No replication with metered input and output exists in anything the Investigator found (hobby videos 2016-2025, a Reddit "blueprint", a 2025 "specification"). That is stated as the state of the record; it does not close the claim, because by Moray's account the detector was never published, and the only recorded copy attempt remains Hayes 1932.

Claim 2 (not induction, hidden wires, batteries). Never properly tested, unchanged. Two quantitative additions from the contrarian, which I accept: stored charge in condensers cannot account for the long runs. [calc, re-done here] 1928 run: 115 W x 157.9 h = 6.5e7 J, needing C = 2E/V^2 = 130 F at 1 kV; the short 1930s loads (4 kW for 10 min = 2.4e6 J) need about 0.2 F at 5 kV; the Strawberry run (about 9e4 J) needs about 1.8 mF at 10 kV, which is tens of kilograms of oil capacitors [memory, medium]. So priming or a charged store can explain, at most, start-up and the shortest runs. That narrows the open alternatives to a feed through the leads (including the water-pipe ground), a hidden source, or unknown power. It does not exclude them. The mathematician's slow-filling arithmetic [calc: 0.025 microfarad at 2.4 kV stores 0.072 J, Q = 6e-5 C; at 3.5e-8 A it takes about 1,700 s; maximum power transfer from a 4e9 ohm source at 2.4 kV is V^2/4R = 0.36 mW] shows that the 1931 "condensers fill slowly like a bucket" observation by an unnamed physicist (one source, the 1945 booklet) indicates a high-impedance, low-current source, which is ordinary and says nothing about kilowatts.

Claim 3 (source). Labels unchanged. Numbers added for the Tesla-type collector (antenna, ground, condenser). The three reviewers used different assumptions, so I recomputed [calc]: fair-weather field about 100 V/m and air-earth current about 2 pA/m2 [web https://en.wikipedia.org/wiki/Atmospheric_electricity, the physicist's check; the mathematician used a global-current figure of 3.5 pA/m2, medium memory]. A 27 m pole at 100 V/m reaches about 2.7 kV. Power per square metre of collector is current density x voltage: 5.4e-9 W/m2 (physicist) to 8.4e-9 W/m2 (mathematician); the contrarian's 2e-10 W/m2 used a 100 V potential and is the lowest. So 1 kW needs between 1.2e11 and 5e12 m2 of collector; 1,000 m2 gives about 5e-6 W; 10,000 m2 about 8e-5 W. Range of the three: a factor of 40 in the assumed voltage; every version leaves the claim short by 1e7 or more. Ruling: atmospheric-electric collection joins the known sources in sub-verdict (i), Refuted by calculation for kilowatts. The mathematician's correction of his own radio bound is accepted: resonance enlarges the effective aperture of a small antenna (A_e = 1.64 lambda^2/(4 pi) [web https://en.wikipedia.org/wiki/Antenna_aperture]; 1.2e4 m2 at 1 MHz [calc]); then field strength limits the power: 1.4 kW needs A_e about 5e5 m2, reachable by a half-wave dipole only below about 150 kHz, and about 1 V/m [calc], a field found only near a transmitter. At 1 mV/m that 5e5 m2 aperture gives about 1.4 mW [calc: S = (1e-3)^2/377 = 2.65e-9 W/m2; x 5.3e5 m2 = 1.4e-3 W], and the 1 MHz dipole's 1.2e4 m2 aperture gives about 31 microwatts [calc: 2.65e-9 x 1.17e4 = 3.1e-5 W] (corrected 2026-10-04 after QE; the earlier text attached the 30 microwatt figure to the wrong aperture). Hence (i) stands for sources of any ordinary field strength at the remote sites; the one thing the record cannot give is the distance from a transmitter at Dugway or the 1928 house. Solar-wind particles (the mathematician's bound of about 3 W for 10,000 m2 in space, and they do not reach the ground) are an upper bound for Tesla's own stated mechanism and do not change the verdict. Bearden's zero-point reading: Not testable as stated (no frequency, flux or detector); parallel collection across 29 stages adds no energy (each stage is limited by the flux through its aperture; resonance multiplies voltage, not power); his "over fifty kilowatts" in the 157-h run disagrees with the primary record; his Soviet reading of the REA episode rests on a Salt Lake Tribune of 2 Dec 1941 that nobody has read, and the printed Senate pages point to Technocrat investigators, not Soviet ones. The Soviet contact that is documented is the FBI memo of 1951 (1929 visit by a Yakovlev), unrelated to REA.

Claim 4 (valve). Refined, not reversed. (i) A rectifier-type mineral or semiconductor detector in the box: Plausible, untested. (ii) Others' solid-state amplification before 1947 (Losev, biased zincite crystal amplifiers and oscillators, 1923-24; Lilienfeld, US 1,745,175, Canadian filing 1925 [web https://en.wikipedia.org/wiki/Oleg_Losev; https://en.wikipedia.org/wiki/Julius_Edgar_Lilienfeld, both via the Investigator, Wikipedia only]): Established as history. (iii) Moray's priority ("preceded the transistor"): Never properly tested and Not testable as stated without a dated exhibit or the stone, as before; it is now also weakened as a claim of uniqueness, since others had gain devices earlier. Physicist and contrarian differ on the label (Never properly tested versus Not testable as stated); I keep both, as in the earlier text, because the claim has a testable form (examine the stone) and an untestable form (priority without an exhibit). (iv) Gain or net power without a bias or other supply: Never properly tested; Losev's devices needed a battery; and the net power from a detector alone is Refuted by calculation (energy conservation). (v) 1968 cascade: Plausible, untested. The Patent Office's "cathodes not heated" ground (Dec 1932) stays "partly weak" in the light of cold-cathode and crystal devices known by then; the "no natural source" ground stays physically sound.

Claim 5a, 5b, 5c. Unchanged. The contrarian's recheck restores the energy budget after an arithmetic slip of 1e6, which agrees with this file (1.25e8 J per gram of residue; re-computed here: 9.75e19 atoms per gram x 8 MeV = 1.25e8 J [calc]). A robust form: the Institute's top reported dose of 200 J per gram allows about 13 eV per gold atom [calc], while any nuclear reaction costs of order MeV, so the shortfall exceeds 1e5 even if each atom needed only 1 MeV. The contrarian's adjacent-Z remark (mercury to gold, as in the 1941 Harvard cyclotron result [memory, medium], cited in the 1945 booklet next to a September 1943 transmutation letter) is physically conceivable at tiny yield but needs mercury in the solution, which the Institute never reported (Hooper's solution held elements of Z of 19 or below), and the energy budget applies to any route. It dates Moray's transmutation interest to 1943, twenty years before the Institute's runs. Hagmann (probably Mark J. Hagmann, BYU physics 1960, Utah PhD 1978 [web IEEE biography via the Investigator]; Paracelsus Research Society lecturer 1966; defendant in the Research Institute suit of 1970, case 195569; printed endorsement in the 5th ed. of unknown date): a possible trained witness and a record that may explain how the collaboration ended. Identity probable, not proven; the reviewers differ on his degrees (a 1966 "M.A." against the IEEE biography). No verdict moves.

Claim 5d (Cosray therapy). The family's 1946 "Short Treatise on Alpha, Beta, Gamma Ray Therapy" (copyright 1946 [web Catalog of Copyright Entries, via the Investigator]; authors John Eugene and Richard R. Moray, not T. H. Moray) says the applicators emit alpha, beta and gamma rays, have a half-life of about 1,600 years, penetrate the body and half an inch of lead "(proven by Geiger Muller Counter tests)", need "no special shielding", and will not injure healthy tissue "because of the internal phantom characteristics" [web https://archive.org/details/thomas-henry-moray-radiant-energy]. The text does not name radium; 1,600 years is radium-226's half-life [web https://en.wikipedia.org/wiki/Radium, the mathematician's check], so radium or an ore carrying it is an inference, a likely one. My rulings: - (a) "Alpha, beta and gamma radiation all penetrate the body and half an inch of lead": alpha particles of a few MeV stop in tens of micrometres of tissue or in paper, beta of about 1 MeV in millimetres of tissue [memory, high]; only gamma can pass through the body and a significant part of 1.27 cm of lead. Gamma from a radium chain at about 1 MeV has a lead half-value layer near 1 cm [web check this round: https://en.wikipedia.org/wiki/Half-value_layer gives 4.8 mm for Ir-192 gammas of 0.3-0.47 MeV, and the value rises with energy, so about 1 cm at 1 MeV is consistent but I did not find the figure itself], so the contrarian is right that the lead statement can be true of the gamma component. The sentence is true of gamma only. As a statement about alpha and beta: Refuted by calculation (established ranges); as a statement about the gamma component: consistent with a radium source. The Geiger-counter "proof" shows a counter rate (Betz's 1954 handwritten note, partly illegible, reports a counter rate rising near the device and lead not shielding it, which also fits gamma). - (b) "No special shielding is necessary" and "will not injure healthy tissues": the treatise gives no activity, so the hazard cannot be computed. Penetrating radiation deposits energy along its path, and radiation does not distinguish tissue by the design of its source; "internal phantom characteristics" has no stated physical content. Ruling: Not testable as stated (no activity or dose given), and contrary to standard radiological practice for any gamma source of therapeutic strength [memory, high]. I do not apply the mathematician's "Internally inconsistent" to this: wrong physics is not self-contradiction. - (c) Book against treatise: the book condemns radium therapy and describes a non-ionizing "radiant" agent (the physicist's reading; I did not re-read the book's wording), while the treatise sells a radioactive applicator and the patent's tube carries a uranium or carnotite coating. The contrarian notes that carnotite ore carries radium in equilibrium [memory, medium-high], so the patent and the treatise are reconcilable; the family does not say so. A discrepancy in the family's own record, to be asked of the family: Internally inconsistent only if the book and the treatise describe the same object, which has not been shown; I record it as an unresolved inconsistency, with no change of label. - (d) Conditional effect: if the applicator is radium-bearing, any benefit or harm would come by ordinary ionizing radiation (radium therapy was standard in this era [memory, high]) and not from a cosmic-RLC field. As benefit: Plausible, untested (no dosimetry, no control); as hazard: unquantified. The contrarian's split is accepted. - (e) Efficacy of the Cosray device, the 785-patient tally and the FDA labelling default: unchanged (Never properly tested; tally Internally inconsistent; FDA charge is the agency's characterization, decided by default). - (f) Perreault's tubes ("no radioactive material" on inspection, no method) may be valve tubes, not applicators; the two statements concern different objects until shown otherwise. Gamma spectroscopy of any surviving applicator and of the tubes tests both. The safety fact is the one the earlier arbiter lacked: by its makers' own 1946 text the applicator was radioactive and sold as needing no shielding. That bears on the FDA action (a labelling and hazard matter) and not on whether it helped anyone.

Claim 6 (blocked or ignored). The record now splits cleanly. (i) Events as documented history, Established as documents (an unusual use of the scale): the patent rejections; FDA, NASA, AEC, Army; the REA engagement (17 Feb 1939 to 28 Feb 1941, $25 a day, at most 742 days = $18,550 [calc]; REA bought laboratory equipment), Wood's report (Aug 1939), Frazer's role, Tugman's letter (May 1940), Slattery's second-hand Craig account (2 Jun 1943), and Slattery's agreement that Moray's claims and those of "Count" Tomadelli (convicted of mail fraud in 1920 over an atmospheric-energy bulb) were "pretty similar", with the Tomadelli article sent to Mr. Lake on 23 Nov 1942 and Slattery saying he was responsible for ending Moray's employment [read: Slattery interview pp.1372-1373, saved copy]. These come from the printed 1944 Senate hearing [web https://babel.hathitrust.org/cgi/pt?id=umn.31951d03587118y, pp.1344-1345 and 1370-1376, per the Investigator]. Stakes: the investigators' staff (V. A. Young) were hostile to Administrator Slattery; Slattery defends himself; Tugman has a university's standing; Frazer and Wood are probably Technocracy believers (identity "probable"); Electrical World (11 Nov 1944, private-utility weekly) and Rep. Halleck (Chicago Tribune, 11 Dec 1944) used the episode against REA and add no test data. Each is reported as that party's characterization. Added Army record: Shinkle, Dec 1936 (above). (ii) "Dismissed without testing because of orthodoxy or suppression": Never properly tested, the standoff reading, now with more weight on both sides. For the standoff: a federal agency paid for two years and no result is on record; a qualified local sceptic was refused for 12 years (his account); the termination reason, on the accused Administrator's word, was political. Against the claim, and from the same Administrator: he judged Moray's claims "pretty similar" to those of a man convicted of mail fraud for an atmospheric-energy bulb, had become "very suspicious", and sent the Tomadelli article to a colleague (23 Nov 1942) before ending the employment; that is a suspicion of fraud by a party with a stake, with no test behind it, and it makes the termination a judgment by analogy rather than either a negative test or a purely political act. Against suppression: the same record shows access, equipment and payment, and no hostile act by REA beyond the end of funding; Wood and Frazer, if Technocrats, wanted it to work. Millikan, NASA, the AEC and the Patent Office did dismiss without a test; their stated reason (a concealed detector, no disclosure) is theirs. The new records do not show data omitted, reanalysed or publication blocked. (iii) Specific events, split: - A story that the Germans tried to assassinate Moray circulated by 1940-43 and reached the REA Administrator (Slattery, second-hand, "got into the Salt Lake papers"): documented in one government record, first non-family source for the story. A rumour, as Slattery relays it, from a man who never met Moray. The Investigator's roughly 25-40 Utah Digital Newspapers searches did not find it in print; the Tooele paper of 26 Mar 1940 has Moray speaking at a funeral three weeks after the family's date without mention of an attack (poor OCR; later-1940 pages not opened). Neither confirms nor contradicts. - The shooting of 2 Mar 1940 and Frazer's smashing of the device in 1939: Never properly tested, as ruled 2026-10-03. The three reviewers write "Not testable as stated". The scale reserves that label for claims too vague to test, or interpretations without a measurable difference; a dated, named event with a known file (FBI 917273, Salt Lake police and medical records, the REA files) is testable against records, though none has been examined. The Senate pages seen say nothing about a hammer or a shooting; the investigators' text is a summary written to ridicule REA, so silence there is weak. I keep my label. The broad "organised conspiracy" stays Not testable as stated. - "Soviet agent" readings (Bearden, "Z"): Not testable as stated and unsupported by the printed pages seen; kept in the record as his characterization.

