Lakhovsky George: Primer (Step 4)
Date: 2026-10-08. Written for the reader profile in prompt.txt. Built from arbiter.md (authoritative verdicts), physicist.md, mathematician.md, contrarian.md, deep/, sites.md and research/. No new verdicts here: every verdict is the arbiter's. Worked example reproduced in example.py.
Tags: [read] project document or Investigator file, [web] with URL, [calc] arithmetic done here or in example.py, [memory] recalled, with confidence.
QE revision (2026-10-08). Updated after qe.md (verdict MINOR) and the arbiter's "QE revision" section, which rests on new deep reads of Radiations and Waves pp.1-48, The Secret of Life pp.5-16, 35-37 and 57-59, and Waves That Heal pp.9-24. Corrections to this primer's earlier text, stated plainly: 1. Section 6 row 4c said de Cigna treated "several hundred ulcer patients with radiological control". That was wrong. De Cigna wrote that his cases "reach into the hundreds", mostly nervous illness he chose not to report; his ulcer series is eight radiographed cases (six presented to the Genoa Academy, later "18") [read: arbiter QE revision, from Radiations and Waves pp.30, 33, 35]. 2. Section 6, last bullet, said the 1941 New York hospital table had "no outcome record". That was wrong. It gives a result for each of 19 patients; what is missing is the hospital's own report, any follow-up and any control [read: Radiations and Waves pp.45-47, via arbiter and qe.md]. 3. Row 4c compressed five deaths by haemorrhage after treatment into "failures admitted in advanced cases". They are now stated (section 6), and the "no contra-indication" verdict has changed with the arbiter's. New in this revision: a claim 8 (radio and animal orientation), the silver-spiral sterilisation experiment, the 1925 copper spiral, Attilj's collar cases, the animal reports, de Cigna's co-treatments and caveats, the MWO's mains power draw, the priority claim, and Lakhovsky's eugenic framing of claim 6.
Orchestrator fix (2026-10-08, after the QE recheck). Added rows 5e-5g (full-moon stall, conducting-soil response, Thiais/Orly and Simeray water supply) and two bullets under "Material kept on the record" (the soil side-predictions; the Hulin, leucocyte and Institute Nos. 1-4 cases). No verdict changes.
Source note (carried forward from the arbiter). All six folder documents have readable text, but none is a primary laboratory or clinical record. The cited primary papers (Comptes Rendus de la Societe de Biologie 91:626, 1924; Roffo; Metalnikoff; Nodon; Rivera; Rigaux; de Cigna; Attilj) were not obtained, so what they found is Lakhovsky's report, except the 1924 note, which is known from a quotation in a 1925 press article on a device seller's site (colysa.ch, a party with a commercial stake) and from citations in Kotzareff 1931 and Roussakow 2013. The Secret of Life copy has no figures and carries editorial layers ("Hatonn", "E.Y.") that are not Lakhovsky. Claim 6 (heredity) rests on the 1941 book and The Secret of Life alone. Where a statement rests on [memory] alone it says so.
1. The idea in one paragraph
Georges Lakhovsky (born 17 Sept 1869 in Ilya, then in the Russian Empire; died in Brooklyn on 31 Aug 1942 [read: Investigator, ncip.by note]), an engineer who worked in Paris, took the radio engineering of the 1920s and applied it to biology. A radio receiver works because a coil and a capacitor form a circuit that "rings" at one frequency and picks that frequency out of the air. Lakhovsky proposed that every living cell contains such circuits: the twisted filaments in the nucleus (what we call chromosomes) would be tiny insulated tubes filled with salty, conducting fluid, ringing at very high (infrared) frequencies and kept going by radiation from space. Health would be a cell ringing at its proper frequency; disease, including cancer, would be a cell pushed off it by microbes or other influences (a "war of radiations"). Cure would then mean restoring the cell's oscillation, not killing the invader. He built devices on that idea: a 2-metre-wave oscillator that, in a 1924 experiment co-signed with a surgeon and a Pasteur Institute plant pathologist, was reported to make tumours on three geranium plants die while sixteen untreated plants kept their tumours (the tumours were crown gall: a plant tumour caused by the soil bacterium Agrobacterium tumefaciens, called Bacterium tumefaciens in Lakhovsky's books, which transfers some of its own DNA into the plant's cells and makes them grow unchecked [memory, high; name read: Secret of Life p.29]); passive copper collars and belts; and the Multiple Wave Oscillator (MWO), a spark-driven set of concentric open rings meant to bathe the body in many frequencies at once so that every cell would find its own. Physicians in Paris, Genoa, Brussels and elsewhere reported cures with it in the 1930s. He also tied radioactivity, ageing skin, the geography of cancer and heredity to the same cosmic waves.
2. What you need to know first
2.1 The LC circuit and the damped oscillator. An inductor (inductance L, in henries, H) stores energy in a magnetic field; a capacitor (capacitance C, in farads, F) stores it in an electric field. Connected together, energy sloshes between them at the natural angular frequency omega0 = 1/sqrt(LC) (radians per second), i.e. frequency f0 = omega0/(2 pi) in hertz (Hz). Real circuits have resistance R (ohms), which turns the energy into heat. The charge q on the capacitor then obeys L q'' + R q' + q/C = 0, the same equation as a mass on a spring with friction. It oscillates only if R is less than the critical value 2 sqrt(L/C); above that it just drains away without a single swing (overdamped). The quality factor Q = omega0 L / R measures how many cycles the circuit rings before its energy falls substantially (roughly Q/pi cycles to lose most of it). A radio tuning circuit has Q of tens to hundreds. Read: J. R. Taylor, Classical Mechanics, ch. 5 (oscillations); E. M. Purcell and D. J. Morin, Electricity and Magnetism (3rd ed.), the chapter on alternating-current circuits.
2.2 Wavelength, frequency and the half-wave dipole. An electromagnetic wave of frequency f has wavelength lambda = c/f, where c = 3.0e8 m/s is the speed of light. So 2 m is 150 MHz (megahertz, 1e6 Hz), 400 m is 0.75 MHz, 10 cm is 3 GHz (gigahertz, 1e9 Hz), and 10 micrometres (um, 1e-6 m) is 3e13 Hz, in the infrared. A straight conducting rod in free space, open at both ends, resonates most strongly when its length is about half a wavelength: this is Hertz's dipole, and it is why Lakhovsky estimated a cell filament's wavelength as twice its length. The rule assumes a good conductor in empty space. Read: D. J. Griffiths, Introduction to Electrodynamics, ch. 9 (electromagnetic waves) and ch. 11 (radiation).
