Lakhovsky George: Arbiter (Step 3)

Date: 2026-10-08. Inputs weighed: claims.md, facts.json, physicist.md, mathematician.md, contrarian.md, deep/, sites.md and research/ (Investigator). Tags: [read] project file or Investigator file, [web] with URL, [calc] my own arithmetic (numpy), [memory] with confidence.

Source gap note

No source gap in the strict sense: all six folder documents have readable text. Limits that every later file must repeat: - None is a primary laboratory or clinical record. The cited primary papers (C. R. Soc. Biol. 91:626, 1924; Roffo; Metalnikoff; Nodon; Rivera; Rigaux; de Cigna; Attilj) were not obtained. What they found is Lakhovsky's report, except the 1924 note, which exists in the file set only as a quotation in a 1925 press article on a seller's site (colysa.ch, a stake) and as a citation in Kotzareff 1931 and Roussakow 2013. BHL, Gallica and SUDOC were blocked to the Investigator. - The Secret of Life copy has no figures and carries editorial layers ("Hatonn", "E.Y.") that are not Lakhovsky. - Claims 5 (soil section) and 6 rest partly on books the reviewers read only through the Investigator or in part (L'Universion, Contributo, Pour rester jeune). The mathematician read the Contributo table and L'Universion scale directly; nobody read the end of the "Child of the Waves" account. Claim 6 therefore rests on the 1941 book and The Secret of Life alone. - Where a conclusion rests on [memory] alone I say so below.

Where the three reviewers agree (and so I do not re-argue it)

• Lakhovsky's arithmetic is usually right (2 m = 150 MHz, 400 m = 0.75 MHz, 10 cm = 3 GHz, the skin-depth figures of The Secret of Life p.53, the 1924 durations); the problems are in the model it is applied to.
• The strong cell-circuit model fails by his own stated materials (damping). All three computed it independently and agree.
• The Roffo, Metalnikoff and thorotrast citations support mainstream radiocarcinogenesis, not oscillation; "radio-active" is Lakhovsky's word, Roffo wrote "photo-active and luminescent agent".
• The 1924 geranium note is the single concrete experimental datum, is not an anecdote, and has never been repeated with temperature recorded. The 1940 Nature review (J. Ritchie) separated the clinical reports from the theory: the oscillator "may work and apparently has worked marvellous cures in cases of cancer, goitre and enlarged prostate", which "does not prove that its electric radiations acted upon hypothetical cell-oscillations" [read: Investigator, research/2026-10-08-internet-archive.md]. I keep both halves on the record.
• Case histories have no denominator, are from the maker and his associates, and are internally inconsistent in small details.

Recomputed numbers (where reviewers differ or a verdict leans on them)

Corrections to earlier documents

State plainly, as the owner asked: 1. claims.md claim 4: "US 1,962,565, 1931" is filing and priority only; granted 12 June 1934 (filed 13 Nov 1931; French priority 2 May 1931). Also "no wavelength, voltage or medical claim" is true of the US patent only; the Swiss family member CH 164027 (1933) names therapy and an "artificial sun" [read: Investigator; web https://patents.google.com/patent/CH164027A/fr]. Duodenal ulcers and lupus are summary statements by named clinicians, not case histories. 2. claims.md claim 2 and facts.json: Roffo's agent is "photo-active and luminescent", not "radio-active". Species of the 55 percent figure differs between the two books (mice vs rats). 3. claims.md claim 3 and facts.json: the 1924 note has four authors (Gosset, Gutmann, Lakhovsky, J. Magrou of the Pasteur Institute), reports 3 treated plants against 16 controls with 6, 33 and 9 hours exposure and onset of necrosis about 16 days after the first session in each; it is C. R. Soc. Biol. 91:626-628, 26 July 1924. "Caused tumours to necrose" should read "in three treated plants". The deep-read "undated" is superseded. The 1941 book's "fifteen days" is Lakhovsky's later summary; the note's 15 days is the time from onset to complete necrosis. 4. claims.md claim 5: the lead sphere is "25 m in diameter" and then "25 m thick" in one passage; the penetration limit is 2 m (L'Universion 1927, for a screen that barely suffices), 7 m (The Secret of Life, objection as quoted) and 11 m (1941). The protective age is 40, 45 and 50 in different places. Hertzian row of the Rivera table: see recomputed table row 8. 5. The Secret of Life, bracket "[approximately 100-200 million cycles per second]": wrong (2-10 m is 30-150 MHz), and an editor's insertion. 6. physicist.md, claim 4: ring inductance for the stated 0.5 m ring (table row 5). 7. physicist.md, Web checks, and contrarian.md, Claim 2: the Kirby citation. PMC2008681 is Hieger and Orr 1954, "On the carcinogenic activity of purified cholesterol", which does not mention Roffo [web: https://pmc.ncbi.nlm.nih.gov/articles/PMC2008681/]. Kirby is PMC2007542 (A. H. M. Kirby, "Biological tests with cholesterol esters of unsaturated acids", Br J Cancer 2:70-74, 1948) [web: https://pmc.ncbi.nlm.nih.gov/articles/PMC2007542/]; the page text is images, so the statement that Roffo's 1938-41 stomach adenocarcinomas in rats fed fat or cholesterol heated to about 350 C "have not been confirmed" (Beck and Peacock 1941; Morris, Larsen and Lippincott 1943; Kirby 1943-45) rests on a search excerpt of its opening paragraph, medium confidence. Kirby adds that other routes in mice gave positive results with such pyrolysis products; that belongs in the record. Contrarian's "tested by others, not supported" is about the heated-fat variant, not irradiated cholesterol. 8. physicist.md and contrarian.md, Claim 5: "cosmic muons were detected under hundreds of metres of water by the 1930s, so 11 m of lead was not the state of knowledge" is only partly right. Regener's Lake Constance work reached 231 m of water, ionisation 0.05 ion pairs per cm3 per s [web: https://arxiv.org/pdf/1411.6217]; on Lakhovsky's own 1927 conversion (35 m water = 1.8 m lead) that is about 12 m of lead, i.e. his 11 m. By density it is about 20 m. His water-lead equivalence is itself not the density ratio. So the 11 m figure was not contradicted by Regener on his own terms; either sphere dimension (12.5 m radius or 25 m thick) exceeds it, and the decisive test is the one below, not the 1930s depth data. That the 1930s ionisation was by muons is later understanding [memory, medium; the search could not confirm it]. 9. contrarian.md, Claim 1 and mathematician: the TTFields EF-14 figures (median overall survival 20.9 vs 16.0 months, HR 0.63, P<0.001) are confirmed from secondary coverage; the JAMA paper itself was not opened [web: https://www.novocure.com/jama-publishes-final-analysis-ef-14-phase-3-pivotal-trial-optuner-together-temozolomide ; a press release, a stake; and https://www.onclive.com/conference-coverage/aacr-2017/tumor-treating-fields-substantially-improve-survival-in-gbm ]. Note: that evidence is a different mechanism at 200 kHz. 10. physicist.md summary says Roffo's UV skin cancers were "corroborated by other authors" from a 1939 Nature note; the Nature page would not open (login redirect), so that rests on a search snippet [web: https://www.nature.com/articles/144246a0], medium confidence.