Claim 7 (Uppsala doctorate). Unchanged: Never properly tested. The Swedish-newspaper searches under "Henry Moray" and "Cain Larson" were negative, which is weak for a student. Alter, Utah, the Storied Domain (1932) pp.439-440 is the next record to read.

3. Absence-of-evidence analysis: what changed

• FBI (Silence B). The Investigator found more of the log: request 917273 "MORAY, T HENRY", received 25 Feb 2000, closed 17 Aug 2000, 74 pages reviewed, 72 released (requester Ravnitzky); a second 2000 request (924753) closed for lack of a notarised consent; request 1178194-002 "MORAY, THOMAS HENRY DR.", received 1 Dec 2011, reopened 8 Feb 2013, closed 11 Feb 2013 (three days after reopening), 0 pages reviewed, 0 released, "No Records" (requester Badia) [read: research/files/FBI-FOIA-log-2013Q1_Moray-1178194-002_excerpt.txt] [web https://archive.org/details/foia-log-2000-2002; https://archive.org/download/UnsortedFBIDocuments/foia-log-1-1-2013-to-3-31-2013-pdf_djvu.txt]; CIA F-2000-01342 (22 Jun 2000), outcome not found; the 72 pages are not on the FBI Vault, MuckRock, governmentattic or theblackvault. This turns Silence B from "a file not posted" into a sharper anomaly: a file released in 2000 and "No Records" in 2013. Competing explanations: (1) destroyed or transferred under a retention schedule after 2000; (2) different name form or index; (3) the "-002" request had a narrower or follow-on scope (for: the "-002" suffix, and the log's "Date Reopened" column shows the request reopened 8 Feb 2013 and closed 11 Feb 2013, a three-day interval that fits a quick re-search of a narrow or follow-on question better than a fresh review of a 74-page file; against: the log has no scope field, so what was searched in the reopened request is not stated); (4) the main file was transferred to NARA; (5) reclassification or withholding. The physicist argues against (1) and (2) from FBI practice (releases are usually retained; the index cross-references names); both are [memory, medium] and I did not check them, so they weigh little. Nothing in the record supports (5), and nothing excludes it. The reviewers and I agree no physics verdict depends on it. The record that decides: the 2000 release itself, requested as previously processed records citing 917273; NARA for any transfer.
• Silence C (missing hearing and agency records). The printed Senate hearing (78th Congress) has been located and partly read, so that item moves from "a gap in our access" to a located record: investigators' report pp.1344-1345, Slattery interview pp.1370-1376. It dates the REA episode 1939-41, with testimony in 1943-44, not the 5th edition's "1942". The investigators say fuller REA project files are in "a separate report" not seen. The FDR Library Carmody finding aid lists REA Chronological File Boxes 102-103, card indexes 91-94 and Box 99 "Senate Hearings [1944]" and names no Moray (9.6 MB, not read page by page); NARA RG 221 and RG 46 remain unsearched. Craig's "lost or destroyed in the 1941 move" explanation is no longer the only one: a Senate staff had "voluminous information", and the Carmody boxes exist.
• New Silence F: the Army's 1936 report. Four NARA War Department index sheets (RG 407) show a report by Brig. Gen. Shinkle of 8 Dec 1936 filed as "sound detector". The report is not found. Explanations: filed under the 413.68 decimal file or the 095 Steffensen name file and not yet requested (most likely); misfiled; destroyed under a schedule; withheld. Nothing indicates the last two. Decides: the file itself.
• Silences A, D, E. Unchanged in substance. For A: the Ottawa Bray fonds (letters 1928-56 from Schiess, Betz, Buchli and Hooper) is an unread record located by the Investigator, not a silence. For E: Hagmann's 1970 suit over papers is a record that may explain how the collaboration ended; the 1966 Paracelsus bulletin and the 1943 transmutation letter add dates, not results. D: the Slattery rumour is a first non-family mention of a story; it does not supply a police or press record. Business and court records (debt suits 1930-56, Sanford v. Moray 1942, a Moray Products notice in Washoe County 1932; no online Nevada index) are consistent with the "commercial strain" explanation and are silent on the device.
• Replications. None with metered input and output is on record after 90 years. 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. It is a datum to weigh, not a verdict.

4. Interested parties in the new material

Moray family (heirs, selling books); Senate investigators (hostile to Slattery); Slattery (accused; second-hand from Craig); Tugman (department head; stock sold in his state); Wood and Frazer (probable Technocrats, wanted a positive result); Electrical World and Halleck (against REA); Bearden (own scalar-vacuum theory); Perreault (plans a commercial generator); the FBI and CIA (logs only; no content seen); Shinkle (Army; report not seen); the family authors of the 1946 treatise.

5. Unread records (kept in the record, not weighed)

University of Ottawa, Arthur Bray fonds 30-003 (F4 Radiant Energy correspondence 1928-56; F24 Schiess 1928-57 "witness accounts of the machine in use" and drawings; SS4-F4; F6 Betz; F8 Buchli; F17 Hooper): finding aid only. These are letters kept by a Canadian government engineer outside the family's editing; they are the best route to test whether the family books quote the witnesses in full and whether Fletcher's letters are as quoted. Alter (1932); A Brief Analysis of the Moray Radiant Energy Device (50 pp.); the 1943 and 1956 printings of Radiant Energy (the folder's printing citing post-1945 physics may be the 1956 one); the Wood report in full; the Shinkle report; the FBI 74 pages; Washoe and Utah court files; the 1931 physicist's name.

6. Web verification (this section; budget 8)

1 check this round: https://en.wikipedia.org/wiki/Half-value_layer (lead 4.8 mm for Ir-192 gammas; the figure for about 1 MeV was not on the page; the "about 1 cm" is still [memory, medium] and no verdict rests on it). Taken from the reviewers' checks and retagged: atmospheric electricity (100 V/m; 2 pA/m2) [web https://en.wikipedia.org/wiki/Atmospheric_electricity]; radium half-life 1,600 y and gamma emission through the chain [web https://en.wikipedia.org/wiki/Radium]; antenna aperture [web https://en.wikipedia.org/wiki/Antenna_aperture]; Losev and Lilienfeld [web Wikipedia pages above]. Not checked and still [memory]: FBI retention and indexing practice (medium); electrostatic fans on high-voltage sources (medium); oil-capacitor energy density (medium); Harvard 1941 mercury-to-gold result (medium); carnotite radium content (medium-high). No ruling above rests on one of these alone.

7. Verdict changes from this section

8. QE remediation (2026-10-04, after the "Investigator update check" in qe.md)

• Fix 1 (retention): restored Slattery's Tomadelli comparison ("I think they are pretty similar"; Tomadelli article to Mr. Lake 23 Nov 1942; "I think I did" / "I think I was" on ending Moray's employment; "very suspicious") to the "Why REA stopped paying" row and Claim 6 (i) and (ii), as an accused party's characterization with its stake; corrected the opening sentence that said nothing had been dropped; added it to the open question on the termination. Read directly from the saved Senate OCR text pp.1372-1373; the Tomadelli conviction detail is the Investigator's bracketed summary of an omitted passage and is labelled so.
• Fix 2 (arithmetic, Claim 3): recomputed [calc, venv Python]: S = E^2/377; at 1 V/m S = 2.65e-3 W/m2, so 1.4 kW needs 1400/2.65e-3 = 5.3e5 m2; at 1 mV/m S = 2.65e-9 W/m2, giving 1.4e-3 W (1.4 mW) on 5.3e5 m2 (1.3 mW on 5e5 m2, as QE found), and 2.65e-9 x 1.17e4 = 3.1e-5 W (31 microwatts) on the 1 MHz dipole's aperture (A_e = 1.64 x 300^2 / (4 pi) = 1.17e4 m2). The 30 microwatt figure now sits with the 1.2e4 m2 aperture. No verdict changes.
• Fix 3 (record detail, Silence B): added the FBI log's reopen date (1178194-002: received 1 Dec 2011, reopened 8 Feb 2013, closed 11 Feb 2013, 0 pages reviewed, 0 released, "No Records") from research/files/FBI-FOIA-log-2013Q1_Moray-1178194-002_excerpt.txt, to section 3 and to the Silence B body, and added the three-day reopen-to-close interval as evidence for, and the missing scope field as the limit on, explanation (3). physicist.md's FBI paragraph and explanation 3 were given the same date.
• Reviewer file: physicist.md Section B now attributes the Tomadelli comparison correctly (the investigator Young put the question; Slattery answered "pretty similar"), with the termination answers and page numbers.

Investigator findings (2026-10-08)

Added 2026-10-08. Inputs: the "Investigator findings (2026-10-08)" sections of physicist.md, mathematician.md and contrarian.md, which weigh sites.md leads 29-43 (the follow-up leads from Q01, Q02, Q04, Q08, Q09, Q11 and Q12) and the research notes 2026-10-06-followup-Q01, -Q02, -Q04, 2026-10-07-followup-Q08, -Q09 (hypotheses put to the review, not web-researched), 2026-10-08-followup-Q09, -Q11, -Q12. I read all three sections in full. I did not re-read the Investigator's notes; where a ruling turns on a figure I redid the arithmetic [calc] and checked the quotations that carry the provenance finding against the saved pages in research/pages/ (section 7). Numbering: the arbiter's 5a-5d stay as above (5d = Cosray). I add 5e (resonance transmutation, the Isotopes booklet) and 5f (physics in the books that comes from other authors, and the Le Bon observations). The reviewers number these differently (the physicist as a new sub-line of Claim 5, the mathematician as 5A(v) and 5E, the contrarian as 5d and in the 5a-5c split); I use 5e and 5f throughout. Tags: [read] project documents as quoted in the Investigator's notes; [web] URL given; [calc] redone here; [memory] with confidence. Nothing found is dropped because a reviewer judged it negatively; the weighing sits beside the material. Findings are leads.

1. The largest finding: the physics pages of the books are largely other authors' text

All three reviewers accept the Investigator's matches, and so do I (checked, section 7). In order and nearly word for word, with the credits visible in the source missing in Moray's printing:

Agreement. All three: this is a fact about what the pages are, not about whether the physics in them is right and not about what the lamps did in 1925-38. Physicist and contrarian add that nothing in the record shows who prepared each printing or why a credit is missing; I draw no conclusion about intent either. The mathematician adds a caution I accept: "nearly verbatim" is a comparison against the Rex Research transcription and Google Books snippets, not a collation of the editions Moray used.

Where they differ. None on the facts; two on emphasis. The contrarian stresses the direction of the edits (what was removed made checking harder; what was added carries no content; the one changed number, "1 to 2 per minute", lost a number rather than gaining one). The physicist stresses that single searches that return only Moray copies "do not prove originality", and the contrarian says the same of the "no source found" items. I accept both: the five rows' last line is "no source found yet", not "Moray's own invention".

My ruling. 1. The passages are evidence of what Moray read and chose to print, not of what he measured. The two items that look most like findings in them, "ten thousand times more active" and "silver to cadmium", are respectively Rutherford's 1900 measurement via Le Bon and a sentence added to a medical-physics text with no datum. A reader may not cite Radiant Energy pp.53-59 or Isotopes pp.4-6, 20-25 as Moray's experiments or analysis. 2. It also answers the owner's question (Q01): "the researches of Rutherford ... taken up again" is Le Bon's 1905 call to repeat Rutherford's alpha-velocity measurements, not a research programme of Moray's. Rutherford's 1906 result was that alphas stop ionising, exciting phosphors or darkening plates below a critical speed of about 9,000 km/s [web https://www.nobelprize.org/prizes/chemistry/1908/rutherford/lecture/, via Q01; I did not re-fetch]. Le Bon's footnote (after Strutt, 1906), dropped from Moray's printing, hoped ordinary matter emits slower, undetected alphas; see 5f for the hypothesis and its answer. What the sons say is that Moray "became deeply involved ... in the study of synthetic radioactivity as described by Gustave LeBon" [read: 5th ed. pdf p.187, via the Investigator]; the borrowing shows he used the book heavily. 3. The provenance does not bear on Claims 1 and 2 (the lamp demonstrations and the induction question), and does not make the passages false. Each passage is weighed on its content below. 4. Dating. The folder's Radiant Energy printing labelled "1926" mentions the atomic pile and plutonium, and the same Rutherford sentence is in the 1945 printing (Q01). The label is therefore not a date of composition; any document that gives "1926" as a date should say "a printing labelled 1926, compiled after 1945". My 2026-10-04 section (5. Unread records) guessed that the folder's printing "may be the 1956 one"; that guess is open.