2.3 Conductors, electrolytes and resistance. A wire of length l, cross-sectional area A and conductivity sigma (siemens per metre, S/m) has resistance R = l/(sigma A). Copper has sigma about 6e7 S/m; sea water about 5 S/m; body fluids about 0.5-1.5 S/m [memory, high]. Salt water carries current by ions drifting through water, about ten million times less well than copper. Tissue is a lossy electrolyte: electric fields in it are absorbed and turned into heat. The heating rate per kilogram is called SAR (specific absorption rate, W/kg); the body's resting metabolism is about 1 W/kg (about 100 W for an adult) [memory, high]. Read: Griffiths ch. 7 (Ohm's law, inductance).
2.4 Thermal noise. Anything at absolute temperature T jiggles with energy of order kT per degree of freedom, where k = 1.38e-23 J/K is Boltzmann's constant. At body temperature (310 K) kT = 4.3e-21 J. A signal that delivers much less energy than kT to a structure is lost in the noise unless something amplifies it. A warm body also radiates heat as infrared, peaking near 9 um (Wien's law, lambda_peak = 2.898e-3 m K / T). Read: any intro thermodynamics text (e.g. Schroeder, An Introduction to Thermal Physics), sections on equipartition and blackbody radiation.
2.5 Radioactive decay. An unstable nucleus decays at random with a fixed probability per unit time, the decay constant lambda_d, so a sample shrinks as N(t) = N0 exp(-lambda_d t). Radium-226 decays by emitting an alpha particle; Gamow, and independently Condon and Gurney, explained this in 1928 as quantum tunnelling through the nuclear barrier, a process with no external driver [memory, high]. Read: K. S. Krane, Introductory Nuclear Physics, the chapter on alpha decay.
2.6 Small-sample statistics and controls. With few subjects, the question is: how likely is the observed split by chance if the treatment does nothing? If 3 of 19 plants are chosen at random to be treated and exactly those 3 are the ones whose tumours die, the chance is 1 in C(19,3) = 969 (Fisher's exact test; C(n,k) is "n choose k"). That reasoning is valid only if allocation really was random and the scorer did not know which plants were treated. Read: any introductory statistics text, section on Fisher's exact test and randomised controlled design.
3. The textbook view
3.1 Resonance in a circuit. For a series circuit (symbols as in 2.1):
- natural frequency: f0 = 1 / (2 pi sqrt(LC))
- oscillates only if R < R_crit = 2 sqrt(L/C)
- quality factor: Q = omega0 L / R = (1/R) sqrt(L/C)
- the solutions of L q'' + R q' + q/C = 0 are q ~ exp(s t) with s = -R/(2L) +/- sqrt(R^2/(4L^2) - 1/(LC)). When R > R_crit both values of s are real and negative: no oscillation, just decay.
For a thin straight wire of length l and radius a (a much less than l), standard estimates are L = (mu0 l / 2 pi)(ln(2l/a) - 1) and C = 2 pi eps0 l / ln(l/a), where mu0 = 4 pi e-7 H/m and eps0 = 8.85e-12 F/m are the magnetic and electric constants of free space [memory, high; these are the formulas the physicist and mathematician used]. At the micrometre scale these lumped formulas are only rough (the wire is not small compared with its own wavelength), which is why the half-wave rule and the lumped LC rule disagree by a factor of three; neither matters once damping is included.
3.2 Tissue and fields. Cells are genuinely electrical. A cell membrane carries a resting potential of about -70 mV across about 7 nm, a field of about 1e7 V/m; membranes act as capacitors of about 1 microfarad per square centimetre; nerves and the heart produce real electrical oscillations at hertz to kilohertz [memory, high]. In tissue, an applied radio-frequency field is absorbed mostly as heat. At 150 MHz the field falls to 1/e of its surface value in roughly 5 cm of muscle-like tissue (sigma about 0.7 S/m) [calc, mathematician; conductivity memory, medium]. Controlled heating to 41-43 C (hyperthermia) is an established adjunct to radiotherapy [memory, high]. Whether weak, non-heating fields have specific biological effects is a real research question, contested. The best-documented non-thermal case is Tumor Treating Fields: 200 kHz alternating fields on the scalp, added to chemotherapy in the EF-14 phase III trial on glioblastoma, gave median overall survival 20.9 vs 16.0 months (HR 0.63; HR is the hazard ratio, the death rate at any given moment in the treated group divided by that in the control group, so 0.63 means treated patients were dying at about 63 percent of the control rate) [web: https://www.onclive.com/conference-coverage/aacr-2017/tumor-treating-fields-substantially-improve-survival-in-gbm ; also a Novocure press release, a party with a stake]. The proposed mechanism is disruption of dividing cells by a non-uniform field at the cleavage furrow, not resonance [memory, medium].
3.3 Energy budgets. A passive antenna (no battery) can collect at most the power density S of the ambient wave times its effective area A_e. For a half-wave antenna A_e is about lambda^2/(4 pi) times a small gain factor; for lambda = 2 m that is about 0.3-0.5 m2. Power density from field strength E (V/m) is S = E^2/377 W/m2.
3.4 Cancer and radioactivity. Cancer is understood mainly as accumulated mutations in a cell lineage; incidence rises steeply with age; UV causes skin cancer by damaging DNA in the outer skin at every age [memory, high]. Decay constants of alpha emitters do not depend on cosmic-ray flux; deep underground laboratories, where cosmic flux is cut about a millionfold, measure 226Ra with no time dependence at the parts-per-million level [web: https://arxiv.org/abs/1802.09373, abstract level].
3.5 What Lakhovsky got right on the way. The arbiter records that his arithmetic is usually right: 2 m = 150 MHz, 400 m = 0.75 MHz, 10 cm = 3 GHz, his skin-depth figures for long waves in sea water and soil (16,000 m wave: about 1.8 m in sea water, about 82 m in dry soil [calc, mathematician]), and the 1924 exposure durations. The problems are in the model he applied the arithmetic to.