Corrected claim list with final verdicts

Claim 1. Every living cell is an electrical oscillating circuit (chromosomes and chondromes as insulated sea-water-filled filaments, resonating in the infrared, powered by cosmic radiation)

Reviewers: all three give Refuted by calculation for the strong form and a softer plausible form. Disagreement: wording of the soft form (physicist "Plausible, untested"; mathematician "Not testable as stated"; contrarian "Plausible, partly supported by others"). Weighing: the strong form is refuted by Lakhovsky's own stated filling: Q = omega L / R is 8e-7 to 4e-6 and the resistance is 1e5 to 6e5 times critical [calc]. His text prints no L, C or dimension, so the only numbers are the reviewers' reconstructions; but any reconstruction with salt water inside fails by orders of magnitude. The energy supply fails separately (Claim 5 heat budget). The Nature reviewer's identification of the Corynactis "circuit" as a cnidocyst thread is a contemporary specialist objection to the very figure used as the model [read: Investigator]. The soft form (cells emit and respond to weak fields; ultra-weak photon emission exists; TTFields act on dividing cells at 200 kHz; Frohlich-type coherence is a contested proposal) is a different, vaguer idea. Mathematician is right that without a frequency, amplitude and coupling it has no test; physicist and contrarian are right that parts of that neighbourhood are live research. A literal reading of "cells respond to specific frequencies" is partly testable and was never tested by Lakhovsky. Final verdict: strong form (LC filament, cosmic power): Refuted by calculation. Soft form (cells interact with EM fields at selected frequencies): Plausible, untested as Lakhovsky meant it; Not testable as stated in his version (no frequencies). What would settle it: measure the dielectric response of isolated chromatin and cytoplasm from GHz to THz (a measured Q above 1 would revive it; the calculation predicts none).

Claim 2. Disease and cancer are oscillatory disequilibrium: a "war of radiations" (microbe, irradiated cholesterol, altered cell constants)

Reviewers: all agree the general form is Not testable as stated. Differ on the Roffo prop: physicist "Never properly tested (irradiated form); radioactivity part refuted"; contrarian "Tested by others, not supported"; mathematician "Never properly tested here". Weighing: the physicist is more exact. Roffo's irradiated-cholesterol feeding result (more than 55 percent of "thousands" of mice, or rats, none in controls, as relayed by Lakhovsky) is single-source in the project record; I found no direct replication or direct failure. What others failed to confirm (Kirby, Beck and Peacock, Morris et al.) is Roffo's heated-fat and heated-cholesterol gastric tumours, a different experiment, though from the same laboratory and line of work. That is evidence against Roffo's wider programme, not a test of the irradiated form. A 1939 Nature note, as relayed, judged no reaction product isolated [web search snippet]. The printed formula C21H36O8 for irradiated cholesterol (C27H46O) is unbalanced as a simple photoproduct and has no cited source [read]; it could be a mistranslation. UV does not make material radioactive [memory, high]. UV carcinogenesis itself (which Roffo reported in animals) is Established, through DNA damage. The typhoid/coli opposite-pole experiment is ordinary electrophoresis; the step to "frequency" is his inference. Final verdict: general oscillation/"different frequency" claim: Not testable as stated. Irradiated cholesterol as carcinogen: Never properly tested (single reported result; related heated-fat claims Tested, not supported). "Radio-active" carcinogen: Refuted by calculation (accepted nuclear physics; and not what Roffo wrote). UV skin carcinogenesis (background): Established. What would settle it: a repeat of the irradiated-cholesterol feeding experiment with coded animals, plus the original Roffo protocols from the Buenos Aires Institute; any assigned frequency to a cancer line, measured.

Claim 3. Restoring oscillation cures: the 2 m (150 MHz) Radio-cellulo-oscillator necrosed crown-gall tumours on geraniums; an open copper spiral, collars and belts relieved illness

Reviewers: all give Never properly tested for the experiment and for clinical relief, and Refuted by calculation for the energy supply of a passive loop. They differ on how seriously to take heat. Weighing: the geranium note is a real, four-author communication to a learned society (Parisian surgeon Gosset, Gutmann, Lakhovsky and J. Magrou of the Pasteur Institute, who made the tumours) with a stated control group; an independent rival, Kotzareff, called it "the only experimental study of its kind". Against that: n = 3 treated; allocation and blinding not stated; exposure 6, 33 and 9 h all gave onset at about 16 days (a dose-independent lag; consistent with a threshold, a natural time course, or something unrelated to dose; three plants cannot decide); no field, distance or temperature recorded. The only power figure is 40 W absorbed (input) power, reported by Kotzareff from a conversation, together with Kotzareff's statement that useful power must exceed 200-300 W for a sure effect on human cancer [read: Investigator and physicist]. My recompute (table row 6) shows a far-field heating of a thin stem of at most a few kelvin at 0.1 m and negligible beyond 0.3 m, so the physicist's heating upper bound is generous; but near-field coupling to a resonant coil and the unknown geometry mean heating is neither shown nor excluded. Crown-gall tumour formation fails above about 29-32 C and viticulture uses hot-water treatment [web search summary, medium; contrarian], so even a modest local warming is a live hypothesis; so is a non-thermal effect (Roussakow, a position). The nearest mainstream parallel is Schereschewsky (8-135 MHz, mouse tumours, "without substantial heating", then the thermal reading in 1933 after criticism) [read: Investigator; web Roussakow 2013 https://onlinelibrary.wiley.com/doi/10.1155/2013/428027]. That the field then turned thermal after 1933 is a superseding interpretation, not a decisive experiment on Lakhovsky's plants. Lakhovsky says he gave up the continuous 2 m device in 1929 for damped pulses to avoid prolonged heating [read: Waves That Heal p.30, per contrarian], which shows the thermal question was in his mind. Collars and belts: the patent (CH 135072 / FR 657295, 26 Oct 1927) itself says there is no energy source of its own. A resonant antenna of effective area about 0.3-0.5 m2 in even a strong 1 V/m ambient field captures well under a milliwatt; typical fields give nanowatts to microwatts, against 100 W metabolism [calc, physicist and mathematician agree]. The testimonials (1928-29) are from users of a product sold by Lakhovsky's company, uncontrolled, subjective endpoints; a patient's ulcers recurring when a collar broke and clearing again with a new one (Perineau, 6 Dec 1928) is a documented on-off-on observation in one person, single-source, and testable with a sham collar. It was not tested. Final verdict: 1924 geranium experiment: Never properly tested (one small experiment, unreplicated, not controlled for heat; not dismissed after a failed repeat). 1925 passive copper-spiral geranium (n = 1 treated): Never properly tested [QE revision]. Passive collar/belt as an energy supply: Refuted by calculation. Clinical relief from collars (including Attilj's 1928 tumour regressions): Never properly tested [QE revision]. Silver-spiral sterilisation as proof of oscillation: see QE revision. What would settle it: repeat on Pelargonium crown gall, 150 MHz, about 40 W, at least 10 plants per arm, randomised and blind scoring, fibre-optic temperature in the gall, field mapping, sham and heated-only controls; a sham-collar crossover for the clinical claim.