2. Claim 5: dissociation, the Le Bon material and resonance transmutation

5a (dissociation of matter into energy) is split by origin. Physicist: Le Bon's thesis "Historically superseded but instructive", beside my earlier "bulk decay Refuted by calculation" and "unquantified Not testable as stated". Mathematician: 5A(i) Not testable as stated unchanged, new 5E "Historically superseded but instructive". Contrarian: the thesis is Historically superseded but instructive; the plain energy-from-mass statement is Established; as Moray's power source it stays Not testable as stated. The three agree on the substance; they differ only in how many rows they open. Ruling: - Le Bon's 1900s thesis that all matter dissociates and that light-driven charge loss shows it: Historically superseded but instructive. Reasonable c.1900-06 (he was nominated for the 1903 physics Nobel [web https://arxiv.org/pdf/1208.3114, Kragh pp.13-14, via the contrarian]); replaced by nuclear structure (1911 on), which gives stable nuclei and a specific list of unstable ones. Ordinary silver does not decay [memory, high]; the proton-lifetime limit stands as before. - His energy point, that a mass loss too small to weigh could free large energy, is right in plain form: 50 kW is 4.8e-8 kg per day, 17.5 mg per year [calc: 5e4 W x 3.156e7 s / (2.998e8)^2 = 1.76e-5 kg], and arrived two to four years before mass-energy equivalence was accepted. Established as E = mc^2; no Moray rate or mechanism. (My 5a sentence "Moray gives no rate or mechanism" stands; the concept is Le Bon's.) - Moray's use of it as the source of his power: Not testable as stated (unchanged). Bulk decay at relevant rates: Refuted by calculation (unchanged). - The reasonable-in-1905 verdict is not a verdict on the books' use of it. The books present it, without the experiments and footnotes that made it Le Bon's argument.

5f. Le Bon's observations and what became of them. Reviewers agree on the history; they differ slightly on labels. - Thorium "ten thousand times more active" (Rutherford 1900; Radio-activity 1904 sec.166: thoria at +300 V, negative wire, "over 10,000 times as active per unit surface as the thoria" [web https://archive.org/details/radioactivity00ruth; checked in the saved text]). Explained by Rutherford and Soddy (1902-03): thoron decays to charged daughters (Pb-212) that the field drives to the wire; concentration, not new energy. Mathematician's check reproduces the order of magnitude (about 3e4 from a thick-source surface density against a collected-deposit surface density [calc, assumptions his]). Energy: 100 g thoria, 10% thoron escape, about 36 kBq of Pb-212 on the wire at about 11 MeV per tail: 6e-8 W, against about 2.4 microwatts for the whole chain [calc, physicist, mathematician and contrarian agree; thoria activity and chain energies [memory, medium]]. Established as an effect; no energy gain. - Amalgamated tin "extraordinarily radioactive" in sunlight. Every test the sources allow says photoelectric effect: threshold near the solar ultraviolet cut-off (295 nm = 4.2 eV [calc]), dependence on sign of charge, on cleaning and oxide, and Le Bon's own identification of his particles with Lenard's electrons. Elster and Geitel had the amalgam result in 1889-92 [web Lodge, via https://www.gutenberg.org/files/69253/69253-h/69253-h.htm, via Q12]; Ramsay and Spencer (Phil. Mag. 12, 1906) repeated Le Bon's experiments and confirmed the main rates (Table IX: mercury 216 s, tin amalgam 0.6 s); Millikan and Winchester (1907) read the same data as photoelectric; Campbell (Athenaeum, 1906) objected that radioactivity cannot be provoked by light. The observation is Established (it is the photoelectric effect); the reading as "radioactivity" or "dissociation of matter" is Tested, not supported (the contrarian's label; physicist and mathematician say the effect was answered "within his own decade" and give no separate label; I adopt the contrarian's because controlled repetitions exist). - Two Le Bon claims not confirmed. Ramsay and Spencer found no measurable positive leak and no discharge through metal shields, and blame weaker ultraviolet sources than his spark discharges. The contrarian reads this as not a clean negative: Never properly tested with spark-strength ultraviolet. The physicist and mathematician record the non-reproduction. I take the contrarian's distinction, which the verdict scale asks for. Campbell is a rival worker (stake). He writes that he "investigated one of these [new phenomena], and found it capable of a totally different explanation", and that a "Mr. Carse has tried and failed completely to repeat another of these experiments"; Campbell names neither experiment (checked in the saved Athenaeum text, research/pages/athenaeum-1906-p1-djvu.txt). - The defending reply (added 2026-10-08, QE remediation). The same saved Athenaeum volume prints a reply signed "F. L." (no. 4087, 24 Feb 1906), the writer of the journal's "Research Notes" whose favourable references to Le Bon Campbell had attacked; sites.md lead 29 gives the English translator of The Evolution of Matter as Legge, and the initials match, but the extract does not identify him [read: saved Athenaeum text; identification an inference]. Stake: he is defending his own column and, if he is the translator, Le Bon's book. He says (i) "I am not acquainted with Mr. Carse's work", so the Carse experiment stays unnamed; (ii) the case Campbell himself investigated "seems to be that of the quinine sulphate", which "Prof. Kalahne, of Heidelberg, has lately repeated, with the result that he finds the reaction due to chemical change, as originally pointed out by Dr. Le Bon, and not to heat, as contended by Mr. Campbell"; (iii) Parodi, in a memoir to the Institut Egyptien, "has repeated his chief experiments on phosphorescence and on the variability of chemical species with perfect success"; (iv) Rutherford's Radio-Activity mentions Le Bon's "lumiere noire" and quinine-sulphate luminescence "without questioning their validity". His follow-up letter (no. 4091, 24 Mar 1906, same file) sets out the quinine record: Le Bon (Revue Scientifique, 14 Apr 1900; 15 Nov 1902) attributed the effect to hydration; Miss Gates (Physical Review, from McGill, 1903) agreed the cause is chemical but found the radiation "completely absorbed by a thin sheet of aluminium" and "no evidence for believing that the ionization ... is due to the spontaneous projection of charged masses from the atom"; Kalahne (Annalen der Physik, 1905) found no phosphorescence or ionisation without water vapour; Gates (Physical Review, Jan 1906, Cavendish) produced both in dried, unheated quinine sulphate exposed to damp air. Weighing: the Kalahne and Parodi reports reach us second-hand through an interested party and I have not read them; on quinine the record he cites supports Le Bon against Campbell on the cause (chemical hydration, not heat) and, through Gates, against Le Bon's wider reading (a molecular, not atomic, process, stopped by thin aluminium). So it does not change 5f: the observation stands, the "dissociation of matter" reading is not supported. Parodi's "perfect success" is a single, unread report; no second source found yet. - Le Bon's detection-threshold footnote (added 2026-10-08, QE remediation). The footnote dropped from Moray's printing quotes Strutt (Nature, 25 Jan 1906): "Ordinary matter may be emitting as many or more alpha particles than uranium, if only their velocity is less than that minimum velocity which has been found necessary to produce the characteristic phenomenon" [web https://www.rexresearch.com/lebonmat/lebonmat.htm, via Q01]; Le Bon used it to explain why ordinary matter seems inert. Rutherford's critical speed of about 9,000 km/s is about 1.7 MeV [calc: 0.5 x 6.645e-27 kg x (9.0e6 m/s)^2 = 2.69e-13 J = 1.68 MeV]. The physicist's answer (physicist.md, Investigator findings 2026-10-08; the QE check attributed it to the contrarian, whose file does not carry it): a slowed alpha is a helium ion that picks up two electrons and becomes a helium atom; and alpha decay is energetically allowed only for heavy nuclei, so stable silver, copper or carbon emit none [memory, high]. The hypothesis was reasonable in 1906 and is superseded by nuclear binding energies. - Le Bon's arithmetic. From "40 degrees in 10 seconds" against "6 in 60" he infers "40 times" the uranium screen and, multiplying a per-gram alpha rate, "3,000,000 particles/second". All three find the step wrong in kind: an alpha makes about 4.2 MeV / 35 eV = 1.2e5 ion pairs [calc], a photoelectron one charge, so equal currents mean about 1e5 times more particles for the photo-emitter, each far less energetic. His 70,000 alphas per gram per second fits equilibrium uranium within a factor 1.4 and overstates thorium about threefold [calc, mathematician; half-lives [memory, high]]. His alpha speeds (30,000 km/s = 18.7 MeV) are an overestimate; no natural alpha exceeds about 8.8 MeV [memory, high]. - Why it matters beyond Moray. All three note that here an anomalous claim was repeated within a year by a leading chemist (Ramsay), reinterpreted by a competing group, and answered in print. The contrarian uses it to ask whether Moray was "uniquely ignored" (answer: his claims were not subjected to equivalent tests largely because no outsider was given the device, Tugman, Claim 6). The mathematician and physicist treat it as a comparison case, not evidence about Moray. I keep it as a comparison only; it does not change Claim 6.

5e. Resonance transmutation (Isotopes pp.20-26, Radiant Energy p.34): a new row. What is claimed: current, rays and sound "in step" with the "natural frequency" of a substance cause "an atomic change such as the change from silver to cadmium"; "if these oscillations become violent enough, disintegration and transformation will take place ... This can be proven mathematically." No mathematics, silver or cadmium datum is in the folder [read: Q08 via the reviewers].

The decisive arithmetic is shared and I recomputed it [calc]. 107Ag to 107Cd directly is uphill 1.416 MeV and 109Ag to 109Cd 0.215 MeV; the real routes need a neutron (109Ag + n releases 6.8 MeV, then 110Ag beta-decays, 2.89 MeV, 24.6 s, to 110Cd) or a proton over a Coulomb barrier of about 9.8 MeV [calc: 47 x 1.44 MeV fm / (1.2 (107^(1/3) + 1) fm)]. A 1 GHz photon carries about 4e-6 eV, visible light about 2 eV, a phonon a few hundredths of an eV; bubble collapse reaches 12,000-20,000 K, kT about 1.7 eV, which is 9.8 MeV / 1.7 eV = 5.8 million times short. Raising intensity gives more quanta, not bigger ones. One gram of silver at 1 MeV per atom needs 5.6e21 atoms x 1.6e-13 J = 8.9e8 J, about 250 kWh. The mathematician adds two points the others do not: (1) the Preston passage is a free damped LC discharge, and a free damped oscillation cannot grow (successive peaks fall by exp(-pi alpha / omega_d)), so "violent enough" needs an external drive or negative resistance the text does not supply; Q is finite in any real medium and multiplies amplitude, not available power; (2) a 450 J Leyden-jar discharge (1 nF at 30 kV) into 1 g of silver is 0.5 eV per atom against 10^5-10^7 eV needed. The physicist notes a tension in the booklet itself: its own p.20 (from Glasser) says a quantum's energy is proportional to frequency, which is the premise that defeats its p.23 conclusion. Three competing readings of the booklet are kept as in the contrarian's note: (a) a theory assembled from reading, silver to cadmium an illustration; (b) real chemistry (ionisation, radiolysis, electrolysis, cavitation) with the wrong name; (c) a real transmutation, which needs a neutron or MeV source the booklet never describes and has no data. Reading (b) is the sceptic's artefact route and goes into "What would settle it".

Related work kept in the record. Keely (c.1882): his own company's engineer reported that "water ... is capable of being, by vibratory motion, disintegrated" [web https://en.wikipedia.org/wiki/John_Ernst_Worrell_Keely, citing the NYT 14 Dec 1882, via the contrarian]. That is a party with a stake; Wikipedia's characterisation of Keely as a fraud rests on an 1899 investigation that found hidden compressed-air tubing, which the company's president disputed. Tesla's oscillator story concerns mechanical resonance of structures; a MythBusters trial (2006) could not shake a bridge [web https://en.wikipedia.org/wiki/Tesla%27s_oscillator]. Both are forerunners of the idea, not support for an element change. Nothing here moves 5a, 5b or 5c: the energy budget for gold (1.25e8 J per gram of residue) is unchanged. The contrarian's artefact route for 5b (field-and-solution treatments with electrodes in contact with radium-bearing solutions would plate radium daughters on a negative electrode and read high radiometrically) is an inference by the contrarian, not shown; it is the kind of control the data lack.

Moray's page 60 (physicist only). Radiant Energy printed p.60: "the work of Sir Rutherford with helium and nitrogen transmutation into oxygen and hydrogen" as proof that "the alchemists were right"; Moray says he has split "the nuclei of paper, of lead, tantalum, mercury, tungsten, gold, copper and aluminum by a method far less expensive than the cyclotron" and made tritium by electrolysis [read via Q01]. Only the physicist weighed this; I read his ruling and the arithmetic behind it, not the page. Rutherford's 1919 N-14(alpha,p)O-17 is Established. Splitting those nuclei "far less expensively than a cyclotron": Never properly tested (no measurement is given), with the same energy-scale objection as 5b unless an MeV particle source is stated. Tritium by electrolysis: Established as isotope enrichment (the heavy isotopes stay in the cell [memory, high]); as production of new nuclei Not testable as stated. Noted as single-reviewer.