4. Where the inventor departs
The real physics Lakhovsky starts from is sound: tuned circuits do resonate; a tuned receiver does pick a very weak signal out of a crowded band (his "Australia to Europe" argument, Secret of Life pp.62-63 [read: mathematician]); open conductors do resonate at about twice their length; cells are electrical; a damped spark does emit many frequencies at once. He also proposed, in the 1920s, that cells interact with fields at specific frequencies, a question later science took up in other forms: biophotons (the ultra-weak emission of light by living cells, a few to some hundreds of photons per second per square centimetre, generally attributed to oxidative chemistry; the line runs back to Gurwitsch's 1920s "mitogenetic rays", whose replications were mostly negative in the late 1930s [web: Volodyaev and Beloussov 2015, via arbiter; emission level memory, medium]); TTFields (section 3.2); and Fröhlich's coherence proposal (in 1968 the physicist Herbert Fröhlich suggested that metabolic energy pumped into vibrating molecules could gather into one coherent mode at about 1e11-1e12 Hz, a hypothesis still contested [memory, medium]). The departures are these:
| Topic | Textbook | Lakhovsky | Where it breaks |
|---|---|---|---|
| What rings | A low-loss conductor (metal) with L and C | Nuclear filaments "composed of organic materials of mineral conductors, covered by a tubular membrane of insulating material consisting of cholesterol, plastin and other dielectric substances" [read: Secret of Life p.23]; "filled with a conducting fluid containing all the mineral salts found in sea water" [read: Waves That Heal p.13-14] | Salt water is about 1e7 times worse a conductor than copper. Q comes out near 1e-6: no ringing at all (section 5) |
| How to find the frequency | Solve for L, C and R, including damping | "I calculated approximately the probable self-inductance ... Taking certain average values, I found a radiation localized in the infra-red region"; or wavelength = 2 x filament length, "as the electric dipoles of Hertz"; "these methods are not accurate" [read: Secret of Life p.23] | No dimensions, L, C or R printed; damping never computed |
| Weak signals | A tuned receiver needs Q much greater than 1 and a signal above thermal noise kT | Infinitesimal energies suffice if tuned | Q is below 1 and the predicted band (2-20 um) is exactly where the body's own heat radiation peaks (9.3 um at 310 K) [calc, mathematician] |
| Power supply | Body heat comes from metabolism, about 100 W | Cell oscillation "maintained by cosmic radiation"; body heat is Joule heating of cell circuits by cosmic, atmospheric and telluric waves [read: Radiations and Waves pp.49-50] | Cosmic muons deliver about 2e-7 W to a body, about 5e8 times too little [calc, arbiter row 7] |
| Driving the cell | A drive works best at the resonant frequency | 2 m (150 MHz) and 10 cm to 400 m waves reach the cell; "harmonics" reach the infrared | Cell frequency (about 1e13 Hz) is 1e5 times above the highest MWO fundamental; nothing shows the coupling |
| Radioactivity | Spontaneous tunnelling, fixed decay constant | Radium's activity is driven by cosmic "universion"; a 25 m lead sphere would stop it [read: Radiations and Waves p.59] | Decay constant unchanged under a millionfold cut in cosmic flux underground |
| Heredity | Carried by DNA (1944, 1953) | Fixed by "a wavelength determined by the diameter of the circuits" [read: Secret of Life p.62] | Superseded; no wavelength or tolerance was ever given |
The common thread: each borrowed radio principle is right for metal wires in air and is applied without checking the materials, the losses, the noise and the energy source of a wet cell at 310 K.
5. Worked example: can a salt-water chromosome ring?
This is the one place where the strong form of Lakhovsky's model can be checked with numbers using only his own description of the materials. Lakhovsky printed no filament size, so we must choose one. We take a chromosome-scale filament, length l = 10 um and radius a = 0.1 um (the same assumed inputs the physicist, mathematician and arbiter used); step 8 shows the conclusion does not depend on the choice. Reproduced in example.py (run: .venv/bin/python "review/Lakhovsky George/example.py"; it ran cleanly).
Step 1. Lakhovsky's own estimate (half-wave rule). lambda = 2l. For l = 1 um, lambda = 2 um, f = c/lambda = 3.0e8 / 2e-6 = 1.5e14 Hz. For l = 10 um, lambda = 20 um, f = 1.5e13 Hz. Both are infrared, matching his statement "a radiation localized in the infra-red region". So far he is consistent.
Step 2. Inductance and capacitance. L = (mu0 l / 2 pi)(ln(2l/a) - 1) = (4 pi e-7 x 1e-5 / 2 pi)(ln 200 - 1) = 2e-12 x 4.30 = 8.6e-12 H. C = 2 pi eps0 l / ln(l/a) = 2 pi x 8.85e-12 x 1e-5 / ln 100 = 5.56e-16 / 4.61 = 1.2e-16 F.
Step 3. Natural frequency. omega0 = 1/sqrt(LC) = 1/sqrt(8.6e-12 x 1.2e-16) = 3.1e13 rad/s, so f0 = 4.9e12 Hz, wavelength 61 um, period T = 2.0e-13 s. The reactance omega0 L = 3.1e13 x 8.6e-12 = 267 ohm. (The lumped estimate gives 61 um against the half-wave rule's 20 um: a factor of three, expected at this scale. Both say "far infrared to infrared".)
Step 4. Resistance of the salt-water core. Cross-section A = pi a^2 = 3.14e-14 m2. With sea water, sigma = 5 S/m: R = l/(sigma A) = 1e-5 / (5 x 3.14e-14) = 6.4e7 ohm. With a more cell-like sigma = 1 S/m: R = 3.2e8 ohm.
Step 5. Does it oscillate? Critical resistance: R_crit = 2 sqrt(L/C) = 2 sqrt(8.6e-12 / 1.2e-16) = 534 ohm. R / R_crit = 6.4e7 / 534 = 1.2e5 (sea water) to 6.0e5 (1 S/m). Q = omega0 L / R = 267 / 6.4e7 = 4.2e-6 (sea water) to 8.4e-7 (1 S/m). The fast decay time 2L/R = 2.7e-19 s is about a million times shorter than one period (2e-13 s). The circuit is overdamped by five to six orders of magnitude: it cannot complete a single swing. What remains is a slow resistive-capacitive relaxation with time constant RC = 7.7e-9 s (8 nanoseconds), a non-resonant response with no preferred frequency.