Claim 4. The Multiple Wave Oscillator (US 1,962,565, filed 1931, granted 1934; US 2,351,055, tube, filed 1941, granted 1944 after his death): spark-excited concentric split rings emit many wavelengths (10 cm to 400 m, harmonics to the infrared); since 1931 it has cured cancer and other conditions; no contra-indication ever reported

Reviewers: all agree on a split verdict; no disagreement of substance. Weighing: - Apparatus: the patents are coherent elementary radio (spark gap, coupled open rings with different natural frequencies), give no number and, in the US text, no medical word. A damped spark is broadband; "many wavelengths at once" is credible as a statement of content [memory, high]. - Spectrum: internally inconsistent in the books ("3 m to infrared" vs "10 cm to 400 m" in one paragraph; Rigaux's "10 to 400 m" with the 750,000 to 3 billion Hz figures is a unit slip; "electrostatic" vs "electro-magnetic" with a visible brush discharge). Ring geometry gives fundamentals from decimetres to a few metres (lambda about 2 pi D: D = 0.1 m, 0.63 m; D = 1 m, 6.3 m). A 400 m end needs tens of nF across a 1 m ring [calc, rows 5 and mathematician row 12], not available; it could come from the driver circuit, which the books do not say. "Harmonics to the infrared" would be the 1e5-th harmonic of the top fundamental; light seen is the discharge itself [mathematician, physicist]. - Dose: a 1 W radiated source gives about 0.08 W/m2 at 1 m, SAR about 1e-4 W/kg [calc, physicist]: the far field does not heat; the patient sits in a strong non-uniform near field with corona (ozone, nitrogen oxides, ions); those are non-"frequency" agents. - Clinical record: Mme C. (68, biopsy-confirmed rodent ulcer, 8 Sept to 19 Nov 1931), M.M. (80, naevo-carcinoma, follow-up to 90 in 1941), Mme S. (82, epithelioma after operation and radium, 26 Apr 1932, healed in 16-35 days depending on the account), J.S. (61, basal-cell carcinoma, 13 Oct to 29 Dec 1931), goitre (7 weeks, 1938), three prostate cases (de Cigna, Rigaux), Dr. H.'s five prostatitis cases, and de Cigna's report to the Genoa Academy and the Vienna short-wave congress (July 1937). [QE revision: the "several hundred" are all of de Cigna's patients since 1934, mostly nervous illness he chose not to report; his ulcer series is eight radiographed cases, later "18"; see the QE revision section.] Strengths: named hospitals and clinicians, dates, photographs, some biopsy confirmation, years of follow-up, a 1933 Rigaux admission of "a few negative results", the 1941 admission of failures, and a contemporary Nature reviewer who accepted that the apparatus "apparently has worked marvellous cures". Weaknesses: no denominator; the maker's best cases; slow-growing, sometimes regressing skin cancers; prior X-ray and radium; silver nitrate applied under the apparatus (Rigaux); fluctuating conditions (Graves disease, prostatism); retellings that differ in counts and dates; no sceptical clinician examined them. Retained fully. - Safety: "no contra-indication ever reported" is absence of reports from the maker's users, not a safety study. Final verdict: apparatus emits several simultaneous damped frequencies: Plausible, untested (no spectrum measurement on record). Stated spectrum "10 cm to 400 m plus harmonics to infrared": Internally inconsistent. Cures of cancer, goitre, prostate, ulcers: Never properly tested (case series by interested and by independent clinicians; no controlled or sham-controlled trial; not refused a trial so far as the record shows). "No contra-indication" / "not the slightest accident": [QE revision] Internally inconsistent as printed (the same book records five haemorrhage deaths after treatment-induced necrosis and lists acute phlebitis as a contra-indication); whether the device caused any death is Not testable from this record. See the QE revision section. What would settle it: (1) spectrum analyser and near-field probe measurements (0.1 MHz to 3 GHz) on an MWO built to the patents, with SAR; (2) a blinded or sham-controlled trial on biopsy-confirmed basal-cell carcinoma, or mouse xenografts, with histology before and after, and corona by-products measured (ozone, NOx) to separate them from RF; (3) retrieve de Cigna's Genoa Academy and Vienna 1937 reports and the Saint-Louis case files with biopsy reports.

Claim 5. Radioactivity and carcinogenic radiation are driven by cosmic "universion" waves; a 25 m lead sphere would remove radium's radioactivity; young skin screens such radiation until 40-45 and the oscillator restores it; cosmic radiation also sets species, heredity and the geography of cancer (soil and water electrical constants)

Reviewers: agree on a split verdict. Disagreement on labels: radium/lead (physicist "Refuted by calculation and by experiment"; contrarian "Tested, not supported"; mathematician "Refuted by calculation"); skin screen (physicist "Tested, not supported"; mathematician and contrarian "Internally inconsistent"). Weighing: - Radium and lead. The energy-budget refutation (radium emits 2.6e9 times the power cosmic rays deliver on 1 cm2 [calc, mathematician]) only refutes cosmic radiation as the energy source; Lakhovsky's wording is that of a trigger ("removed from the action of cosmic waves"). A trigger is testable: cosmic flux underground is suppressed by about 1e6, and Gran Sasso measurements of 226Ra and 40K found no time dependence in the decay constant at 3.5 and 3.4 parts per million (1 sigma) [web: https://arxiv.org/abs/1802.09373, read as abstract/summary]. I note that paper tests time modulation underground, not a surface-underground difference directly; the direct comparison rests on the agreement of Ra-226 half-life values from many laboratories at different depths and on the Curie and Kamerlingh Onnes 1913 low-temperature test (unopened; search summary), medium confidence, [memory] for the half-life agreement. Alpha decay is quantum tunnelling (Gamow, Condon and Gurney, 1928) [memory, high]. The passage itself is internally inconsistent (diameter vs thickness; lead as shield against incoming rays and against the hydrocarbon's own emission; limit 2 m, 7 m, 11 m across the books). The contrarian's label fits the experimental side better; the calculation label fits only the "energy supply" reading. I give both parts. - Body heat from cosmic waves: Refuted by calculation (table row 7), all reviewers agree. - Skin screen to 45: the three ages (40, 45, 50) are inconsistent, the screen is invoked for external and internal sources, and cell spacing is irrelevant at the wavelengths involved. No controlled test of the claim exists, so "Tested, not supported" overstates; mainstream explanation of the age curve (multistage accumulation of DNA damage, UV absorbed in the outer skin at all ages) does not need it. Evidence offered for the oscillator "rejuvenating" skin is physicians seeing younger texture, without histology. - Soil and cancer: the Contributo (1927) maps 1921-22 Paris cancer deaths per 1000 residents against geology: sand low (0.3 to 1.1), plastic clay and gypsum marl high (1.41 to 2.08). My recompute of the mathematician's exact test gives 3.3e-4 for the 6 vs 8 quartiers he quotes [calc]. This is striking but post hoc and ecologically confounded: not age-standardised, deaths counted where they occur (the high list includes Saint-Vincent-de-Paul, Hopital Saint-Louis and Pere-Lachaise; the highest value in the table, the Salpetriere at 2.80, is not discussed in the text read), industrial riverside districts, social class. Radon is a measured geological factor in lung cancer [memory, high]. Lakhovsky's electrical-constant mechanism is untested. The skin-depth arithmetic of The Secret of Life p.53 checks out (16,000 m, 18.7 kHz: 1.8 m in seawater, 82 m in dry soil) [calc, mathematician] but links nothing to cancer. Final verdict: lead sphere / radium decay driven by cosmic waves: Tested, not supported (indirectly, by decay-constant constancy; the energy-source reading is Refuted by calculation); the thought experiment as printed is also Internally inconsistent. Cosmic radiation as body heat: Refuted by calculation. Skin screen to 40-45: Internally inconsistent (and unsupported by the known mechanism of UV carcinogenesis). Cancer geography via soil conductivity: Never properly tested. What would settle it: compare a Ra-226 source's activity directly at the surface and in a deep underground laboratory with the same spectrometer; age-standardised 1925 Paris yearbook mortality for all 80 quartiers, hospitals and cemetery districts excluded, geology pre-specified, and radon added; current registry data against measured soil conductivity and radon.