3. Claim 3: the "hidden reservoir" arguments, the amalgam collector, Compton and the 1968 infrared paragraph

All three examined the hypotheses put to the review in Q09 as physics questions, not as Moray claims; none of its arguments is in a Moray text, but they bear on whether a hidden source could supply the power.

• Neutrinos. The Sun's flux is about 6.5e10 per cm2 per second, about 30 W/m2 [memory, high; all three]. Weber's claim of a 1e23 enhancement in stiff crystals was tested at least four times with null results (Franson and Jacobs 1992; an Eot-Wash test; Luo et al. 1995; Feng et al. 2006, "a null result ... 526 times smaller than the predicted value" [web https://www.colorado.edu/physics/sites/default/files/attached-files/gravitywavesandneutrinos.pdf, Franklin 2010; the 526 figure checked in the saved text]). Franklin notes the idea is "still alive" in a 2008 analysis (Apostol). This corrects the earlier Q09 note's "never confirmed": it was tested and not supported. Credit to the testers: an anomalous claim here received neutral repeated tests.
• Stellar bounds. The contrarian, rightly, applies the Q09 point that stellar models are not absolute. The bound rests on independent checks: Borexino (Nature 562, 505, 2018) puts the Sun's neutrino luminosity at 3.89 (+0.35/-0.42) e33 erg/s against 3.846e33 from photons (within about 10%); CNO neutrinos have been observed (arXiv 2006.15115); helioseismic sound speed agrees within 0.2% rms (1990s models); N_eff = 2.99 +/- 0.17 (Planck 2018) [via Q09 notes; the Borexino and arXiv pages were saved by the Investigator; I did not re-fetch]. So a hidden channel is bounded to a few per cent of the Sun's output, not excluded. The physicist and contrarian add the strongest mainstream form of the objection: the device would need a coupling strong enough for a 100 kg box to extract kilowatts and weak enough that the Sun is not drained. That is argument, not demonstration. The Homestake episode (a third of expected neutrinos for 30 years, resolved by flavour change) is a fair example that anomalies there resolved into new physics; it argues for measuring.
• Entropy and the Q09 Pauli analogy. "Entropy as missing energy" is a category error (all three); the parallel with the beta spectrum and Pauli's neutrino is apt as method (missing energy has been inferred before from a measured deficit). No measured deficit exists in Moray's case.
• Vacancy current. The mathematician checks that for a radial current in a spherical cell the total current (J + epsilon0 dE/dt) is zero, so there is no magnetic field; this is correct and applies only with exact symmetry. It is a device class that exists (Moseley 1913, betavoltaics) and supplies no energy. The charge bookkeeping is right: 109Ag to 110Cd leaves a missing electron, carried by the beta; bare-ion bound-state decay (163Dy, 187Re) needs stripped ions and cannot make neutral silver into cadmium (215 keV uphill; cadmium K-shell binding about 27 keV).
• Plauson US 1,540,998 (1925). In the Nuenergy compilation (pdf p.83): "zinc amalgam or gold amalgam or other photoelectric acting metals in the form of small patches in extremely thin layers (.01 to .05 mm.)" on a collector balloon, with "polonium amalgams" for greatest effect [read via Q12]. It is the Elster-Geitel photocathode on an atmospheric collector, and the same family as Tesla's patent weighed on 2026-10-04. The physicist's ceiling uses a UV flux of about 20 W/m2 (300-360 nm), a quantum yield of 1e-4 and 0.5 V, all [memory, medium]; the contrarian and mathematician scale from Le Bon's electroscope numbers with assumed capacitance (nanoamperes per 100 cm2). The two methods give 5 to 8 orders below 50 kW. Plauson's text is the patentee's description; there are no performance data for the balloon in the folder. Moray's books do not use the amalgam experiments, only Le Bon's word "dissociation" (Radiant Energy printed p.46; he elsewhere writes "photo electric ionization", p.30). Ives (1924, 1926) shares ancestry (Elster and Geitel) but a different mechanism; no bearing on a Moray verdict.
• Infrared paragraph of the 1968 proposal. A correct summary of 1968 detector practice (extrinsic Ge:Hg to about 30 K, cutoff about 14 micrometres; HgCdTe aiming at 77 K [web https://link.springer.com/article/10.2478/s11772-012-0037-7; https://en.wikipedia.org/wiki/Mercury_cadmium_telluride, via Q02]). The microwave background peaks near 1,060 micrometres [calc: 2,898 / 2.725], about 100 times beyond the paragraph's wavelengths, and "microwave" and "3 K" do not occur in it. The mathematician adds that even a hypothetical intrinsic CMB photon detector would need about 0.75 K [his construction]. Second law: a perfect absorber of the 2.7 K background receives 3.1e-6 W/m2, while a 300 K device radiates about 460 W/m2 [calc, contrarian]. No source of the paragraph was found; borrowing is possible and not shown. No change to Claims 3 or 4.

Claim 3 final verdict (split, revised): (i) kilowatts from known radiation (cosmic rays, starlight, microwave background, atmospheric electricity, broadcast RF, now also photoelectric collection and Compton scattering): Refuted by calculation. (ii) "Resonance sustains or creates energy" as written: Internally inconsistent (unchanged; the mathematician's finding that Radiant Energy p.34 admits finite Q and then asserts unbounded growth is a further instance). (iii) Neutrino or similar weakly interacting source at kilowatts: Refuted by calculation; Weber-type enhancement Tested, not supported. (iv) Unknown oscillating radiation or hidden channel with no stated coupling, frequency or detector: Not testable as stated, now bounded. (v) Primed, stored or hidden feed: Never properly tested (unchanged). (vi) Ordinary reception: Established (unchanged).

4. Claim 4: the radioisotope premise, radon plating, phosphors and the Foster pressing

Radioisotope direct conversion (the premise of the 1968 proposal). The proposal cites Safenev's "experiments dating back to 1913" (no person of that name is identified in the review files; added 2026-10-08, QE remediation, from sites.md lead 30) and "1899"; the Investigator identifies these with Moseley's radium beta cell (Proc. R. Soc. A 88, 471, 1913) and Rutherford's electrometer work (the identification is the Investigator's, not the proposal's). Moseley's abstract: radium emanation (20 millicuries) in a bulb thick enough to stop the alphas, insulated in an evacuated flask; a million volts was the aim [web https://royalsocietypublishing.org/doi/10.1098/rspa.1913.0045, via the contrarian]. The 150 kV and 0.01 nA come from Wikipedia's "Atomic battery" (I fetched it: "Moseley first demonstrated high voltage of 150kV, with a current equal to 0.01nA" [web https://en.wikipedia.org/wiki/Atomic_battery]), not from the paper, which no one has read. Power: 1.5 microwatts [calc: 150e3 V x 1e-11 A], or about 75 microwatts per curie. The mathematician's ceiling: 7.4e8 Bq x 2 betas x 1.6e-19 C = 0.24 nA if every beta escaped, so 0.01 nA is 4%, plausible with absorption and finite insulation; even at that ceiling it is 36 microwatts [calc]. That retags the earlier "microwatts per curie [memory, medium]" to [web, medium, via secondary figures]. The correction applies that the potential was on the central bulb (the source), not an outer sphere. Established as a microwatt-scale effect.

At kilowatt scale. The reviewers differ on one number, and I resolve it. Physicist and contrarian: at 2.89 MeV per event (110Ag decay), 1 kW needs 2.2e15 decays/s, 58,000 Ci [calc: 1e3 / (2.89e6 x 1.602e-19) = 2.16e15; / 3.7e10 = 58,400 Ci]; 50 kW about 2.9 million Ci. Mathematician: 2.89 MeV is the beta end-point (equal to Q); the mean beta energy is a third to two-fifths of that [memory, medium], so at a mean of 1 MeV 1 kW needs 1.7e5 Ci [calc: 1.69e5]. The mathematician is right that 58,000 Ci is a floor (all of 2.89 MeV delivered to the load); I use the physicist's and contrarian's number as the minimum and state the mathematician's as a probable value. The 157.9 h run at about 110 W (6.25e7 J = 17.4 kWh [calc]) would need at least 1.35e20 events (about 24 mg of silver changed [calc]) and at least 6,400 Ci [calc: 110 / (2.89e6 x 1.602e-19) / 3.7e10 = 6,420]. The mass is trivially small and unweighable; the radiation would be enormous (physicist: about the dose rate of a sizeable teletherapy source [memory, medium]). Neither the reviewers nor the Investigator found a source reporting a radiation effect at the device, and no witness is recorded as having an instrument (Source gap); the Perreault statement that surviving tubes show no radioactive material (method unstated) is the only measurement-like remark. Electrical power cannot exceed the nuclear power released; self-annealing (the 1968 proposal; Bell Labs remark) is real and protects the device, not the energy account. Refuted by calculation for a kilowatt-scale radioisotope-excited or silver-to-cadmium converter; a radioisotope source also caps any "1968 cascade": about 58 Ci per watt at 2.89 MeV and 100% efficiency (physicist), so the isotope inventory sets the weight (the mathematician's "decay-power cap"). The 1968 cascade as a modest-power concept stays Plausible, untested.

Radon plating on a negative electrode (Perreault's second route). The contrarian corrects his own earlier reading: Perreault's two routes to "artificial radiation" were inconsistent as a pair, but "radon plated on a negative metal" is a real effect (Rutherford 1900). Physicist and mathematician reach the same practical point (a radium-bearing carnotite coating and a negative cathode collect radon daughters; microwatts to milliwatts for practical amounts, no energy beyond the source's own). Plausible, untested for Moray's tubes; no one measured. A consequence the contrarian draws and I accept: Pb-210 (22.2 y, the longest-lived daughter of radon-222) would remain on such an electrode decades later [memory, high], so gamma and alpha spectroscopy of a surviving cathode tests this route and weighs against Perreault's "burnt out" explanation.

Phosphors and the stone. Standard references: ZnS:Ag is the alpha scintillator; ZnS:Cu glows green and was radium luminous paint; ZnS:Mn is strongly triboluminescent; CaS:Eu,Sm is photostimulable, traps "too deep to be thermally emptied", emptied by infrared [web https://pmc.ncbi.nlm.nih.gov/articles/PMC5445864/, https://en.wikipedia.org/wiki/Zinc_sulfide, via Q04]. Le Bon's "residual charge made visible by infrared" matches; his 18-month "invisible light" belongs to his unconfirmed black-light claims. Stored energy is bounded by what was absorbed: 1e19 traps/cm3 at 3 eV is about 4.8 J/cm3 [calc]. Noffsinger's 1984 handwritten list of the stone ("germanium alloy, bismuth, iron sulfide, molybdenite, triboluminescent zinc and radium chloride") fits a radium-doped zinc-sulphide phosphor with a semiconductor; provenance not established, digest-only for the physicist. Together with the family's 1946 Cosray treatise (half-life about 1,600 years, radium-226) and Foster 1990 (below), three separate family-side or advocate-side items put radium in Moray devices or materials; none is a measurement. A radium-doped, phosphor- or rectifier-type stone: Plausible, untested (physicist), implying a slow, microwatt-scale source, not kilowatts; stored-charge phosphors as an effect: Established; that Moray's stone worked that way: Never properly tested (contrarian). No Moray experiment with a sulphide screen is in the record. "Diamond" in Moray's text is Le Bon's radium gauge (Brazilian stones glow under radium, Cape stones do not; footnote dropped; printed p.54, scan p.55) and two other passages; no Moray document read names diamond as his detector. The Q09 diamond hypothesis is not supported by the record and is not excluded.

Foster 1990 (lead 42; the only account in the record of an outside contractor handling Moray detector material). Perreault's Nuenergy (2001) p.30 quotes Arthur Glenn Foster, "Principal Engineer", 11 April 1990: "We had a contract to process some of the Moray 'valve,' fine-ground, 'Swedish Stone' into pressed 'detector' pellets. The pellet material was doped with a tiny amount of unpurified radium ... subjected to a huge number of millibars of pressure by a tetrahedral press ... When a certain critical pressure was reached 'something' in the material 'flashed.' The electric pulse produced was so violent that it burned out the 2000-amp carbon-die heater power supply, blew the main breakers in the building and blew the power line transformer" [read: Nuenergy p.30, via the Investigator]. It is one participant's account, relayed in a compilation by an advocate of the device (Perreault, who also proposes to build a generator on it; stake noted); no letter, date, contractor or contract has been found, and the Investigator found it only in other compilations of the same text. Not found is not the same as untrue. The three reviewers read the physics alike and differ in how far they decide: - Stored energy: 4.8 J/cm3 at the assumed density (physicist, contrarian); the mathematician bounds a radium source's deposit differently (1 microcurie over 30 years, about 170 J, only a few per cent trapped [memory, low]). Either way, joules. - A 2,000 A heater supply at an assumed 10 V is 20 kW; one second holds about 20 kJ, 4,000 times the 5 J example (mathematician; the 10 V is his assumption, not in the account). So breakers tripped and a transformer blown fit an electrical short through a conductive hot pellet or die far better than a release of stored radiation energy. The physicist also lists "both" (a small trigger and a large external supply). The contrarian declines to choose. - My ruling: the energy arithmetic favours the short-circuit reading (readings: stored energy release; short circuit; a flash that starts a short), but it does not decide, because no one measured. The account is consistent with a radium-doped phosphor-type stone and with nothing more; it describes no net-energy device and a pressing, not an operating box. It conflicts, as reported, with Perreault's "no radioactive material" in surviving tubes; different objects (1980s pellets, 1920s-60s tubes), a coarse instrument or an inaccurate account would each reconcile them. - It changes one earlier statement: "never examined by an outsider in a recorded test" (Claim 4, Agree) stays true of the whole device and the stone as a unit, but one unrecorded outside pressing of detector material, undated, is reported. Deciding records: Foster's letter, the press log, the contractor ("Project X"). The physicist notes the question of whether this is the Foster who relayed Fletcher's account in 1973 is unchecked.