Step 6. The same filament with a copper core. R = 1e-5 / (5.96e7 x 3.14e-14) = 5.3 ohm, Q = 267/5.3 = 50. It would ring like a radio circuit. (At 5 THz the skin effect and other losses would lower this, but not by seven orders of magnitude.) This is the fair way to see Lakhovsky's reasoning: his picture is right for a metal wire, and the radio analogy silently assumes metal.
Step 7. What conductivity would rescue it? Setting Q = 1: sigma = l / (omega0 L A) = 1.2e6 S/m, about 200,000 times sea water and one-fiftieth of copper. No aqueous electrolyte comes close. The contrarian's caveat (dc conductivity is only a first approximation at terahertz frequencies) is fair; the arbiter notes it would take a 1e5-fold error to rescue the circuit.
Step 8. Robustness. The script sweeps lengths 1-30 um and radii 0.05-0.5 um at sigma = 5 S/m: Q stays between 5e-7 and 1.3e-4. Surrounding the filament with water (relative permittivity about 80) raises C, lowers f0 by about 9 and lowers Q by the same factor, to about 5e-7.
What the example shows, and what it does not. It shows that the strong form of Claim 1 (an LC filament with a salt-water core, ringing in the infrared) fails by Lakhovsky's own stated materials; this is why the arbiter's verdict is "Refuted by calculation" and why that verdict does not rest on consensus. It does not show that cells have no electrical oscillations of any kind, nor that external fields cannot affect cells: those are the soft form of the claim, which the arbiter rates "Plausible, untested" as Lakhovsky meant it and "Not testable as stated" in his version (no frequency, amplitude or coupling given). The decisive measurement the arbiter names is terahertz dielectric spectroscopy of isolated chromatin: a measured Q above 1 would revive the model; the calculation predicts none.
A second, quick budget (collars and belts). The open copper spiral is passive: the Swiss patent CH 135072 / FR 657295 itself says it has no energy source of its own [read: Investigator; web https://patents.google.com/patent/CH135072A/fr]. Its ceiling: S = E^2/377. In a strong 1 V/m ambient field S = 2.7e-3 W/m2; times A_e of about 0.3 m2 gives about 0.8 mW. In a typical 0.01 V/m field it is about 8e-8 W [calc]. Against about 100 W of metabolism, a collar delivers a millionth to a billionth of the body's own power. That is why the arbiter rates the collar as an energy supply "Refuted by calculation", while the clinical relief reported from collars remains "Never properly tested" (a different question; see section 6).
6. Evidence table
Verdicts are the arbiter's. "Replicated?" means an independent repeat on record.
| Claim | Tested? | Result | Replicated? | Verdict |
|---|---|---|---|---|
| 1a. Cell nucleus is an LC circuit (salt-water filament, infrared, cosmic-powered) | By calculation only | Q about 1e-6 with his materials; cosmic power 5e8 times too small | n/a | Refuted by calculation |
| 1b. Cells interact with EM fields at selected frequencies (soft form) | Not as he meant it; related modern work (biophotons, TTFields at 200 kHz, Fröhlich proposal) | Parts of this neighbourhood are live research; no frequencies given by him | n/a | Plausible, untested; Not testable as stated in his version |
| 2a. Disease is oscillatory disequilibrium, "war of radiations" | No | No frequency assigned to any disease or cancer line | No | Not testable as stated |
| 2b. Irradiated cholesterol is carcinogenic (Roffo, as relayed: over 55% of "thousands" of mice or rats, none in controls) | Reported once (Roffo, via Lakhovsky) | Species differs between books; formula C21H36O8 unbalanced, unsourced | No direct repeat found; Roffo's heated-fat gastric tumours not confirmed by Kirby, Beck and Peacock, Morris et al. | Never properly tested (heated-fat variants: Tested, not supported) |
| 2c. The carcinogen is "radio-active" | By accepted nuclear physics | UV does not induce radioactivity; Roffo wrote "photo-active and luminescent" | n/a | Refuted by calculation |
| 2d. UV causes skin cancer (background) | Yes, widely | DNA photodamage | Yes | Established |
| 3a. 1924 geranium crown-gall experiment (Gosset, Gutmann, Lakhovsky, Magrou; 150 MHz) | Once: 3 treated, 16 controls | All 3 treated galls began necrosis about 16 days after the first session (exposures 6, 33, 9 h); controls kept growing. No temperature, field or distance recorded; allocation and blinding not stated | No; never repeated with thermometry | Never properly tested |
| 3b. Passive collar/belt as energy supply | By calculation | Microwatts or less vs 100 W | n/a | Refuted by calculation |
| 3c. Clinical relief from collars (1928-29 letters; Attilj, Florence 1928) | Testimonials from users of his company's product; one hospital radiologist's report (Attilj), relayed in the seller's publications | Relief reported; one on-off-on case (ulcers returned when the collar broke, cleared with a new one; Perineau, 6 Dec 1928); Attilj: relapsing hand sarcoma "almost completely disappeared" in two months, relapsing breast sarcoma "no longer palpable", both after earlier surgery, no histology, follow-up at most about six months | No sham control | Never properly tested |
| 3d. 1925 passive copper spiral on a geranium | Once: 1 treated plant, controls undefined | 30 cm open copper ring; controls dead by 30 Jan 1925, tumour shed by end of February, plant 1.40 m tall in March 1928 | No | Never properly tested |
| 3e. Silver spirals sterilise water (d'Arsonval presenting, with Sesari of the Pasteur Institute, 15 April 1929), offered as "proof of cellular oscillation" | Once, one vessel per condition, no replicate | B. coli 1,128,000/cm3 fell to 0 at 25 h; control rose to 43,680,000. Water re-seeded after the silver was removed still killed (12 and 13 colonies, against 946 in the control) unless passed through a porcelain filter (1,474) | No | Killing by silver: Established (as chemistry). As proof of oscillation: Never properly tested (no inert-metal spiral, no silver assay) |
| 3f. Animal reports offered as excluding suggestion (horses, dogs, a filly) | Owner, trainer and vet reports | Seven horses with "more red cells, lower viscosity" (direction only, no values); originals not obtained | No | Never properly tested |
| 4a. MWO emits several simultaneous damped frequencies | No spectrum measurement on record | Credible for a spark-excited ring set | No | Plausible, untested |
| 4b. Stated spectrum 10 cm to 400 m, harmonics to infrared | Internal check | "3 m to infrared" vs "10 cm to 400 m" in one paragraph; 400 m needs tens of nF across a 1 m ring | n/a | Internally inconsistent |