Claim 6. The method works on heredity and the next generation: treating prospective mothers gives healthier children; heredity is fixed by the wavelength of the nuclear circuits; municipal pre-marital treatment

Reviewers: agree (superseded theory; offspring benefit never tested). Rests on the 1941 book and The Secret of Life alone (source gap above). The count of 48 chromosomes was the accepted figure in 1941 (46 established 1956) [memory, high]; that is a period fact, not an error of his. Final verdict: heredity by nuclear wavelength: Historically superseded but instructive (and Not testable as stated, no wavelength or tolerance given). Benefit to children of treated mothers: Never properly tested (three anecdotes; a healthy outcome has a high base rate). What would settle it: nothing needs a new experiment for the mechanism (DNA settled it); for the offspring claim, retrieve the case records, and a registry-based comparison would be needed.

Claim 7. Later citation: Lakhovsky "treated and cured a number of cancer patients until his death in 1942" and is a precursor of Rife and Priore; MWO supplies the missing frequencies of a "polychromatic" healthy spectrum (Winey, US 2007/0258329)

Reviewers: agree. Interested party: the applicant has a commercial and legal stake (golf-shaft tuning); the Lakhovsky paragraph is background with no experiment [read: deep/piezo]. Its citation "Lakhovsky, 1939" cannot document treatment to 1942. "Rife and Priore" are cited to Bearden 1988, itself a secondary source. The same classification applies to the Hatonn layer in The Secret of Life ("proven beyond any shadow of doubt"; the editor sells related products) [read: deep reads]; it is not Lakhovsky and no Crile publication is cited. Final verdict: Not testable as stated (the polychromatic framework); the "cured until 1942" sentence is an applicant's characterisation that overstates its own citation.

Absence-of-evidence analysis (where the record goes dark)

Where it goes dark: the American end, 1941 to 1945 and after. Documented: Lakhovsky left France in 1941 and died on 31 Aug 1942 at the Adelphi Hospital, Brooklyn, three days after being hit by a limousine; a "George Lakhovsky Institute of Multiple Waves" was founded in New York in January 1945 (ncip.by, a state patent office's unsigned biography; one source, no second found) [read: Investigator, research/2026-10-08-ncip-by.md]. US 2,351,055 was granted 13 June 1944 posthumously, with the assignment to Henry S. Blum of Chicago; the administratrix is Anne-Marie Louise Lakhovsky [read: deep/2351055]. [Corrected by QE revision] The 1941 book prints a 19-patient New York hospital table with a per-patient result and seven case histories, and an 8-case Brooklyn urologist table; what is missing is the hospital's own report, any follow-up and any control. Google showed Lumen legal-removal notices on three Lakhovsky queries; content unknown; I draw no inference. No FDA or Post Office file, no trial report in indexed literature, and no clinical follow-up from the 1930s series (de Cigna, Rigaux, Araujo, Reilly, "thousands of cases") has been found. The geranium result was likewise never repeated with temperature recording after 1933.

Competing explanations: 1. Death, war and displacement removed the sponsor (inventor aged 72, emigration 1941, occupied France, death in 1942). For: dates are documented. Against: it does not explain the silence of Genoa, Brussels, Rome, Montevideo and Buenos Aires clinicians. Decider: their own publications (Genoa Academy of Medicine; Vienna 1937 congress proceedings; Florence 1928 radiology congress). 2. No controlled results existed to publish (case-series only; the maker's cases). For: nothing controlled is on record anywhere; Nature 1940 called for none and found none. Against: absence of a record is also what this explanation predicts and cannot distinguish it. Decider: the clinicians' original files and the 1941 New York hospital data. 3. Mid-century medicine moved to radiotherapy and chemotherapy and ignored low-power devices; short-wave oncology went thermal (hyperthermia) after Schereschewsky's 1933 reinterpretation. For: Farrell 1982 files Lakhovsky inside hyperthermia history; the 1950 Le Haut-Parleur list of about 30 later workers shows a field that continued as a different programme. Against: that was not a refutation of his plants. Decider: reading Farrell and the Public Health Reports series. 4. Regulatory or professional pressure on device-based cancer claims. For: documented in the US for other devices [memory, medium; not shown for Lakhovsky]. Against: nothing found for him. Decider: FDA/Post Office/AMA Bureau of Investigation files (the AMA archives), NARA. 5. Commercial continuation under another name (COLYSA at 26 rue des Marronniers, Paris, 1929; Editions S.A.C.L. 1939; a present-day Colysa SA in Val-de-Ruz, Switzerland with no linking entry; present-day sellers). For: devices are still sold under the Lakhovsky name. Against: no succession document found. Decider: Paris Registre du commerce and the Neuchatel cantonal extract. 6. Suppression. For: nothing positive found beyond the unexplained Lumen notices. Against: no record of blocked funding or publication was found; the 1924 note appeared in a society journal, and Nature reviewed the book. Silence is not contradiction, and the record of dissent is thin by nature for an inventor outside institutions; I treat this as not shown and not excluded. Records that would decide: NYC death certificate and 1942 obituaries (New York Times, Brooklyn Daily Eagle), New York State incorporation of the 1945 institute, Henry S. Blum and Chicago corporation records, French RCS for COLYSA, Gallica/BHL copies of C. R. Soc. Biol. 91 and C. R. Acad. Sci. 188 (p.657-659, d'Arsonval), the de Cigna and Attilj reports, and any US Post Office or FDA file.

What would settle each disagreement (single deciding step)

Open questions

1. What did the original 1924 note actually record (temperatures, plant placement, how the 3 treated plants were chosen, how often untreated galls necrose)? The journal text was not obtained.
2. How many patients were treated with the MWO in the 1930s and how many failed? [QE revision: de Cigna's reported series is 8 radiographed ulcer cases (later 18) plus 5 skin cases and a dozen others; his "hundreds" were mostly nervous cases he did not report. Lakhovsky gives "three failures" in "innumerable treatments" and no denominator.]
3. Why does a 6 h, 9 h and 33 h exposure all give necrosis onset near 16 days?
4. Is the MWO's effect, if any, thermal, corona-chemical, non-thermal RF, or placebo? Nothing in the record separates them.
5. What happened to the US venture after 1942 (Institute of Multiple Waves, Blum), and is there a succession to the present-day sellers?
6. What exactly did Roffo publish about irradiated cholesterol (the formula, the species, the counts)? Is C21H36O8 a mistranslation?
7. Did Lakhovsky ever reply to the Nature 1940 identification of the Corynactis figure as a cnidocyst thread?
8. Soft form of the idea: are there frequency-specific, non-thermal field effects on dividing cells beyond TTFields? Mainstream research, not Lakhovsky's, now decides it.