Claim 4 final verdict (split, extended): (i) rectifier-type detector in the box: Plausible, untested; a radium-bearing stone Plausible, untested. (ii) "Preceded the transistor": Never properly tested / Not testable as stated (unchanged). (iii) Detector alone as net power: Refuted by calculation (unchanged). (iv) 1968 cascade as a modest-power concept: Plausible, untested (cap added); radioisotope direct conversion as an effect at microwatts: Established; at kilowatts: Refuted by calculation. (v) Radon-daughter plating on tube cathodes: Plausible, untested. (vi) Stored-charge phosphors: Established as an effect, Never properly tested for Moray's stone. (vii) Losev and Lilienfeld: Established as history (unchanged). The 1968 infrared paragraph needs no label (no claim).

5. Claim 6 and the absence-of-evidence analysis

• Claim 6: no verdict change. One family-only item is added: the sons write that Le Bon's book was "withdrawn from libraries" and that the FBI asked about it in 1942 [read: 5th ed. pdf p.187, via the Investigator]. Physicist: I do not weigh it as suppression or as ordinary weeding of a 1905 book; neither reading is excluded. Contrarian: the borrowing shows Moray used the book heavily; it does not corroborate the withdrawal or the FBI question. I agree. The 74-page FBI file (917273) is the record that could settle it, and now has a specific thing to look for (a 1942 entry on a "LeBon enquiry"); see Silence B.
• Silence A (the device and detector). Two additions. (1) The Foster pressing, if it happened, shows detector material was still being handled in 1990, 16 years after Moray's death, by a contractor; it bears on the "what remained" question, not on the demonstrations. (2) The deciding test for the tubes is now sharper: gamma and alpha spectroscopy of a surviving cathode (Pb-210 at 46.5 keV, Po-210 alpha, Ra-226 daughters) and a ZnS:Ag screen, which test the radon-plating route, Perreault's "no radioactive material" and any silver-neutron route (108mAg, 439 y, which decays mainly by electron capture to palladium, not cadmium [web https://en.wikipedia.org/wiki/Silver-108m, isotope table checked]) in one measurement.
• New Silence G: Moray's own laboratory experiments with these materials. The record contains no Moray experiment with a sulphide screen, an amalgam, a thorium plate or a silver/cadmium assay, although his books print the physics of all four. Competing explanations: (1) he ran none and copied the physics to explain a device built by other means; (2) he ran some and they are in his notebooks (held by the family; by Perreault's account the widow gave a notebook to Stan Mahurin); (3) the pages were compiled by the family or a typist after his death from books he owned (the "1926" printing was compiled after 1945). For: (1) every passage found is traceable to a printed source and none carries a datum of his; (3) the post-1945 content. Against: (1) "and sometimes other properties", "Herein be the door to greater discoveries" and the "silver to cadmium" sentence are insertions that suggest he thought in these terms; the sons say he studied "synthetic radioactivity as described by Gustave LeBon". Silence in other sources is not contradiction. Records that decide: Moray's notebooks (family; Mahurin material), the Ottawa Bray fonds (letters 1928-56), the 1943 and 1956 printings of Radiant Energy.
• Silence B (FBI 74 pages). Unchanged in substance. If the 1942 Le Bon enquiry is real it should appear there; it is a family-only claim until then.
• Silences C, D, E, F and the replication note: unchanged. The Le Bon repetitions are not a Moray replication; no Moray result has been repeated.

6. Where the three reviewers disagree, and my rulings (summary)

7. Web verification (this section; budget 8)

Done this section: (1) https://en.wikipedia.org/wiki/Atomic_battery: confirms "Moseley first demonstrated high voltage of 150kV, with a current equal to 0.01nA" and the globe with the radium emitter at the centre (so the source charges positive); (2) https://en.wikipedia.org/wiki/Silver-108m (isotope table): 439 y, decay mode electron capture to 108Pd (and isomeric transition); (3-7) five local checks against the Investigator's saved pages in research/pages/: Le Bon (lebonmat.htm): "the experiments of Rutherford and his pupils should be taken up anew", "this plate becomes 10,000 times more active, surface for surface, than the thorium itself" ("according to Rutherford"), "Brazilian diamonds alone --- Cape diamonds never", "a loss of substance too small to be detected by our balances"; Rutherford, Radio-activity (rutherford-radioactivity00ruth.txt): "over 10,000 times as active per unit surface"; Franklin 2010 (franklin-2010-weber.txt): "526 times smaller than the predicted value of Weber's theory"; Ramsay and Spencer 1906 (philmag-6-12-1906-djvu.txt): "unable to get a measurable discharge"; the Athenaeum 1906 text: "Mr. Carse has tried and failed completely to repeat another of these experiments"; (8) the Moray text "taken up again" is in the Sea of Energy 4th ed. extracted text. All match what the reviewers quote. Taken from the reviewers and Investigator and not re-fetched: Moseley's abstract, Planck, Borexino and Bahcall figures, the isotope tables, Smet et al. 2010, Keely and Tesla pages, Elster-Geitel via Lodge. Still [memory], not web-checked, and no verdict rests on any alone: mean beta-energy fraction (medium), solar neutrino flux 30 W/m2 (high), work functions of tin, mercury, zinc, aluminium, magnesium (medium), UV flux and quantum yield assumptions (medium), thoria and radium chain activities and energies (medium), teletherapy comparison (medium), the Fermi-constant scattering estimate (high), Wigner-energy storage fraction (low).

8. Verdict changes from this section

9. Corrections to earlier documents (for the orchestrator)

1. Physics attributed to Moray. Any statement in facts.json, claims.md (Claim 5 wording), the reviewers' earlier sections, primer.md and the rank entry that Moray "argues", "writes" or "analyses" the alpha-particle, spinthariscope, phosphor, induced-radioactivity, folded-cylinder or dissociation pages (Radiant Energy pp.53-59; Isotopes pp.4-6) should say these reproduce Le Bon (1905); Isotopes pp.20-25 are in part Glasser (1944) and Preston (1928). In this file: Claim 5a's "Moray gives no rate or mechanism" is amended by a note in 5a (concept is Le Bon's).
2. "1926". The folder's Radiant Energy printing labelled "1926" cites the atomic pile and plutonium and was compiled after 1945; the same Rutherford sentence is in the 1945 printing. Anywhere "1926" is a date of composition is wrong (physicist.md 2026-10-04, section H item 1, is the reviewer's own correction).
3. Q09 research note (2026-10-07) and A09 (followups.json). (a) Moseley's 150 kV was the potential of the central source bulb, not an "outer sphere", and the 150 kV and 0.01 nA figures are secondary (Wikipedia). (b) Weber's neutrino claim was tested at least four times with null results, not "never confirmed". (c) 108mAg (439 y) decays mainly to palladium, not cadmium. (d) The diamond sentence is on printed p.54 (scan p.55). (e) "and sometimes other properties" is Moray's insertion, not Le Bon's. (f) "58,000 curies per kilowatt" is a lower bound (all 2.89 MeV delivered); about 1.7e5 Ci at a mean beta energy of 1 MeV. (g) The note's neutrino stopping fraction (1e-19 for a metre of lead) is a low estimate; a coherent-scattering estimate gives about 2e-17; the conclusion is unchanged.
4. mathematician.md, Claim 5 verdict (i): "17.5 g per year" should read 17.5 mg per year (his own correction in this round; this file and the physicist already had mg).
5. Contrarian's earlier Claim 4 and 5c sentences: the Perreault two-routes inconsistency holds only for the bismuth-to-polonium route; the radon-on-negative-metal route is real (Rutherford 1900); "never examined by an outsider in a recorded test" needs the unrecorded Foster pressing; "microwatts per curie [memory, medium]" is now [web, medium]; "matter dissociates into energy is E = mc^2 restated" is Le Bon's thesis.
6. Q12 wording: plain tin is not in the solar light-leak table (amalgamated tin 1000; mercury 2); Figure 45 uses a positive electroscope. The question's wording is left as is.
7. Search for stale copies: the rank entry (work/rankings/entries/Morray-TH.json), primer.md Sections 3-6 and the evidence table, facts.json entries for the Radiant Energy and Isotopes documents, claims.md Claim 5 and Claim 4, and followups.json A01, A02, A04, A08, A09, A11, A12 should be checked for Moray-attributed physics, "1926" dates, and the A09 figures in item 3.

10. What would settle these (additions; also folded into "What would settle it" #10)

1. Gamma and alpha spectroscopy and a ZnS:Ag screen on a surviving tube cathode or coating (Pb-210, Po-210, Ra-226 daughters), and on any Cosray applicator: tests radon plating, Perreault's "no radioactive material" and a radium-bearing stone together.
2. A radiation survey beside a running device (any kW claim from nuclear conversion needs radiation of at least that power), and an assay of electrodes and process metals for cadmium in silver, with a blank.
3. Foster's 1990 letter, the press log and the contractor (a measurement of the flash would separate stored energy from a short).
4. Moray's notebooks (family; Mahurin material) for "and sometimes other properties", any screen, amalgam, thorium-plate or silver/cadmium experiment.
5. Moseley's full 1913 paper (the 150 kV and 0.01 nA figures), and Rutherford's 1900 paper (the contrarian did not open it).
6. A blank-controlled repeat of one Le Bon amalgam experiment with spark-strength ultraviolet (positive leak; aluminium screen), the part Ramsay and Spencer could not test; Nye (1974, Hist. Stud. Phys. Sci. 4:163-195) and the Athenaeum correspondence (pp.237-366) for the reception.
7. The complete texts of Glasser (1944) and Preston (1928) (the match is on snippets), and the printing of Le Bon that Moray's editor used (the Rex transcription may differ).
8. Why tin in mercury beats both metals (fresh oxide-free surface, lower work function or impurity) is not decided in the sources; it does not bear on a Moray claim.
9. (Added 2026-10-08, QE remediation.) Kalahne's 1905 Annalen der Physik paper ("Ueber die Strahlung des Chininsulfates", pp.451 ff.), Gates's two Physical Review letters (1903, 1906) and Parodi's Institut Egyptien memoir: the first-hand texts behind the "F. L." reply in the Athenaeum (5f).

11. QE remediation (2026-10-08, after the "Investigator update check (2026-10-08)" in qe.md)

• QE defect 1 (retention, Athenaeum reply): fixed. 5f now gives the "F. L." reply of 24 Feb 1906 (no. 4087) and his follow-up of 24 Mar 1906 (no. 4091), both in the saved research/pages/athenaeum-1906-p1-djvu.txt, with his stake and the second-hand status of the Kalahne and Parodi reports, beside Campbell's objection. Correction to the QE wording, from the saved text: the reply identifies as quinine sulphate the case Campbell investigated himself, not the Carse experiment ("I am not acquainted with Mr. Carse's work"), so the Carse experiment stays unnamed; 5f now quotes Campbell's own sentence for both. The quinine record the reply cites (Gates 1903, 1906; Kalahne 1905) supports Le Bon on the cause (chemical hydration, not heat) and does not support the atomic "dissociation" reading; the 5f verdicts are unchanged. Unread first-hand texts added to section 10 item 9. This material was first left out of this section although the arbiter had checked the page; the opening sentence "Nothing found is dropped" was not true of it until this fix.
• QE defect 2 (retention, Strutt footnote): fixed. 5f now quotes the footnote (Strutt, Nature 25 Jan 1906, via Q01), converts the 9,000 km/s critical speed to 1.68 MeV [calc shown in 5f], and gives the mainstream answer; section 1 item 2 points to it. The answer is in physicist.md, not contrarian.md as the QE check says. Safenev (lead 30) is now named in section 4 beside the Moseley identification. No verdict changes.
• QE defect 3 (primer acronyms and notation): not applicable to this file (primer.md only).

Silences in the record (absence-of-evidence analysis, added 2026-10-03; updated 2026-10-04: see section 3 of "Investigator findings (2026-10-04)" for Silences B, C, F and replications; updated 2026-10-08: Silences A and B extended and Silence G added, section 5 of "Investigator findings (2026-10-08)")

Where the record goes quiet where it would be expected to continue, I treat it as an anomaly, set out the competing explanations with the evidence for and against each, and name the record that would decide. None of these silences, by itself, moves a physics verdict; they bear on Claims 1, 4, 5 and 6.