| 4c. MWO cures cancer, goitre, prostate, duodenal ulcers | Case series by Lakhovsky and by named clinicians (Saint-Louis hospital cases; de Cigna, Genoa: five skin cases and eight radiographed ulcer cases, later "18"; Rigaux; Dr. H.; an unnamed New York hospital and Brooklyn urologist, 1941) | Biopsy-confirmed skin cancers healed in weeks; de Cigna's ulcer patients also had diet, lemon juice, medication stopped or nasal cauterisation; five patients with advanced cancer died of haemorrhage after tumour necrosis that followed four or five treatments (three in Lakhovsky's account, two from Rigaux); no denominator; prior X-ray/radium in some | No controlled or sham trial on record | Never properly tested |
| 4d. "No contra-indication ever reported" / "not the slightest accident" | Internal check against the same book | The same book records the five haemorrhage deaths; Gentile lists acute phlebitis as a contra-indication | n/a | Internally inconsistent as printed; whether the device caused any death: Not testable from this record |
| 5a. Radium's decay driven by cosmic waves; 25 m lead sphere would stop it | Indirectly | Decay constants unchanged underground (Gran Sasso, ppm level) | Many laboratories | Tested, not supported (energy-source reading: Refuted by calculation); passage Internally inconsistent |
| 5b. Body heat from cosmic waves | By calculation | Ionising dose 0.17 J/yr vs 3.2e9 J/yr metabolism | n/a | Refuted by calculation |
| 5c. Young skin screens carcinogenic radiation until 40/45/50 | No controlled test | Three different ages in his texts | n/a | Internally inconsistent |
| 5d. Cancer geography follows soil electrical constants (Paris 1921-22) | Ecological comparison by him | Sand quartiers low (0.3-1.1 per 1000), clay/marl high (1.41-2.08); 6 vs 8 groups separate, exact p = 3.3e-4 [calc, arbiter] but post hoc, not age-standardised, hospital and cemetery districts among the high | No | Never properly tested |
| 5e. Patients' improvement stalled for about 8 days at full moon (unnamed Paris professor, weekly weight, blood count and blood pressure) | Once, no data printed | No numbers, dates or patient count; one lunar cycle. An 8-day stall overlaps an 8-day full phase by chance about half the time (15 of 29.5 days) [calc, arbiter] | No | Never properly tested |
| 5f. Collar wearers on "conducting" (high-cancer) soils respond faster than on "insulating" ones (Grenelle, Javel, Auteuil, Neuilly vs Gaillon, Champs-Elysees, Dauphine, Monceau) | No counts or times given | The two lists are the two ends of his own 1921-22 mortality table; the table holds deaths, not treatment response, and no soil conductivity was measured | No | Never properly tested |
| 5g. A distant water supply brings cancer, local wells protect (Simeray's village, 1927; Thiais 3.36 vs Orly 0.36 per 1000 on the same Brie limestone) | Two observations, one second-hand | Thiais/Orly contradicted his soil rule ("did not appear to be in accord with my theory") and the water cause was found afterwards; populations, counts and water conductivity not given | No | Never properly tested (electrical mechanism: Not testable as stated) |
| 6a. Heredity fixed by nuclear wavelength | Superseded by DNA genetics | No wavelength given | n/a | Historically superseded but instructive |
| 6b. Treating mothers gives healthier children | Three anecdotes | High base rate of healthy births | No | Never properly tested |
| 7. Later citation (Winey, US 2007/0258329, applicant with a commercial stake): cured patients "until 1942", precursor of Rife and Priore, "polychromatic" spectrum | No experiment | A 1939 citation cannot document 1939-42 | n/a | Not testable as stated; citation overstates its source |
| 8a. Radio transmitters disorient homing pigeons (Paterna, Valencia, 1924 and 7 Nov 1926, under Spanish military control, report by M. J. Casamajor; Kreuznach, March 1926) | Reported by others, as relayed by Lakhovsky | During transmission birds released 8 km away took no direction; within minutes of switch-off they flew home. No bird count, power, frequency or scoring given; Casamajor report not seen | None in the record; related modern work: Engels et al. 2014 (robins, 2 kHz-5 MHz noise) | Never properly tested |
| 8b. Animals navigate by radiation they emit; semicircular canals are a radio direction-finder | No | No frequency, power or detector given; the canals are known rotation sensors | n/a | Not testable as stated (canal role unsupported by their known function) |
| 8c. Female oak-bombyx attracts males by cell radiation (cotton leaf she lay on, then dipped in alcohol or corrosive sublimate) | Once, by him | No counts; alcohol removes odorants, so odour (a pheromone, characterised for the silkworm moth in 1959) is not excluded | No | Never properly tested |
| 8d. Shaking a duck's wing charges it to about 600 V | One electrometer reading | Frictional charging is ordinary; the step to 25,000 V in flight and altitude "regulation" is assertion | No | Plausible, untested (flight-potential regulation: Not testable as stated) |
Material kept on the record whatever the verdict (SOURCE HONESTY). - The 1924 note is a real four-author communication to a learned society, with a stated control group; Kotzareff, a physician with a rival method, called it "the only experimental study of its kind" (1931) [read: Investigator]. The only power figure is 40 W absorbed (input) power, from Kotzareff's conversation with Lakhovsky, beside Kotzareff's view that 200-300 W useful power was needed for a sure effect on human cancer [read: Kotzareff 1931, via physicist]. The arbiter's bound: far-field heating of a thin stem by 40 W is at most a few kelvin at 0.1 m and negligible beyond 0.3 m, but near-field coupling and the unknown geometry mean heating is neither shown nor excluded. Crown-gall formation fails above about 29-32 C [web search summary, medium], so modest warming is a live hypothesis; so is a non-thermal effect (Roussakow 2013, a historian with that position [web: https://onlinelibrary.wiley.com/doi/10.1155/2013/428027]). - The closest mainstream parallel: Schereschewsky (US Public Health Service), 1926-28, 8-135 MHz fields destroying mouse tumour grafts "without substantial heating", then accepting a thermal interpretation in 1933 after criticism [read: Investigator via Roussakow]. The field then went thermal without a decisive experiment on Lakhovsky's plants. - Lakhovsky says he abandoned the continuous 2 m device in 1929 for damped pulses to avoid prolonged heating [read: Waves That Heal p.30, per contrarian]: the thermal question was in his mind. - Nature, 26 Oct 1940 (reviewer J. Ritchie) separated clinic from 