Web checks done (7 calls; budget 8)

• Gran Sasso Ra-226/K-40 time-dependence, cosmic background suppressed: https://arxiv.org/abs/1802.09373 (opened, abstract level). Retag: [web]. The 1311.7043 and 0910.4338 citations of the physicist and the Curie/Onnes 1913 result were not opened and remain search-summary level.
• Kirby 1948 identity and the "not confirmed" statement: https://pmc.ncbi.nlm.nih.gov/articles/PMC2007542/ (identity opened; the quoted statement from a search excerpt, medium); physicist's PMC2008681 corrected (Hieger and Orr 1954).
• Nature 144:246 (Roffo, 1939): login redirect, not read; remains a search snippet.
• EF-14 TTFields figures: secondary coverage and a company press release only (stake).
• Regener's Lake Constance depth (231 m of water): arXiv 1411.6217 via a search summary; the muon attribution remained [memory, medium].
• Gurwitsch mitogenetic rays: reported in the 1920s, replication attempts overwhelmingly negative in the late 1930s, revived in the 1960s and the root of the biophoton literature; Volodyaev and Beloussov, Frontiers in Physiology 2015 (https://www.frontiersin.org/journals/physiology/articles/10.3389/fphys.2015.00241/pdf), via a search summary. Retag [web]; it is background and no verdict depends on it.
• Still [memory]: thorotrast as an alpha-emitting carcinogen (high), plant tissue conductivity at 150 MHz (low; the heat bound uses 0.5 S/m and 0.3 S/m in different reviews, results scale linearly), the silkworm egg mass (low), fulguration history (medium). No verdict depends on them alone.

Summary table of final verdicts

QE revision (2026-10-08)

Inputs: qe.md (verdict MINOR); the new deep reads appended to deep/ (Radiations and Waves pp.1-13, 14-29 and 30-48; The Secret of Life pp.5-16 and 35-37, 57-59; Waves That Heal pp.9-24); one web check (Engels 2014). Tags as above. The central verdicts (cell circuit refuted by calculation; geranium and MWO cures never properly tested; radium/lead tested, not supported) stand. What changes: one new claim (8), a rewording of the "no contra-indication" verdict, a better account of the clinical record, and several earlier statements of mine that were wrong.

Corrections to earlier documents (this section's own)

1. arbiter.md, open question 2 and Claim 4: "de Cigna's several hundred ulcer cases with radiological control" was wrong. De Cigna writes that his cases "reach into the hundreds", of which he reports only those with "objective clinical observations", and says he is "purposely avoiding" the nervous cases. The ulcer series is "eight cases, all checked and verified by radiographs", "six presented to the Royal Academy", later "18 cases" [read: deep/Radiations and Waves pp.30, 33, 35]. The same applies to the contrarian's and physicist's wording (the physicist's "no individual case described" holds for Waves That Heal, not for the 1941 book).
2. arbiter.md, absence-of-evidence section: "a 1941 New York hospital trial table ... with no outcome record" was wrong; see the Claim 4 item below.
3. qe.md item 3: the throat patients had four sessions and the bleed was found when Lakhovsky "arrived for the fifth", so "after the fourth treatment" (not "at the fifth") [read: pp.41-42].
4. Lakhovsky's "three failures" and the "Mrs. S." operation date "1939" (impossible for April 1932), the 26 July vs 26 August 1924 date of the Gosset communication, and the 11/12 July 1937 Vienna dates are printing or translation inconsistencies in the 1941 book; the July 26 date of my correction 3 stands as the note's date (C. R. Soc. Biol. 91:626 is 26 July 1924 [web, earlier check]).
5. Waves That Heal: Attilj's four-patient list there and the six-patient list in The Secret of Life differ (hand sarcoma, breast sarcoma, post-mastectomy scar, obesity in both; the Secret of Life adds an oral epithelioma and a genital epithelioma). Technique and contraindications often attributed to Araujo belong to Dr. Nicholas Gentile (Medicina Nuova 1935) [read: Waves That Heal p.19-22].

Claim 3 (additions): silver spirals, the 1925 copper spiral, Attilj

Silver-spiral sterilisation (Secret of Life p.35-36; presented by d'Arsonval 15 April 1929, with M. Sesari of the Pasteur Institute, abridged in the book). Measured, as reported: B. coli 1,128,000 per cm3 fell to 73,600 (119 cm2 of silver) and 171,200 (72 cm2) at 18 h and to 0 at 25 h, while the control grew to 43,680,000 (38.7-fold); the 18 h control is blank, one vessel per condition, no replicate [read; calc]. Experiment III: water exposed to the silver, re-seeded after the metal was removed, gave 12 colonies (untreated) and 13 (heated to 101-115 C) against 946 in the control, but 1,474 after Chamberland filtration [read]. Typhoid "slightly more prolonged"; "platonix" repeat, no data. - What is established: silver kills bacteria; the oligodynamic (metal-ion) action has been known since the 1890s [memory, high], and Lakhovsky himself says it "has been known for some considerable time". The measurement is retained as a real, controlled-at-one-time-point observation by a Pasteur Institute collaborator. - Whether it supports "oscillation": no. His premise was that the effect is "purely physical", the metal altering the nucleus's oscillation. The data cannot separate that from released silver ions, because there is no inert-metal or non-silver spiral, no silver assay, and Experiment III shows only that the agent survives heating and is lost on porcelain filtration. QE read this as favouring a dissolved agent; the deep reader as favouring a particle-bound one (dead cell bodies, Lakhovsky's "immunizing effect"), noting that a porcelain candle can also adsorb silver ions [memory, low]. I leave that open: the result fits both, and neither is oscillation. Fouling of the silver by lime and organic films stopping the effect (his p.36 remark) is what contact chemistry predicts as well. - Verdict: killing of B. coli by silver: Established (as chemistry; the reported numbers are plausible and the source is single). Killing as proof of nuclear oscillation: Never properly tested (the discriminating controls were not run; not refuted). Dead bacteria as "immunizing" vaccine source: Never properly tested. - What would settle it: repeat with silver, an inert metal and a silver-free spiral of the same area, with ICP assay of dissolved silver and replicate plates.

1925 copper spiral (Secret of Life p.37, 57-58). A single treated plant "chosen at random", a 30 cm open copper ring on an ebonite support, inoculated 4 Dec 1924, spiral early Jan 1925; controls dead by 30 Jan 1925, tumour shed by end of February, plant 1.40 m on 23 March 1928 [read]. The ring circumference 0.94 m gives a half-wave resonance near 1.9 m, which is where his "about 2 m" comes from: arithmetic, not a field measurement [calc]. This matters for the heat question in Claim 3: a passive ring cannot deliver tens of watts (collar bound, above), so if the result were real the heat explanation is not available; with n = 1 treated and an undefined control group it cannot be told from natural variation in crown-gall course. Verdict: Never properly tested.