A. The device and its detector after 1939, and after Moray's death (1974). What the record says: the 1939 device was smashed (Frazer by the family's account; Moray himself by Frazer's) [read: 5th ed. pp.180-181]; in 1942 Moray rebuilt a smaller unit "using the remaining bit of what was known as the 'Swedish Stone' ... which ... he had never been able to duplicate" and drove it from Salt Lake to Idaho, Wyoming and Denver in summer 1943 [read: 5th ed. pp.187-188]; then "the RE device had burned out during a test ... because of an overload ... The so-called 'detector' was no longer operative, for out of fear of compromising its secret, he had dismantled it months before" [read: 5th ed. p.189]. The sons write elsewhere that "demonstrations became impossible after the last detector-material burned out, although Henry Moray did not wish to emphasize this point ... This could have been a psychological defense that he himself had erected because of his financial inability to re-build his device" [read: 5th ed. p.170]. Yet demonstrations are recorded later: about a dozen witnesses in 1957 (digest only), and Eyring's viewing in 1963, after which the AEC wrote that he "has seen Mr. Moray's demonstration" [read: AEC fragment, 1970]. The 70-ft antenna pole was taken down in 1956 [read: 5th ed. p.189]. After his death, by Perreault's account, Moray's widow gave tubes and a notebook to Stan Mahurin; the tubes passed to a friend, sat in storage for over 25 years, and were bought by Perreault [read: Nuenergy p.12]. The 1978 foreword says Moray "died with his dream unrealized and the original device destroyed" [read: 5th ed. p.18]. No source records an inventory of what remained. Competing explanations: 1. Destroyed or burned out, and not rebuildable because the "Swedish Stone" could not be duplicated (the family's account). For: detailed and consistent in the sons' telling, including their own admission that this may have been a "psychological defense"; the 1960 text's 10-20% tube yield shows a fragile process. Against: the later 1957 and 1963 demonstrations imply a working unit of some kind existed after 1943; what it was is not described in the deep-read pages. 2. Deliberate concealment (trade secrecy) by Moray and his heirs. For: Moray dismantled the detector himself "out of fear of compromising its secret"; he withheld it in every recorded examination; he said his notes were divided and sent abroad; the sons say they hold the notes and could rebuild it [read: 5th ed. pp.189, 197]. Against: in the 48 years since 1978 no rebuild has been reported in the folder; secrecy explains the withholding but not why no unit was ever shown again to anyone. 3. Seizure or suppression. For: the family's account of the 1939 destruction, the 1940 shooting with a recognised "FBI man", FBI questions in 1942 about LeBon's book, the 1943 threat [read: 5th ed. pp.180-189]; 1944 Senate testimony that Frazer was "threatening Moray" [web Salt Lake Tribune 11 Dec 1944]; FBI files on him exist. Against: no source records a seizure of any device; Moray kept his laboratory and, by the family's account, was granted a secret clearance after a 1957 hearing [read: 5th ed. p.172]; some tubes passed through private hands after 1974 with no sign of official interest. 4. The device never did what was claimed, so there was nothing of value to preserve. For: the source arithmetic of Claim 3; the claimed performance fell about 600-1,000 times between the 1930s and 1968 (Claim 4); no metered run ever. Against: the 1928 and 1938 witness record is not explained by any test on record (Claims 1-2). 5. Transfer abroad. For: Russian interest in 1929 (FBI memo) and a 1941 British Admiralty invitation (family account, Bluit) [read: 5th ed. p.187]. Against: Moray declined both by every account; no foreign trace found. (Added 2026-10-08: a second handling of detector material is reported, Foster 1990, an undated contractor pressing of radium-doped "Swedish Stone" pellets, single-source, in an advocate's compilation; and the tube test is sharpened to spectroscopy for Pb-210, Po-210 and Ra-226 daughters plus a ZnS:Ag screen. See "Investigator findings (2026-10-08)", sections 4 and 5.) Records that would decide: the Salt Lake County probate file for T. H. Moray (1974), which would list estate property; the family (House of Moray) archive and the notebook said to be kept by Mahurin; a laboratory examination (gamma spectroscopy, X-ray, composition) of Perreault's tubes, which would test both the dopant account and Perreault's "no radioactive material"; the 1970 Hagmann case file (Salt Lake district court, case 195569), which may say what papers were in dispute; witness statements for the 1957 and 1963 demonstrations; the 1957 security-clearance hearing record.

B. The 74-page FBI file (request 917273). The FBI's FOIA log for 2000 records: "MORAY, T HENRY 917273 ... 2/25/00 8/17/00 74 72 Granted in Part(3)", requester Michael J. Ravnitzky; a second request in June 2000 by Larry E. Nelms was closed for lack of a notarised consent or proof of death [web https://archive.org/details/foia-log-2000-2002]. So a file indexed to Moray himself exists, 74 pages were reviewed and 72 released. It is not on the FBI Vault (search "Moray" returns only the Yakovlev file) [web research/2026-10-03-nara-catalog.md]. The folder holds only the 1951 interview inside the Yakovlev file. Competing explanations for its absence from public view: (1) routine: the Vault posts a selection of released files, and a release to one requester in 2000 need not be posted [memory, medium; not web-checked this round]; (2) mundane contents: the 1951 Yakovlev interview, the 1942 enquiry about LeBon's book, and the 1957 clearance check would by themselves fill a file of this size; (3) contents bearing on the violence: by the family's account Moray promised to report the 1940 shooting to the FBI, and alleged that one assailant was an FBI man; if he reported it, an entry should be in this file. The two pages withheld are a small fraction; the meaning of the "(3)" code is read from the log, not confirmed. The FBI is itself an accused party in the family's account, so the file is an interested party's record; its contents would still be decisive on whether a report was made. Records that would decide: the released file itself, requested from the FBI as previously processed records citing 917273, or from the requester or a FOIA archive. (Added 2026-10-08: the family's statement that the FBI asked about Le Bon's book in 1942 is a specific thing to look for in these pages; it is family-only until then.) (Added 2026-10-04: a later request, 1178194-002, received 1 Dec 2011, reopened 8 Feb 2013 and closed 11 Feb 2013, was answered "No Records"; see section 3 of "Investigator findings (2026-10-04)".)

C. Missing hearing and agency records. - REA reports (Crim, Wood, Frazer). Craig, a former REA deputy administrator, wrote that the engineers' reports "were put in the Administrator's files", that he does not have copies, and that they are probably "lost or destroyed" because the REA moved to St. Louis in 1941 and back after the war, with "much pertinent material ... even personnel records ... lost" [read: 5th ed. p.119]. Explanations: wartime relocation loss (Craig's own explanation; he was an REA official); deliberate removal (consistent with the family's suppression account; no evidence found); the reports survive in NARA RG 221 but have not been searched (the NARA catalogue search was blocked) [web research/2026-10-03-nara-catalog.md]. Decides: RG 221 series for 1938-41 and the Senate committee's exhibits. - The 1944 Senate Agriculture and Forestry Committee REA hearing. The press (Salt Lake Tribune and Chicago Tribune, 11 Dec 1944) says Halleck quoted it [web https://newspapers.lib.utah.edu/details?id=28911552]. A printed hearing should exist; it has not been read because HathiTrust and Google Books were blocked, so this is a gap in our access, not a silence in the record. The family book dates Halleck's inquiry to 1942 and says Moray was never called to testify [read: 5th ed. p.184]; the 1944 press date conflicts with that. Decides: the printed hearing (78th Congress) and RG 46. - The 1938 Utah Securities Commissioner's report, the Hayes testimony at Reno (1932) and the 1940 USDA order: none in hand. Explanations: not yet requested (the Washoe County case is now identified); lost; withheld by the family's editors (they printed favourable statements, not the Commissioner's report). Decides: Utah State Archives; Washoe County Clerk or Nevada State Archives; NARA (USDA). - The patent applications. Moray asked for copies in 1954; "the patent office replied that the patent applications had been destroyed because the office was under the impression that the applications were abandoned"; the sons write that this was incorrect and that "these patent applications have all 'disappeared' from the patent office" [read: 5th ed. p.170]. Explanations: routine disposal of abandoned applications (the Patent Office's own explanation; I recall that disposal of abandoned files was routine practice [memory, medium]); misfiling; removal. One examiner action survives in the folder [read: gov-Commerce image]. Decides: NARA RG 241 abandoned-application files and the PTO file-wrapper index for Serial No. 550,611 and the other six applications.

D. No police or press record of the 1940 shooting. The family's own account explains it: the doctor made no report because Moray promised to report to the FBI, and "as far as I know, no police report was ever made" [read: 5th ed. p.186]. Explanations: (1) unreported by choice, as the account says; (2) reported to the FBI only (see B); (3) the event did not happen as told. The silence is predicted by (1) and (2), so it does not count against the account. Decides: the FBI file; Salt Lake City police records for March 1940; the papers of S. E. Bringhurst; year-faceted newspaper searches for March 1940 and the "two other occasions".

E. The Research Institute's gold and uranium work, and the 1968 Air Force proposal, after 1970. What the record says: the Research Institute, Inc. (Salt Lake City; John E. Moray, with Hooper, Hendricks and Rudolph) ran funded irradiation work on rented outside accelerators (High Voltage Engineering, Electronized Chemicals, Radiation Dynamics, Varian) and a 1964 pilot plant of about 800 lb of ore, presented it at the National Western Mining Conference (1964, 1965), and reported assays by "a wholly independent firm" and "an independent assayer company" that it never names [read: Rex pp.2-3, 5-7, 12-13, 17]. Its statistician (Olsen) reviewed 62 tests in 1966-67 and in 1979 projected a plant producing about 238,000 oz of gold a year, with break-even at 0.727 oz/ton and about $6.4 million [read: Rex pp.18, 20]. The last test data in the folder are seven runs of Aug-Oct 1970 [read: Rex p.19]; John Moray's statements of 1981-82 discuss the "environment" and the bombardment tube but report no new test [read: Rex pp.21-22]. The uranium assays date from 1956-58 [read: Isotopes deep read; Rex pp.12, 16]. The 7 Aug 1968 proposal to Air Force Systems Command (T. H. Moray and Stan Taylor) says working models were "demonstrated to competent observers" [read: gov-Airforce]; the folder holds no reply. Not found: any journal paper; any conference paper outside the mining conference; any patent assigned to the Research Institute or to John E. or Richard R. Moray [web research/2026-10-03-google-patents.md]; any named assayer, certificate, plant, mining lease or gold sale; any outside replication; any record of what the Air Force did with the proposal. A process its own consultant priced at millions of dollars of gold a year would be expected to leave such records, so the silence is an anomaly, not a neutral gap. Competing explanations: 1. The results were artefacts that could not be reproduced, so the work faded. For: the Institute's own records show failed duplication (six runs at 4-7 oz/ton), 3 of 9 pilot runs under 5 oz/ton, "very few" worthwhile linac results after Varian, yields that scale inversely with sample mass, and inconsistent statistics [read: Rex pp.6-7, 17-19]; the energy budget and nuclear argument (Claim 5b). Against: the Institute kept presenting the process to 1979-82 and published no retraction; nothing in the record shows the team concluding that it failed. 2. Commercial failure or lack of capital. For: the 1979 projection needed about $6.4 million before break-even [read: Rex p.20]; the Morays' earlier ventures ended in litigation over money (Moray Products, Reno 1932; the 1942 Sanford judgment) [web, see Claim 6]. Against: no record of an attempt to raise that money, or of investors declining, has been found. 3. Trade secrecy. For: the "environment" was withheld ("This report will not deal with the chemical composition of the test solution") [read: Rex p.12]; no patent was sought, which fits a secret process. Against: secrecy hides a recipe, not an output; gold sold or a plant run would still leave commercial and tax records. 4. Absorption into another company or a change of name. For: nothing found yet, but only patents (by assignee) and Utah newspapers ("Research Institute" Moray) have been searched [web research/2026-10-03-google-patents.md; research/2026-10-03-utah-digital-newspapers.md]. Against: no successor found. 5. Classification or official suppression (bears mainly on the 1968 proposal). For: the proposal went to a military command, and the family's account of earlier official interference (Claim 6). Against: no document links the gold or uranium work to any agency after the 1956 Bureau of Mines and AEC evaluations; the 1963 FDA action concerned the Cosray device, not the gold work. Records that would decide: the Utah Division of Corporations file for Research Institute, Inc. (officers, status, dissolution, successor); the assayer's certificates and the firm's name; accelerator facility logs or invoices (Varian, Radiation Dynamics, High Voltage Engineering, Electronized Chemicals) showing runs and doses; the 1964-65 National Western Mining Conference proceedings; Hooper's papers (Principia College); the 1968 proposal's file and reply in Air Force Systems Command records (NARA); mining-press coverage 1963-82; the 1974 Salt Lake County probate file, if it lists Institute shares.

G. Moray's own laboratory 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; no Moray experiment with any of them, and no silver or cadmium assay, is in the record. Most of those pages are Le Bon, Glasser and Preston (section 1 of "Investigator findings (2026-10-08)"). This is an anomaly to analyse, not a neutral gap. Competing explanations: (1) he ran none and copied the physics to explain a device built by other means; (2) he ran some and the notes are in his notebooks, held by the family (Perreault says the widow gave a notebook to Stan Mahurin); (3) the pages were compiled after his death by the family or a typist from books he owned (the printing labelled "1926" cites the atomic pile and was compiled after 1945). For (1): every passage traced has a printed source and none carries a datum of his. Against (1): the insertions "and sometimes other properties", "Herein be the door to greater discoveries" and the silver-to-cadmium sentence suggest he thought in these terms, and the sons say he studied "synthetic radioactivity as described by Gustave LeBon". For (3): the dating. Silence in other sources is not contradiction. Records that would decide: Moray's notebooks and the Mahurin material; the Ottawa Bray fonds (letters 1928-56); the 1943 and 1956 printings of Radiant Energy.