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"; the same review identified Lakhovsky's Corynactis "oscillating circuit" figure as the thread of a cnidocyst (stinging cell) [read: Investigator, research/2026-10-08-internet-archive.md]. - Named cases: Mme C. (68, biopsy-confirmed rodent ulcer, 8 Sept to 19 Nov 1931); M.M. (80, naevo-carcinoma, alive at 90 in 1941); Mme S. (82, epithelioma after operation and radium, 1932, healed in 16-35 days depending on the account); J.S. (61, basal-cell carcinoma, 13 Oct to 29 Dec 1931); a goitre (7 weeks, 1938); prostate cases; de Cigna's duodenal ulcer series (Genoa Academy of Medicine; Vienna short-wave congress, July 1937). Weaknesses as the arbiter lists them: no denominator, maker's best cases, slow-growing skin cancers, prior radiation, silver nitrate applied under the apparatus (Rigaux), fluctuating conditions, retellings that differ in counts and dates. - Smaller case reports, kept but not load-bearing (none changes a verdict) [read: arbiter, deep/Radiations and Waves pp.6-7, 27-29]. Hulin, a Salpetriere dentist (Paris Society of Odontology, 20 July 1933), showed four radiographs of one patient's pyorrhoea tooth re-ossifying. Weight: low. It is Lakhovsky's summary of Hulin, with no follow-up after the fourth film. The oscillator was used to drive calcium-chloride ionisation after six sclerosing injections, so its own share cannot be separated. An English lady of about 52 (1935) had a white-cell (leucocyte) count of 20,000 that fell to 5,000 after a week of 7-minute daily sessions, and her fibroma operation then went ahead. Weight: low. There are two counts only, the cause of the high count is not stated, a raised count often falls by itself once its cause passes [memory, high], and the surgeon's praise is reported by Lakhovsky alone. Institute of Biological Physics cases Nos. 1-4 (December 1939 to March 1940; a man with a nipple tumour called "neoplasic scirrhus", an anaemic radium worker, a skin plaque and a lip growth) were chosen "at random" by the chief of staff of Lakhovsky's own institute. Weight: low. Only No. 1 has a blood count (red cells under 3 million rising to 4.8 million, unit not printed), there is no histology, follow-up runs from one week to one month, and the outcomes are phrased "we believe". - De Cigna's series, read in full [read: arbiter QE revision, from Radiations and Waves pp.30-39]. Five skin cases: P.A. (basal-cell epithelioma, healed at 24 days, unchanged at three years), G.C. (histology; lost to follow-up), O.G. (no histology; "ulcerated wart" by prudence; healed in 7 days), S.G. (histology, basal-cell; pain gone, lesion "remained fixed", not claimed cured), P.N. (lupus of 20 years after X-ray, Curie, Finsen, cryotherapy and gold salts; "all lesions have disappeared"). In the ulcer cases he also stopped all other medication (P.G., R.A., Father N.), gave his own lemon-juice regimen (up to four large glasses a day for 25 days, which he says "always given me immediate results" for gastric pain) and strict diets, treated constipation by nasal reflexotherapy or Bonnier cauterisation (R.A., D.F.), and in the Carmelite case rest stopped a bleed before the MWO. Only P.G. has a post-treatment radiograph quoted. His own caveats: "six cases are not sufficient, and the time elapsed is too short to speak of a positive cure"; "the oscillatory factor ... certainly has the most significant value, but for the moment, we cannot completely prove it". No adverse event in any of his cases. It remains the best-documented series in the record, with more than one candidate cause for the ulcer relief on his own account. - Deaths after treatment [read: Radiations and Waves pp.41-43, via arbiter QE revision]. Two throat-cancer patients at Saint-Louis (December 1931) had four sessions each and were found haemorrhaging when Lakhovsky came for the fifth; both died. Captain H. at Val-de-Grâce (thoracic tumour of 8-10 kg that necrosed, a piece "the size of a new born baby's head" detaching at dressings) died of haemorrhage in January 1932. Rigaux reported two further throat cases: "the hemorrhage and death always followed the fourth treatment". Lakhovsky's explanation: the necrosing tumour freed arteries it had eroded. Two readings fit and the record cannot separate them: treatment-induced necrosis (evidence that the device acted on tumours, and an adverse-event signal), or the late course of the disease (fatal erosion bleeding also occurs in untreated advanced head-and-neck cancer [memory, high]). These deaths sit beside the book's "no contra-indication" and "three failures ... in innumerable treatments" (no denominator). The directions call for a pause of two to three weeks after the fourth treatment; the book does not connect this to the deaths. - Attilj (hospital radiologist, Rome; Florence radiology congress, May 1928), passive bracelet or waist circuit: a relapsing hand sarcoma "almost completely disappeared" in two months; a relapsing breast sarcoma became "no longer palpable"; a scar recurrence after mastectomy and arm oedema were relieved [read: Waves That Heal pp.21-23; Secret of Life p.58]. Both sarcoma patients had earlier surgery (the breast case also X-ray), there is no histology, and Attilj called it "merely the beginning". His lists differ between the two books (four patients and six). Single-source regression reports from a named clinician, not the seller, relayed in the seller's publications; no control. - Animal reports, offered by the books as excluding suggestion [read: Waves That Heal pp.9-12, 15]: P. Fournier-Ormonde (Institut Physiologique de la Croix Blanche, Vaucresson) reported seven horses fitted with collars with more red cells and lower blood viscosity (direction only, no values); a Brisbane dog (January 1940), a Cannes fox-terrier, a parrot and the racing filly Ballerina. The arbiter's reply: suggestion acts on the people who report (owners, trainers, vets), not on the animal, and there were no controls or blinding. - Silver-spiral sterilisation [read: Secret of Life pp.35-36]: kept as a real measurement by a Pasteur Institute collaborator, controlled at one time point. Its own Experiment III shows the killing agent stays in the water after the silver is removed and survives heating but is lost on porcelain filtration. That fits a dissolved silver agent (the oligodynamic effect, known since the 1890s [memory, high]) or a particle-bound one; neither is oscillation. Lakhovsky also held patents on silver sterilisation (FR 688887, CH 147806, FR 810280, FR 866036) [read: Investigator, sites.md]. - MWO operating data and dose [read: Radiations and Waves p.48, via