Attilj (Florence, May 1928; Waves That Heal p.21-22; Secret of Life p.58). Hospital radiologist, not the seller, though relayed by the seller's publications. Passive bracelet or waist circuit; relapsing hand sarcoma "almost completely disappeared" after two months; relapsing breast sarcoma "no longer palpable"; scar recurrence and arm oedema relieved; one obese patient's lumbar pain gone in three days. Both sarcoma patients had earlier surgery (the breast case also X-ray), so timing against those is a confound; no histology; follow-up at most about six months; Attilj himself calls it "merely the beginning" [read]. Single-source evidence stated and kept; it is a regression report from a named clinician with no control. Verdict: Never properly tested (collar relief); it does not change the energy-supply verdict.

Animal reports offered against placebo (Waves That Heal p.9-12, 15). Fournier-Ormonde's seven horses (red cells up, viscosity down, direction only, no values, no controls), the filly Ballerina (owner "had backed her"), a Cannes fox-terrier, a Brisbane dog (January 1940, the vet had been treating it before), a parrot; originals said to be in L'Oscillation cellulaire (1931), not obtained. The booklet argues animals and plants exclude suggestion; the reply is that the reporters (owners, trainers, vets) are the ones who suggest, not the animals, and there are no controls or blinding. Dr D. Marcus's 1928 insomnia crossover (his wife borrowed the collar and slept, he relapsed; he resumed and recovered) is a single unblinded on-off-on observation like Perineau's. The only plant control is the untreated plants in Plate III. Verdict: Never properly tested.

Claim 4 (additions): the clinical record read in full

• Deaths. Five patients with advanced head, neck or thoracic cancer died of haemorrhage: two at Saint-Louis (Dec 1931, four sessions each, found bleeding when Lakhovsky came for the fifth), Captain H. at Val-de-Grace (necrosing 8-10 kg thoracic tumour, "a piece the size of a new born baby's head" detached at dressings, died January 1932), and two throat cases reported by Rigaux, "the hemorrhage and death always followed the fourth treatment" [read: pp.41-43]. Lakhovsky's reading: necrosis freed arteries the tumour had eroded. Two readings fit and the record cannot separate them: treatment-induced necrosis exposing eroded vessels (which is evidence that the device acted on the tumours and an adverse-event signal), or the late course of the disease (fatal erosion haemorrhage occurs in untreated advanced head-and-neck cancer [memory, high]). The p.30 statement that no user "has mentioned an accident, nor have there been any contra-indications" and the book's own "three failures ... in innumerable treatments" (no denominator) sit beside this. Gentile also lists acute phlebitis as a contra-indication and cardiac asthma as negative [read: Waves That Heal p.20-21].
• De Cigna's cases, tabulated. Five skin cases: P.A. (basal-cell epithelioma, 10 sessions, healed at 24 days, unchanged at three years), G.C. (histology at the dermatology clinic; lost to follow-up, "ultimate results ... not on record"), O.G. (80; no histology, "prudence counselled a diagnosis of an ulcerated wart"; healed in 7 days), S.G. (histology by Solimano, basal-cell; 17 sessions; pain gone, lesion "remained fixed", not claimed cured), P.N. (erythematous lupus of 20 years after X-ray, Curie, Finsen, cryotherapy and gold salts; 20 sessions; "all lesions have disappeared"). Ulcers: P.G., R.A., D.F., P.M. and Father N. have dated histories; only P.G. has a post-treatment radiograph quoted (niche not found, "small irregularities" persist, taken two months after the final treatment). Co-treatments in the ulcer series: all other medication stopped in P.G., R.A. and Father N.; de Cigna's own lemon-juice regimen (up to four large glasses a day for 25 days), which he says has "always given me immediate results" for gastric pain; strict diets; nasal reflexotherapy and Bonnier cauterisation for constipation in R.A. and D.F.; for the Carmelite, rest stopped a bleed before the MWO. Pain relief therefore has more than one candidate cause on de Cigna's own account. His caveats, in his words: "six cases are not sufficient, and the time elapsed is too short to speak of a positive cure"; heat and oscillation are both considered, "the oscillatory factor ... certainly has the most significant value, but for the moment, we cannot completely prove it". Date and count anomalies as printed (P.G. schedule; R.A. 20 sessions in 136 days at 10-12 day spacing; "eight" / "six" / "18" ulcer cases). No adverse event in any de Cigna case. Selection: he reports only cases with photographs, radiographs or histology. This is the best-documented clinical series in the record and it has not become more favourable or less favourable on a close read: documented objective endpoints in a few cases, co-treatments in the ulcers, one lost, one not claimed cured.
• Lakhovsky's other cases. Rigaux's Colonel T. (penis cancer): biopsy negative 1928, positive 1931, one inguinal node "hard, sclerosed and cancerous", radium twice, X-ray, insulin for diabetes, silver nitrate 3% applied under the apparatus, an unexamined excised "segment", a gangrenous 2 cm gland; at the talk (about 3.5 months, 20 Jan 1933) pain gone, +5 kg, scrotum normal, and "As for the local condition, I'll let you be the judge"; Lakhovsky saw him a year later. Rigaux also names "the effluvia and sparks" as an acting agent (the corona question). Roversi's 1934 letter (a list of results, one control radiograph, no histology, a request for Lakhovsky's photo and autograph). Dr H. of Brussels (five prostatitis cases, "25" nocturnal micturitions to one or two, unit unstated). Hulin's one-patient dental radiograph series (reossification with six sclerosant injections plus calcium-chloride iontophoresis). The 1935 English lady (leucocytes 20,000 to 5,000 in a week; unnamed surgeon; temperature printed 32.7 C). Mr C. "Courbe" (a cerebello-pontine angle tumour with no histology, 20 radiotherapy sessions without change in 1929, then 14 MWO sessions, patient letter 31 Dec 1933). All are retold by the maker or his collaborators; each is kept.
• New York 1941. The 19-patient table records a result for every patient (all improved to some degree, none "no change"), mostly arthritis and soft-tissue conditions chosen by Lakhovsky because "these are illnesses for which treatments are known"; the hospital is anonymous and withheld its report pending a presentation to the New York Society of Physiotherapists; case histories differ from the table (F.T. 12 vs 11 sessions; M.M. 7 vs 10 and 9; C.P. endocervicitis 6 sessions vs parametritis 12+ with an exam on 7 October, after the book's 21 August close); four of seven histories end "did not return". The Brooklyn urologist (anonymous, had bought a machine, "hundreds of patients", "picked at random") gives 8 cases and a 6-patient prostate group; S.B., bladder carcinoma, is the only case in the 1941 pages with an instrument check (cystoscopy three weeks apart: "reduced in size"). Retained as stated: favourable outcomes reported by an interested party with a device purchase at stake, no control.
• Priority and scale claims. "Venice 1934 and Vienna 1937 gave official recognition to Lakhovsky as the first to invent the short wave oscillator for therapeutic use" and "more than 50,000 of these short wave therapeutic apparatus are in use" are stated by Presner in the preface without proceedings, votes or maker figures; both are about the earlier single-band 2-10 m apparatus, not the MWO. Kotzareff (1931, independent rival) credits Lakhovsky ahead of Schliephake [read: Investigator]. Lakhovsky's own account of Vienna adds that the communications "presented contradictions, the results obtained, positive or negative". Recorded as an interested party's characterisation; the proceedings would settle it.
• Dose, re-done with the one power figure on record. The directions (p.48) give a mains draw of 110 or 220 V at 2 to 4 A (220 to 880 VA), three power steps, transmitter and receiver 1.3 to 1.5 m apart with the patient between, sessions of 5 to 12 minutes, a pause after the fourth treatment; seller Colysa [read]. Radiated fraction unknown. Range [calc]: 1 percent of 440 VA is 4.4 W, which at 0.7 m from a point source is 0.7 W/m2; 10 percent of 880 VA is 88 W, 14 W/m2. A body absorbing over about 0.5 m2 would take up roughly 0.35 to 7 W, or 0.005 to 0.1 W/kg over 70 kg, against basal metabolism near 1.4 W/kg. The whole-body far-field heating conclusion therefore holds across the range; the patient is in the near field of two resonators with corona, where the estimate is not reliable. Rigaux and Lakhovsky both name sparks and effluvia as acting agents. Neither heat, corona chemistry, RF nor placebo is excluded.
• Directions for use: after the fourth treatment pause two to three weeks; the throat deaths came after the fourth. The book does not connect the two; I note the coincidence without reading anything into it.