Corrected claim list (what the claims should read as)

1. Between 1925 and 1938 Moray showed invited witnesses (about 30 named, many associated with him, some state and REA officials) a box with an antenna and a ground wire that lit lamps and heated a flat iron. Witnesses recorded nameplate loads of 0.1 kW to about 8 kW (about 5.5 kW for a state Securities Commissioner's inspection in 1938), including a sealed 157.9 h run (two segments) at about 110-120 W ending at 635 W. The bulbs often glowed oddly (a dime-sized hot spot), so delivered power is unknown. No power was ever metered. The 50 kW figure has no witness. [read, via the sons' book]
2. Mains induction is unlikely at the remote sites; 1920s batteries are disfavoured for the 1928 and 1938 tests if the lamp loads were real; a hidden lead, a mains-bonded water-pipe ground and priming charge were not excluded. Checks existed (Yates, Welling, Jensen) but none measured power, and the detector was withheld in every recorded examination.
3. The proposed source (oscillating cosmic radiation) would need an energy density 1e9 to 4e9 times the measured cosmic-ray density (or 37 times full sunlight); the book's own resonance equations have no energy source, and Moray's own text says the first stage is excited by an external source. Added 2026-10-04: Tesla's 1901 antenna-ground-condenser collector gives about 2e-10 to 8e-9 W per square metre of collector from fair-weather atmospheric electricity (about 5e-6 W for 1,000 m2), and a resonant antenna in an ordinary radio field needs about 1 V/m to supply 1.4 kW; zero-point-energy readings give no frequency or flux. Added 2026-10-08: photoelectric collection (Plauson's amalgam patches on a balloon, US 1,540,998) gives nanowatts to a fraction of a watt per 1,000 m2 in full sun; Compton scattering degrades photon energy and cannot concentrate it; neutrino capture is short by 1e17 to 1e21, and Weber's claimed 1e23 enhancement in crystals was tested at least four times with null results; the stellar checks (Borexino, N_eff, helioseismology) bound a hidden channel to a few per cent of the Sun's output without excluding it. The 1968 proposal's infrared paragraph is 1968 detector practice, not a reference to the microwave background.
4. The detector is an unexamined mineral or semiconductor stone; a rectifier cannot create power; no gain was shown; Hayes's attempt to rebuild from the patents failed; performance claims fall about 600-1,000 times between the 1930s and 1968. Added 2026-10-04: crystal amplifiers with a bias supply existed earlier (Losev 1923-24; Lilienfeld's patent 1925/1930), so priority is not unique and gain without a supply is untested. Added 2026-10-08: direct conversion of decay particles is real (Moseley 1913, about 1.5 microwatts) and needs at least 58,000 Ci per kilowatt (about 1.7e5 Ci at a mean beta energy of 1 MeV), with radiation at the output power, which no source reports; radon-daughter plating on a negative tube electrode is a real effect (Rutherford 1900) that would leave Pb-210; the stone may have been a radium-doped phosphor-type material (family treatise 1946, Noffsinger 1984 list, Foster 1990 pressing: three family- or advocate-side items, no measurement). The Foster flash is joules of stored energy or a short circuit of the press supply (single-source account).
5. "Slow matter dissociation" is unquantified. The 1963-70 gold results fail an energy budget (1.25e8 J per gram of residue needed, at most about 200 J per gram delivered) and a nuclear argument, have one thin control pair, inconsistent statistics and no outside replication. The uranium "upgrade" (0.04-2.6% to 21-67%) fits discarding 96-99.8% of the mass or radiometric artefacts. Cosray has an uncontrolled, arithmetically inconsistent tally of 785 patients and a contested FDA labelling default, never a clinical trial. Added 2026-10-04: the family's own 1946 treatise describes the applicators as emitting alpha, beta and gamma rays with a half-life of about 1,600 years (radium-226's), penetrating the body and half an inch of lead, and needing no shielding; alpha and beta cannot penetrate so, and the gamma component is a hazard no dose figure addresses. Added 2026-10-08: "matter dissociates" and the pages on Rutherford, screens, thorium deposits, phosphors and amalgam discharge are Gustave Le Bon (1905), reproduced without his footnotes or credits (Radiant Energy pp.53-59; Isotopes pp.4-6); Isotopes pp.20-25 follow Glasser (1944) and Preston (1928). They show what Moray read, not what he measured. Le Bon's thesis is historically superseded but instructive (his amalgam observation was the photoelectric effect, repeated by Ramsay and Spencer in 1906); the plain energy-from-mass point is established. "Silver to cadmium by resonance of current, rays and sound" is refuted by calculation (9.8 MeV barrier, quanta of eV or less) and has no datum.
6. The patent rejections (inoperability and no known natural source, the examiner's grounds) and the FDA action (a labelling charge, decided by default) are documented. Whole-device tests were offered by critics and by Moray's side and not completed by either; the REA's own 1938-41 tests left reports that are now missing. Moray's REA engagement (1939-41), Frazer's "threatening" and the 1929 Soviet contact are documented outside the family's books. The 1939 destruction of the device and the 1940 shooting are documented in one detailed published account (the sons' 1978 book) and have not yet been checked against the FBI, REA, Senate and police records that would confirm or contradict them. Added 2026-10-04: the printed 1944 Senate hearing documents the REA payroll (17 Feb 1939 to 28 Feb 1941), Wood's report ("essentially honest ... stumbled onto something that he himself cannot understand"), Frazer's two years "all to no avail", a physicist's (Tugman, 1940) 12-year refusal to be shown a demonstration, and a circulating story of an assassination attempt; all from interested parties and none a measurement. An Army Ordnance officer (Shinkle) filed a report on a "sound detector" in Dec 1936 (report not found). Added 2026-10-08: the family's statement that Le Bon's book was withdrawn from libraries and that the FBI asked about it in 1942 is family-only; the folder's "1926" printing of Radiant Energy was compiled after 1945.
7. The Uppsala doctorate rests on Moray's own account; the only search made used the wrong name, so no document yet supports or contradicts it.

Summary table

What would settle it

1. The decisive experiment (Claims 1-3): a power meter, spectrum analyser and oscilloscope on the antenna and ground leads; a mains-isolated, shielded room wired by a sceptic; a driven-rod ground chosen by the sceptic (not a water pipe); calorimetry on the box; the same trunk-and-seal protocol; 24 h or more; an antenna-less run (Fletcher's 1928 point 3) and a dummy-detector blank. It shows whether power enters through the leads (hidden or external feed), is released from a primed store, or has no input. All three reviewers converge here.
2. Metered priming audit: a recorded energy budget of the start-up (magnet stroking, terminal shock, electrostatic plate) against the run output, to test "externally excited, then self-sustaining". A decaying primed oscillator and a driven one separate on total energy.
3. Knudsen's hidden-wire test: inspect and trace the antenna pole and lead-in; ground-path test by isolating the building's bonded water pipe.
4. Remote field run with a duration and meters: a site chosen by a third party, a recording ammeter on a battery-free load; replaces the 1938 Dugway recollection. A field survey (RF and other flux) at the site tests the unknown-radiation form of Claim 3.
5. Open the detector at a neutral laboratory (X-ray, composition, rectification and gain bench test), compare with the 1949 patent's tubes, and test whether a rebuild is possible.
6. Blind gold re-test: split sample, unirradiated reagent blank, environment-solution-only assay, irradiated no-solution control with replicates, a named independent laboratory with certified standards, activation and isotopic analysis. Uranium: split homogenised ore, masses before and after, chemical (not radiometric) assay, U-235/U-238 isotopics.
7. Controlled clinical test of the Cosray device against a sham.

8. Primary documents that the sons' books cite but do not reproduce: the Securities Commissioner's 1938 report; Crim's and Frazer's REA reports; any Fletcher or Bell Labs telegram or correspondence about a New York test; Hayes's Reno testimony (Nevada court records); the Patent Office files of the seven applications; the 1940 USDA order; the Boaz Mine and Union Assay certificates; the Bureau of Mines and AEC reports; Moray's Uppsala record under "James Cain Larson". Added 2026-10-03: the 74-page FBI file (FOIA request 917273, released 2000); the printed 1944 Senate Agriculture and Forestry Committee REA hearing and its exhibits (Wood's report, Frazer's testimony) in NARA RG 46; REA Administrator's files 1938-41 in NARA RG 221; the Washoe County case file "In re Moray Products Company" (1932); the Salt Lake County probate file for T. H. Moray (1974); the 1970 Hagmann case file (195569); Salt Lake police records for March 1940; a gamma-spectroscopy and composition examination of the tubes Perreault says he holds. Added in the 2026-10-03 recheck (Silence E): the Utah corporate file for Research Institute, Inc.; the gold assayer's name and certificates; accelerator logs or invoices for the 1963-70 runs; the 1964-65 National Western Mining Conference proceedings; the Air Force Systems Command file and reply for the 7 Aug 1968 proposal.

9. Added 2026-10-04 (Investigator's leads): Shinkle's report of 8 Dec 1936 (NARA RG 407, AGO decimal file 413.68; 095 Steffensen name file); Wood's full August 1939 report and the REA project files (RG 221; Carmody Papers boxes 91-94, 99, 102, 103, FDR Library); the Ottawa Bray fonds files F4, F24, SS4-F4, F6 (contemporary letters outside the family's edits); the 2000 release of FBI 917273 and CIA F-2000-01342; gamma spectroscopy of any surviving Cosray applicator and of Perreault's tubes (radium-226 or not); the 1970 Hagmann file (195569) and a question to Hagmann; Alter (1932) pp.439-440; the 1931 physicist's name. Experiment: a fair-weather measurement at a driven-rod site of what a 27 m antenna and ground deliver into a matched load, to compare with the roughly 5e-6 W (1,000 m2) estimate; and a metered motor-load test (shaft power) in place of lamp counts.

10. Added 2026-10-08 (follow-up investigations): gamma and alpha spectroscopy and a ZnS:Ag screen on a surviving tube cathode or coating and on any Cosray applicator (Pb-210 at 46.5 keV, Po-210 alpha, Ra-226 daughters): tests radon plating, Perreault's "no radioactive material" and a radium-bearing stone together; a radiation survey beside a running device and an assay of electrode and process metals for cadmium in silver, with a blank (any kilowatt claim from nuclear conversion needs radiation of at least that power); Foster's 1990 letter, the press log and the contractor of "Project X"; Moray's notebooks for "and sometimes other properties" and any screen, amalgam, thorium-plate or silver/cadmium experiment; Moseley's full 1913 paper and Rutherford's 1900 paper; a blank-controlled repeat of one Le Bon amalgam experiment with spark-strength ultraviolet (the part Ramsay and Spencer could not test); Nye (1974), the Athenaeum correspondence (pp.237-366), the complete Glasser and Preston texts and the Le Bon printing Moray's editor used; the 74-page FBI file for a 1942 Le Bon entry. The sceptic's artefact route for 5b (radium-daughter plating on a negative electrode reading high radiometrically) is a control the data lack.

Open questions

• Was the antenna lead, the pole or the building wiring traced and metered in any of the 1928-38 events? Was the 157.9 h run continuous and witnessed (it was two segments with the trunk opened between)?
• Did the start-up priming store energy, and could it explain minutes but not 158 h? Was the 1938 meter reading taken before or after start-up (the 4th ed. says before)?
• What did the bulbs actually dissipate? A calibrated power measurement on one lamp would convert every lamp count into watts.
• Who declined the New York test, and is there a telegram? Is the Hayes Reno testimony in the court record, and what exactly did it say?
• Why did the claimed performance fall about 600-1,000 times between the 1930s and 1968, and why did the 1962 and 1960 yield statements differ (2 versus 10-20 in 100)?
• Are the Boaz Mine and Union Assay certificates real, and what was the "before" value for the Bureau of Mines uranium sample?
• Are the Noffsinger 1984 and Wachs 1955 detector lists legible and accurate (the 1984 note says "snow job also")?
• What did the 1943 rebuild, the 1957 demonstrations (about a dozen witnesses) and Fate 1956's "57+ hours" run involve? These are digest-only here.
• Did a Swedish record exist under the alias, and what precisely did the "doctorate" consist of?
• (Added 2026-10-03) What is in the 74-page FBI file: a 1940 report of the shooting, the 1942 LeBon enquiry, the 1957 clearance? Does anything there name an FBI employee present on 2 Mar 1940?
• What did Wood's full report and Frazer's Senate testimony say about the tests and about the destruction of the device? Did Frazer's report survive in RG 221?
• What remained of the apparatus at Moray's death in 1974, who received it, and are Perreault's tubes Moray's? What device was demonstrated in 1957 and 1963, after the sons say the detector material had burned out?
• Is the Felix J. Frazer of the REA episode the Technocracy economist of the same name?
• (Added 2026-10-04) What did Shinkle see in Dec 1936: the receiver only, or the lamp unit? What did the "small motor and fans" in Craig's account draw, and at what frequency? How long and with what output did the REA tests run, and why was the engagement ended (Slattery says a senator's opinion, and agreed the claims were "pretty similar" to those of the convicted "Count" Tomadelli; the investigators say negative conclusions)? What is in the Ottawa files and do they quote the 1928-38 witnesses as the family books do? Are the Cosray applicators and the valve tubes the same material, and is it radium? Why does a 74-page FBI file released in 2000 return "No Records" in 2013? Was the 1934 "100 lights for 154 hours" a count of small lamps (about the primary 115 W) or of 100 W lamps? What was Wood's own basis for "stumbled onto something"?
• (Added in the 2026-10-03 recheck) Why does the Research Institute's gold work go quiet after 1970, with no named assayer, patent, plant or outside replication, and what did the Air Force do with the 1968 proposal? (Silence E.)
• (Added 2026-10-08) Who prepared the printings of Radiant Energy, Sea of Energy and Isotopes, and why are Le Bon's, Rutherford's and Crookes's credits and footnotes missing? Did Moray run any sulphide-screen, amalgam, thorium-plate or silver/cadmium experiment (Silence G)? What did "and sometimes other properties" mean to him? Is Foster's 1990 pressing documented anywhere (letter, press log, contractor), and did the flash come from stored energy or a short? Do surviving tube cathodes carry Pb-210 or Ra-226 daughters? What is the full text of Moseley's 1913 paper (150 kV and 0.01 nA)? What did Le Bon's spark-strength ultraviolet do that Ramsay and Spencer's lamps could not reproduce? Which printing of Le Bon, Glasser and Preston did Moray's editor use? Is the family's statement that the FBI asked about Le Bon in 1942 in the 74-page file?