arbiter]: mains 110 or 220 V at 2 to 4 A (220 to 880 VA, the only power figure on record), three power steps, transmitter and receiver 1.3-1.5 m apart with the patient between, sessions of 5-12 minutes. The radiated fraction is unknown. If 1 to 10 percent is radiated, the body takes up roughly 0.005 to 0.1 W/kg against a resting metabolism near 1.4 W/kg [calc, arbiter], so whole-body far-field heating is small across that range. The patient sits in the near field of two resonators with corona, where this estimate is unreliable; Rigaux and Lakhovsky both name sparks and "effluvia" as acting agents. - Priority and scale: the 1941 preface (Presner) says the Venice 1934 and Vienna 1937 congresses recognised Lakhovsky as the first inventor of the therapeutic short-wave oscillator and that "more than 50,000" such units were in use; no proceedings or figures are given, and both refer to his earlier 2-10 m apparatus, not the MWO. Kotzareff (1931), an independent rival, credits him ahead of Schliephake. Recorded as an interested party's characterisation; the congress proceedings would settle it. Lakhovsky also cites Schliephake (Giessen) and Esau (Jena), 1928-29, on short waves killing mice and sterilising milk as his reason for leaving thermal short waves [read: Radiations and Waves p.10, via qe.md]. - Claim 6 in his own words [read: Radiations and Waves pp.65-66]: the three children of treated mothers were "a new race of unusual strength and perfection of form"; a member of a Paris birth-rate committee proposed, and Lakhovsky reports with approval, an official municipal decree requiring "all engaged couples, a month or two before marriage" to have at least one free treatment a week at the City Hall, continued through pregnancy, so that "a race could be created of strong, perfect children, free from all hereditary stains"; he claims that the 35-45 percent of young men declared unfit for military service would fall to "5 per cent at the maximum". He also concedes "three isolated cases are insufficient" and proposes a test on some twenty pregnant patients in one maternity hospital, which is not on record as done. - Side-predictions of the soil theory (rows 5e-5g; arbiter, "Claim 5, soil section") [read: Secret of Life pp.51-59; Contributo table]. These are three testable predictions that Lakhovsky printed himself: a full-moon pause in recovery, faster recovery on "conducting" soils, and cancer following a change to distant water. None comes with data that tests it. The most instructive is Thiais and Orly. Both communes sit on Brie limestone, yet their cancer death rates differ tenfold (3.36 against 0.36 per 1,000 per year). Lakhovsky says the pair "did not appear to be in accord with my theory" and then explained it by their water: Seine water at Thiais, local wells at Orly. His rock classes also move. Limestone is "insulating" and low-cancer on pp.51 and 53, but Brie limestone is "a fairly good conductor" and high-cancer on p.55, and no conductivity was ever measured. A rock can therefore be placed in whichever class fits its cancer figure. For a learner, this is the lesson. A rule that absorbs every counter-example with a new cause stops being testable. Small communes also give noisy rates: at 0.36 per 1,000, a town of 2,000 records about one or two cancer deaths in two years [calc, illustrative population]. The soil-geography verdict stays Never properly tested. - Where the record goes dark: Lakhovsky left France in 1941 and died on 31 Aug 1942 in Brooklyn, three days after being struck 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 yet); US 2,351,055 was granted posthumously on 13 June 1944, assigned to Henry S. Blum of Chicago [read: Investigator; deep/2351055]. The 1941 book prints a 19-patient table with a result for each patient and seven case histories from an unnamed New York hospital (1 July-21 August 1941; mostly arthritis and soft-tissue conditions; four of the seven histories end "did not return"; the hospital withheld its own report pending a presentation to the New York Society of Physiotherapists), and an 8-case table from an unnamed Brooklyn urologist who had bought a machine; the hospital's own report, any follow-up and any control are not on record [read: Radiations and Waves pp.45-47, via arbiter]. No FDA or Post Office file and no follow-up of the 1930s clinical series has been found. The arbiter sets out competing explanations (death, war and displacement; no controlled results existed; medicine moving to radiotherapy and to the thermal view; regulatory or professional pressure, not shown for him; commercial continuation under another name; suppression, not shown and not excluded) and the records that would decide (NYC death certificate and 1942 obituaries, 1945 incorporation, Blum records, French RCS for COLYSA, Gallica copies of C. R. Soc. Biol. 91 and C. R. Acad. Sci. 188 pp.657-659, the de Cigna and Attilj reports). Google showed Lumen legal-removal notices on three Lakhovsky queries; content unknown; no inference drawn.
7. What a learner should take away
Solid. The radio engineering is real and, in his arithmetic, usually correct. A spark-excited set of open rings does emit many damped frequencies (credible, never measured on a patent-built device). Cells are electrical, and the general question he raised (do cells respond to fields at particular frequencies?) is live research today in forms he did not foresee. The 1924 geranium note is a genuine, controlled, co-signed experiment with a striking result.
Wrong, and why the mistake is instructive. The strong cell-circuit model fails because an analogy was carried across without its hidden conditions. A tuned circuit needs a good conductor (the worked example: Q of 50 for copper, 4e-6 for salt water), a signal above thermal noise, and a power source. Each condition is checkable with undergraduate physics, and each fails by many orders of magnitude for a wet nucleus at 310 K. The general lesson: when you borrow a formula, list the assumptions it was derived under and check each one in the new setting. The same habit catches the cosmic-ray body-heat claim (5e8 short), the passive collar (microwatts against 100 W) and the radium claim (decay constants do not change underground).
Watch the citations. Several supports shrink when traced: Roffo's "photo-active and luminescent" agent became "radio-active"; Roffo's "compatible with" became "confirmed"; the Hatonn editor's insertion "[approximately 100-200 million cycles per second]" for 2-10 m is wrong (it is 30-150 MHz) and is not Lakhovsky; the Winey application's "cured ... until 1942" is cited to a 1939 source.