Revised Claim 4 verdicts: apparatus emits several simultaneous damped frequencies: Plausible, untested. Spectrum as stated: Internally inconsistent. Cures: Never properly tested. "No contra-indication / not the slightest accident": Internally inconsistent as printed; deaths' causation Not testable from this record. Priority and 50,000-units claims: unverified, an interested party's statement. What would settle it (adds to the Claim 4 list): de Cigna's Genoa Academy session (31 May 1935) and Vienna 1937 case files; the Saint-Louis, Val-de-Grace and Rigaux records of the five deaths with autopsy if any; the Venice 1934 and Vienna 1937 proceedings.

Claim 6 (addition, from qe.md)

The 1941 book frames the three children as "a new race of unusual strength and perfection of form", proposes a municipal decree requiring treatment of engaged couples, and claims conscription unfitness would fall from 35-45 percent to 5 percent [read: qe.md, from Radiations and Waves pp.65-67; not re-read by me]. These belong beside the verdict. Verdicts unchanged.

Claim 8 (new). Animals orient by radiation they emit and receive; radio transmitters disorient homing pigeons; the semicircular canals are a radio direction-finder; living things charge electrically in flight

Reviewers: none reviewed it (it entered at QE); this is my first and only weighing, from the deep read of Secret of Life pp.5-16, a [web] check and [memory] where marked. - Pigeons. Reported: Paterna radio station near Valencia, 2 July 1924 (observation "came to my notice", observer unnamed); repeats under Spanish military control (the year 1925 is an editor's insertion) and on 7 November 1926; release one bird every three minutes, pigeon station about 8 km away, transmission more than half an hour, no bird took a direction, a wavelength change did not help, and within minutes of the transmission stopping the birds flew straight home, including those from the first run; Kreuznach, March 1926, with dove-cot and station on opposite sides so the station lay on the homing line, birds lost bearing near the aerial and recovered outside "the intense electromagnetic field" [read: pp.5-6]. Not given: number of birds, power, frequency, how disorientation was scored, any citation for M. J. Casamajor's "detailed report", the name of any experimenter. Single secondary source (Lakhovsky); the French contribution is not described. The on-off contrast is a real design feature if it happened as stated. - Related modern work. Engels et al., Nature 509:353-356 (2014): European robins tested at Oldenburg lost their magnetic-compass orientation in unscreened huts and regained it when shielded from broadband electromagnetic noise; the Oldenburg press office puts the noise at about 2 kHz to 5 MHz, mostly from electronic devices, well under ICNIRP and WHO limits; the authors state earlier low-intensity effects had a poor replication record and that their own was reproduced under blinded conditions [web: https://ideas.repec.org/a/nat/nature/v509y2014i7500d10.1038_nature13290.html ; https://uol.de/en/news/article/electrosmog-disrupts-orientation-in-migratory-birds ; https://www.nationalgeographic.com/science/article/electromagnetic-noise-disrupts-bird-compass]. That is the same class of observation, a different species and a different organ (the magnetic compass, which the authors and others tie to light-dependent radical-pair chemistry in the eye [memory, medium]), and a lower frequency. It does not test Lakhovsky's pigeon reports and says nothing for his mechanism, but it shows that "radio noise disturbs bird orientation" is not an absurd observation. - Mechanism. The canals are known as the rotation sensors of the vestibular system [memory, high]; Lakhovsky's own argument has his canal fluid called "insulating" and then "conducting" within a few lines, and the support offered is that birds without canals lose equilibrium, which is what a rotation sensor predicts. No experiment applying fields to canals is reported. "Radiation emitted by birds and insects, with frequency discrimination" has no measurement. - Moths. The Fabre material is accurately reported; the attractant that Fabre called an emanation is a chemical sex pheromone, characterised for the silkworm moth in 1959 (bombykol) [memory, high]. Lakhovsky's own test (a cotton leaf the female lay on keeps attracting males; dipped in pure alcohol or corrosive sublimate it stops) has no count, trial number, or control with another solvent; alcohol dissolves and removes odorants, and mercuric chloride would also destroy cells, so the test does not separate "living-cell radiation" from "odour" [read; chemistry memory, medium]. That is where it fails, not on whether it was done. - Duck wing. One reading of about 600 V from shaking a duck's wing in front of a "radium electrometer" with the experimenter insulated on two 2 cm ebonite discs, no repeats, humidity or distance given [read: p.10]. Frictional charging by plumage in air is ordinary; 600 V is plausible. The step to 25,000 V for a flying bird and a bird "regulating" altitude by Q = CV is assertion, not data [read: p.11], and a separate use of the 1 V/cm fair-weather gradient is stated roughly (about 100 V/m is the right magnitude [memory, high]). - Retention. The editors' interjections ("Hatonn", "E.Y.") in these pages are not Lakhovsky and are not evidence; kept out. Final verdicts: radio transmission disturbs homing in pigeons, as reported: Never properly tested (reported, never replicated in the record; the underlying reports not obtained; nothing found that dismisses it). Canals as radio direction finders: Not testable as stated as framed, and unsupported by their known function. Orientation by the animal's own emitted radiation: Not testable as stated (no frequency, power or detector). Oak-bombyx cotton-wool test: Never properly tested. Duck-wing charging: Plausible, untested; flight-potential regulation: Not testable as stated. What would settle it: Casamajor's Paterna report and the Spanish pigeon service records (Valencia 1924-26), the German Kreuznach reports (1926), then a blinded release of homing pigeons with a transmitter on and off in the 150 kHz to 10 MHz band, power and field measured.