Changes after QE (verdicts that changed from the previous arbiter.md)

• Claim 1: label unchanged (Never properly tested), but the factual basis changed. "4 kW and above never observed" and "record covers 1925-28" are withdrawn; nameplate loads of 1.5-8 kW (1930-38) are recorded; only 50 kW is unwitnessed. The battery argument is now stated as conditional arithmetic (strong for 1928 and 1938, nil for Strawberry). Added: the non-neutral rebuild failed (Hayes 1932), the Securities Commission inspection, the capacitor test, priming ("excited externally").
• Claim 2: label unchanged (Never properly tested), but "critics did not insist on instruments" and "no instruments" are withdrawn (Yates, Welling, Jensen checked, none measured power). Classification extended with "test offered, inventor withheld" and "critics' offers not carried through"; Jensen's "electrically dead" reading was taken before start-up and is weak evidence.
• Claim 3: verdicts unchanged. Added: Moray's own "externally excited" statement; Moray's own 1,193,600 per-person figure (1.4 times sunlight rather than 1.95); a new "primed/stored/hidden-feed: Never properly tested" part; Nuenergy rebutted.
• Claim 4: verdicts unchanged; added the failed Hayes rebuild, Perreault's tube inspection (one published source, low weight, no instrument or method given), and the performance-figure inconsistencies.
• Claim 5: split into 5a-5d. (i) Gold: still Refuted by calculation. The re-run briefly withdrew the energy-budget refutation on a figure of 125 J per gram of residue; that figure dropped a factor of 1e6, and the correct value is 1.25e8 J per gram [computed], so (corrected 2026-10-03) the verdict rests on both the energy budget and the nuclear argument. "No blanks" and "Institute's own assayers" are corrected (two thin controls; unnamed independent assayer); assay claim Never properly tested. (ii) Uranium upgrading is new: creation Refuted by calculation; upgrade claim Never properly tested (fits 96-99.8% mass loss or radiometric artefact). (iii) Cosray: still Never properly tested, but "no clinical data" is corrected to an uncontrolled, inconsistent tally of 785; USDA item is a build order; the FDA default was contested and vacatur denied, so "finding of no therapeutic value" is withdrawn. (iv) Mass-loss figure stays 17.5 mg per year.
• Claim 6: event verdicts unchanged (Established as history); "no known natural source" is a stated examiner ground (earlier "sons' gloss" claim withdrawn). "Dismissed without testing" is replaced by a standoff: Never properly tested, with tests offered and not completed on both sides, a disputed Fletcher New York test, and a failed non-neutral rebuild; only some parties (Millikan, NASA, AEC, Patent Office) dismissed without testing. NASA/Bennett 1961 both sides added.
• Claim 7: unchanged.

Web verification log (budget 8; 1 new check this round)

• 2026-10-08: see section 7 of "Investigator findings (2026-10-08)" (Atomic battery page: Moseley 150 kV and 0.01 nA; Silver-108m isotope table; six local checks of quotations against the Investigator's saved pages).
• New: Li-ion typical specific energy 160-300 Wh/kg (up to 450 specialty), so the earlier "[memory, medium] 250 Wh/kg" is confirmed [web https://en.wikipedia.org/wiki/Lithium-ion_battery].
• Confirmed earlier (this review chain): lead-acid 35-40 Wh/kg [web https://en.wikipedia.org/wiki/Lead%E2%80%93acid_battery]; MMRTG about 2.8 W/kg [web https://en.wikipedia.org/wiki/Multi-mission_radioisotope_thermoelectric_generator]; proton lifetime [web https://en.wikipedia.org/wiki/Proton_decay]; Seaborg 1980 Bi to Au [web https://en.wikipedia.org/wiki/Nuclear_transmutation]; antenna aperture [web https://en.wikipedia.org/wiki/Antenna_aperture]; solar constant, cosmic-ray density, transistor 1947, Knudsen at UCLA [web https://en.wikipedia.org/wiki/Vern_Oliver_Knudsen], Pb-210 half-life, General Order 40 date, US 2,460,707 [web https://patents.google.com/patent/US2460707A/en], betavoltaics (reviewers' fetches; I did not re-fetch them).
• Still [memory], not web-checked: Coulomb barrier scale and photonuclear behaviour (high for "MeV to tens of MeV", medium for "100 MeV"); fire-assay flux gold and reagent blanks (high); water pipe bonded to the service neutral (high); radiometric U3O8 equivalent and radium daughters (high); cold-cathode tube history (high); patent utility doctrine (medium); lamp brightness exponent and flat-iron mass (medium, low-medium; used only for a qualitative point); global fair-weather current 1,800 A (medium); germanium purity in 1927 (medium). No central verdict rests on one of these alone.
• 2026-10-04: one web check (Half-value layer, Wikipedia, partial confirmation only); other new [web] items are the reviewers' and Investigator's, listed in section 6 of "Investigator findings (2026-10-04)".
• 2026-10-03 revision: two web searches attempted (the FBI file 917273 on FOIA archives; the printed 1944 Senate REA hearing). Both failed: the session's web-search budget was exhausted. New [web] items in this revision come from the Investigator's notes (sites.md, research/2026-10-03-*.md), which give the URLs; I did not re-fetch them. Still [memory], not web-checked: that the FBI Vault posts only a selection of released files (medium); that the Patent Office routinely disposed of abandoned applications (medium). No verdict rests on either.

Evidence-rule revision (2026-10-03)

Audit of every claim in this file against the three SOURCE HONESTY rules (no suppression of evidence; interested parties; absence of evidence), using sites.md and the research/ notes. Changes:

Verdict changes 1. Claim 6 (iii): changed. The 1939 destruction of the device, the 2 Mar 1940 shooting, the later attacks and the 1943 threat move from Not testable as stated to Never properly tested. Reason: they are specific, dated claims with named people, documented in one detailed published account, and they can be checked against records that exist or should exist (FBI file 917273, NARA RG 221 and RG 46, the 1944 Senate hearing, Salt Lake police records). Nothing found contradicts them. Only the broad "organised conspiracy" form stays Not testable as stated. 2. Claim 6 (i): extended. Moray's REA engagement (Feb 1939-Feb 1941, $25 a day), Frazer's role and "threatening", the Wood report and the 1929 Soviet contact (FBI memo, 1951) are added as Established documented history, from records outside the family's books. No other verdict changed: none of the newly weighed evidence includes a measurement, so the physics verdicts (Claims 1-5) stand.

Dismissal by label, removed 3. Source gap: added that family provenance is to be weighed, not treated as disqualifying, and listed the new outside sources; "rest on friendly testimony" became "testimony of named observers", noting the state and REA officials among them. 4. Claim 1: "treat the testimony as a fact about the documents, not about the device" is withdrawn. The testimony is now weighed as strong evidence that the effects were seen, and as unable to settle the energy source. 5. Test classification and Claim 1 verdict and summary table: "Witnessed by friends" renamed "Witnessed by named observers". 6. Reconciliation table (Fletcher New York test): "I give it no weight in either direction" became a weighing of two conflicting first-hand accounts, each from a party with a stake. 7. Eyring's wattage-test refusal: "John Moray's account only" became "documented in John Moray's account, no second source found yet". 8. Hayes rebuild: noted that this evidence against Moray is also from one source and is weighed on the same terms; the court that holds the record is now identified. 9. Claim 4: Perreault's tube inspection, "hearsay ... not evidence either way", is rewritten. It is quoted as one published first-hand source: evidence that tubes survived Moray's death and that a casual check found no apparent radioactivity. Its low weight is explained, and a test of the tubes is proposed. 10. Claim 4 (ii): "assertions only" became a statement of what the evidence is. 11. Claim 5c: "Dr. Marven ... is hearsay" became a recorded remark that does not bear on the question. 12. Claim 5d: "an anecdote" (Senate lessee) and "one handwritten anecdote" (Betz) are replaced by quoted accounts with their sources. Added the 1949 newspaper report of cancer claims. 13. Claim 6: "rest on the family's account ... show no suppression" is replaced by a quoted, weighed account of the destruction, the shooting and the threats. Added Frazer's counter-claim (Moray destroyed it himself), Bearden's conflicting "Soviet agent" version, and the 1944 Senate corroboration of Frazer's role. Added the FBI's 1951 memo, which corroborates the Russian interest, and noted that silence on the violence in that memo is not contradiction. 14. Claim 7: "unsupported" became "no document yet supports or contradicts it", with the record that would decide. 15. Claim 3: Nuenergy "one-author reconstruction" became a sourced description with Perreault's stake.

Interested-party characterizations, now attributed 16. Patent Office grounds are presented as the examiner's characterization beside Moray's own description, with the family's relayed counter-allegation ("honeycombed"). 17. FDA: the misbranding charge is presented as the prosecuting agency's characterization, decided by default, not a finding about the device. 18. NASA, Senator Bennett and the AEC (Larson relaying Eyring): flagged as the characterizations of parties with institutional stakes. 19. Millikan and Bell Labs: the "concealed part" reason is marked as their stated reason. 20. Knudsen's hidden-wire suspicion is marked as his, with the Morays' counter-claim that his switching burned out the valve. 21. REA evidence added with stakes: Crim and Craig (REA officials; Craig's letter, 5th ed. p.119), Wood (REA engineer; report quoted in 1944 Senate testimony), and Halleck (a political opponent of the REA who made it public). Wood's "essentially honest, unable to ... perpetuate a hoax" is weighed against the deliberate-fraud hypothesis, not against priming or observer error. 22. Business and court record added because the rule cuts both ways: Washoe County 1932, Sanford v. Moray 1942 (a judgment against Moray), and the 1970 Hagmann suit beside Hagmann's printed endorsement.

Silences analysed (new section "Silences in the record") 23. A: the device and detector after 1939 and after 1974. This sets out five competing explanations (destroyed and not rebuildable; trade secrecy; seizure or suppression; never worked; transfer abroad) with the evidence for and against each. It notes the tension between "detector burned out" (1943) and the demonstrations recorded in 1957 and 1963, and lists the deciding records (probate, family archive, Perreault's tubes, the Hagmann case, the clearance hearing). 24. B: the 74-page FBI file (917273), which is not on the Vault. Covers routine, mundane-contents and violence-related explanations; notes that the FBI is an accused party; the deciding record is the file itself. 25. C: missing REA reports (Craig: "lost or destroyed" in the 1941 move), the 1944 Senate hearing (a gap in our access, not a silence in the record; it conflicts with the family's 1942 date), the Securities Commissioner's report, the Hayes testimony, the USDA order, and the patent applications (the Patent Office says "destroyed" as abandoned; the sons say "disappeared"). 26. D: no police or press record of the 1940 shooting. The family's own account predicts this silence, so it does not count against the account. 27. Claim 2: Frazer's 1939 twelve-sheet glass test was added and assessed as a series-capacitor check that does not separate the hypotheses. 28. "What would settle it" #8, the open questions, the corrected claim list (1, 6, 7) and the summary table (rows 1, 6, 7) were updated to match.

Evidence-rule recheck fixes (2026-10-03, after qe.md "Evidence-rule check") 29. Claim 4 in "Changes after QE": "the hearsay tube inspection" replaced by "Perreault's tube inspection (one published source, low weight, no instrument or method given)", matching the Claim 4 body. 30. Claim 5b gold energy budget: 124-125 J per gram corrected to 1.25e8 J per gram of residue [computed], with the arithmetic in the reconciliation table; the withdrawal of the energy-budget argument is itself withdrawn. Changed in the reconciliation table, 5b "What changed", 5b refutation and verdict lines, summary table row 5b, corrected claim 5 and "Changes after QE". The verdict (Refuted by calculation) is unchanged; it now has two independent supports. 31. New Silence E (the Research Institute's gold and uranium work and the 1968 Air Force proposal after 1970), with five competing explanations and the deciding records; added to "What would settle it" #8 and the open questions.