Claim 5, soil section (addition, orchestrator fix 2026-10-08): the side-predictions

qe.md (item 7 and the Recheck) asked for this. It said "Claim 6 or the soil section"; in this file the soil and cancer geography is Claim 5, so the weighing goes here. The sources are The Secret of Life pp.51-59 (_text/, read for this fix) and the Contributo table the mathematician read (research/files/1927-Contributo-eziologia-cancro_djvu.txt). Correction to qe.md's Recheck: it says these items "are in facts.json and claims.md". facts.json has the full-moon and conducting-soil observation (Secret of Life entry, key_claims), but before this fix neither file had Thiais/Orly or Simeray, and claims.md had none of the four. I have added them to both (see qe.md, "Orchestrator fix").

1. Full-moon stall. Claim [read: Secret of Life p.59]: "a distinguished French professor", unnamed, in "one of the great Paris hospitals", weighed his patients, took blood counts and measured blood pressure once a week. During "about eight days" the improvement "came to a definite standstill in all the patients", and the period "coincided with the full phase of the moon". Lakhovsky explains that the moon changes the field of cosmic waves, so the circuit absorbs less. The prediction: response to the circuits should dip at every full moon. Does the printed data test it? No. No numbers, dates, patient count or second lunar cycle are printed, and the Contributo holds no treatment data. How likely is the coincidence by chance? The synodic month (new moon to new moon) is 29.5 days [memory, high]. Take "full phase" as the 8 days around full moon. Then an 8-day stall that starts on a random day overlaps it whenever it starts within 7 days either side of the window's start: 15 of 29.5 days, a probability of about 0.5 [calc]. With weekly readings, an 8-day stall is one flat reading. One episode therefore cannot separate a lunar effect from chance. Verdict: Never properly tested. It would be settled by a pre-specified record of the same measures over at least six lunar cycles.

2. Conducting soils respond faster. Claim [read: p.59]: in "all the patients wearing oscillating circuits", the effect was "immediate and rapid" for those living on "highly conducting soils, that is to say naturally carcinogenic", such as Grenelle, Javel, Auteuil and Neuilly. It was "much slower" on "insulating soils" such as Dauphine, the Champs-Elysees, Gaillon and Monceau. Hence his "paradoxical conclusion" that "the worst soils ... finally turn out to be the best". The prediction: among circuit wearers, time to response should be shorter in the high-cancer districts. Does the printed data test it? No. The Contributo prints cancer deaths per 1,000 residents per year (1921-22), not response to treatment. What it does show is that the two lists are the two ends of his own mortality table: Grenelle 2.08, Auteuil 1.76, Javel 1.61 and Neuilly 2.25 (the Neuilly figure is from the Seine commune list, Secret of Life p.51), against Gaillon 0.324, Porte-Dauphine 0.53 and Champs-Elysees 0.68. Plaine Monceau is mid-table at about 1.07-1.08; the OCR does not align the two figures to the names with certainty [read: Contributo table, lines ~1403-2016]. No measured soil conductivity for any district appears in the texts read. The "conducting" label therefore comes from rock type and, in effect, from the same mortality ranking. No patient counts or times are given. Verdict: Never properly tested. It would be settled by circuit wearers grouped by district before treatment, with time to a pre-defined response compared between groups.

3. Thiais and Orly, and Simeray's village (water supply). Claim [read: Secret of Life p.55]: Dr. Simeray told the Academie de Medecine (15 March 1927) of a village that was free of cancer "as long as" it drank from its own wells, and had "a series of cases of cancer" after the authorities brought water from outside. Lakhovsky then compared two neighbouring communes on "the same kind of soil", Brie fresh-water limestone: Thiais, 3.36 per 1,000, drinking Seine water drawn at Alfortville, and Orly, 0.36 per 1,000, drinking from its own wells. In his words: "As this case did not appear to be in accord with my theory on the subject, I decided to investigate." The prediction: water from a distant source, with "electrical constants" different from the local soil, raises cancer; local wells protect. Does the printed data test it? Only in part, and not in his favour as a soil test. The Thiais and Orly figures are a tenfold difference on one rock type, which is a counter-example to the soil rule. The water explanation was found afterwards, for this one pair, and no water conductivity was measured. The Simeray village has no name, counts or dates in his account, and the Academie paper was not read. The Thiais and Orly values are not legible in the OCR of the Contributo's Seine map. Its legend puts "Calcaire de St-Ouen et de Brie" in the 1.26-2.00 band ("terrains mauvais"), so Thiais (3.36) lies above its class and Orly (0.36) far below it [read: Contributo, lines ~2595-2737]. Two further points weigh on it: - Small numbers. The texts do not give the communes' populations. As an illustration only: a commune of 2,000 people at 0.36 per 1,000 per year records about 1.4 cancer deaths in two years [calc]. One or two deaths more or fewer would then change the rate several-fold. - Limestone is classed both ways. "Sand, limestone, etc." is "insulating soil" and goes with low cancer [read: pp.51, 53], but Brie limestone is "a fairly good conductor and therefore characteristic of a high cancer density" [read: p.55]. The Contributo legend separates coarse Paris limestone (good band) from Brie and Saint-Ouen limestone (bad band). Since no conductivity is measured, each rock can be put in whichever class matches its cancer figure. That makes the soil rule, as printed, hard to falsify.

The water source as a cancer factor is a reasonable question in mainstream terms, through chemistry: arsenic in drinking water is an established carcinogen [memory, high]. Lakhovsky's electrical-constant mechanism is a separate claim and is untested. Verdict: water-supply effect, Never properly tested (one unnamed village, reported second-hand, and one pair of communes explained after the fact). Its electrical mechanism is Not testable as stated, because no constant is measured or specified.

Bearing on the Claim 5 soil verdict. These internal predictions do not test the soil-cancer correlation. They show how it was used. A counter-example (Orly) was absorbed by an added cause (water), and the soil classes are flexible enough to fit the figures. This weakens the soil claim as evidence; it does not refute it. Verdict unchanged: cancer geography via soil conductivity, Never properly tested. What would settle the side-predictions: the 1925 Paris statistical yearbook and Seine figures for Thiais and Orly with populations; Simeray's paper (Bulletin de l'Academie de Medecine, session of 15 March 1927); the unnamed professor's weekly records, if they survive.

Updated records and open questions

New records: C. R. Acad. Sci. note of 15 April 1929 (Lakhovsky, Sesari, presented by d'Arsonval); Attilj's Florence 1928 report; L'Oscillation cellulaire (1931, animal data); Rigaux, "Cosmic Rays and Oscillating Circuits" (1932); Medicina Nuova XXVI no. 5 (1935, Gentile); proceedings of Venice 1934 and Vienna 1937; Casamajor and Kreuznach pigeon reports. Open questions added: Did an independent group ever repeat the Paterna pigeon on-off result? What did the silver-spiral water contain (dissolved silver) after exposure? Were the five haemorrhage deaths among all advanced cases treated, or all of the advanced cases? (The denominator.)

Web check (1 call; budget used 8 of 8)

Engels et al. 2014: citation and band confirmed through the Oldenburg and Nature index pages (search result summaries, not the paper text), [web] as above. The Gallica check of the 1929 silver note was not done (budget).

Other changes to the tables above

The verdict summary table now carries Claim 3 (silver, 1925 spiral, Attilj), Claim 4 ("no contra-indication") and Claim 8 as above. Nothing else changes; the five-point agreement list, recomputed numbers and absence-of-evidence competing explanations stand, with the New York correction.