Genuinely open. 1. What killed the geranium tumours: heat, a non-thermal RF effect, or chance and selection? Only a blind, randomised repeat with in-gall thermometry, sham and heated-only controls can tell (Pelargonium crown gall, 150 MHz, about 40 W, at least 10 plants per arm). 2. Did the MWO help the patients reported in the 1930s, and if so by what: heat, corona by-products (ozone, nitrogen oxides, ions), non-thermal RF, or placebo and natural course? Nothing in the record separates them; de Cigna's series (eight radiographed ulcer cases with co-treatments, five skin cases) is the best documented, and his original Genoa Academy and Vienna 1937 files have not been retrieved. Were the five haemorrhage deaths caused by treatment-induced necrosis or by the disease, and were they all of the advanced cases treated or a few among many? 3. What happened to the US venture after 1942, and why did a device said to cure cancer leave no trial record? 4. Why do 6 h, 9 h and 33 h exposures all give necrosis onset near 16 days? 5. Do radio transmitters disturb homing birds, and by what mechanism? Lakhovsky relays on-off pigeon trials at Paterna (1924-26) and Kreuznach (1926), never replicated in the record. The modern lead: Engels et al., Nature 509:353-356 (2014), found that European robins at Oldenburg lost their magnetic-compass orientation in broadband electromagnetic noise (about 2 kHz to 5 MHz, well below safety limits) and regained it when shielded, reproduced under blind 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]. That is the same class of observation in another species, organ and band; it does not test the pigeon reports or Lakhovsky's mechanism. Deciding records: Casamajor's Paterna report and the Kreuznach reports, then a blinded transmitter on/off release with power and field measured. 6. What did the silver-spiral water contain after exposure? A repeat with silver, an inert metal and a silver-free spiral, with dissolved-silver assay, would decide between ions and anything else.
8. Check yourself
- Lakhovsky's Radio-cellulo-oscillator worked at a wavelength of 2 m. What frequency is that, and how far is it (as a ratio) from a 10 um "cell wavelength"?
- A series RLC circuit has L = 8.6e-12 H, C = 1.2e-16 F and R = 534 ohm. Will it ring? What if R = 5 ohm? What if R = 6.4e7 ohm?
- A collar acts as a half-wave antenna with effective area 0.3 m2 in an ambient 0.1 V/m field. What is the most power it can deliver? Compare with metabolism.
- In the 1924 experiment, 3 plants were treated and 16 left as controls; all 3 treated tumours necrosed and no control did. Under random allocation and no treatment effect, what is the probability of that outcome? Name two things the note does not tell us that would make this number meaningless.
- Lakhovsky said a 25 m lead sphere around radium would stop its radioactivity, because the cosmic waves that drive it would be shut out. Describe a modern measurement that tests this, and what it found.
9. Further reading
Mainstream. - D. J. Griffiths, Introduction to Electrodynamics: ch. 7 (Ohm's law, inductance), ch. 9 (waves), ch. 11 (radiation, dipoles). - J. R. Taylor, Classical Mechanics, ch. 5 (damped and driven oscillators; the same mathematics as the RLC circuit). - E. M. Purcell and D. J. Morin, Electricity and Magnetism, chapter on alternating-current circuits (resonance and Q). - D. V. Schroeder, An Introduction to Thermal Physics (equipartition, blackbody radiation). - K. S. Krane, Introductory Nuclear Physics (alpha decay as tunnelling). - R. K. Adair, "Constraints on biological effects of weak extremely-low-frequency electromagnetic fields", Phys. Rev. A 43, 1039-1048 (1991): the thermal-noise argument [web: citation confirmed by mathematician]. - V. G. Roussakow, 2013, on the history of non-thermal RF oncology, including the 1924 note and Schereschewsky [web: https://onlinelibrary.wiley.com/doi/10.1155/2013/428027]. - R. A. Weinberg, The Biology of Cancer (the mutation model of cancer). - S. Engels et al., "Anthropogenic electromagnetic noise disrupts magnetic compass orientation in a migratory bird", Nature 509:353-356 (2014) [web: citation confirmed by arbiter through index pages; exact title memory, high].
The inventor's own key works (titles as in the project files and Investigator notes). - Le Secret de la Vie (1925); English The Secret of Life (copy in the folder carries non-Lakhovsky editorial layers). - L'Universion (1927); Contributo all'eziologia del cancro (1927, the Paris soil and cancer table); La Sante par les ondes (COLYSA, 1929, collar testimonials); Pour rester jeune a 100 ans (1939); Radiations and Waves, Sources of Our Life (1941). Not found: L'Oscillateur a ondes multiples (1932), La Science et le Bonheur. - Gosset, Gutmann, Lakhovsky and Magrou, C. R. Soc. Biol. 91:626-628 (26 July 1924): the geranium note (not yet obtained in the original). - Lakhovsky and Sesari, silver-spiral sterilisation note presented by d'Arsonval to the Académie des Sciences, 15 April 1929 (abridged in The Secret of Life pp.35-36; the original not obtained). - Patents: FR 657295 / CH 135072 (open circuit, 1927); FR 732276, FR 42146E, CH 164027 (MWO family; the Swiss text names therapy); US 1,962,565 (filed 1931, granted 12 June 1934); US 2,351,055 (tube, filed 1941, granted 1944). - Contemporary assessments: Kotzareff, thesis, Paris 1931; Nature 146, 26 Oct 1940 (J. Ritchie's review). Later secondary: Mark Clement, Waves That Heal; T. J. Brown, MWO Handbook (1994, Borderland Sciences, a publisher with a market for MWO material).
Answers to Check yourself
- f = c/lambda = 3.0e8 / 2 = 1.5e8 Hz = 150 MHz. A 10 um wave is 3.0e13 Hz, so the ratio is 2e5: the cell "frequency" is two hundred thousand times higher than the device's. Nothing in the books shows how one drives the other.
- R_crit = 2 sqrt(L/C) = 534 ohm, so R = 534 ohm is exactly critically damped: no ringing, the fastest non-oscillating return. R = 5 ohm: Q = 267/5 = 53, it rings (this is the copper case). R = 6.4e7 ohm: Q = 4e-6, hugely overdamped (the salt-water case in section 5).
- S = E^2/377 = 0.01/377 = 2.7e-5 W/m2; P = S x A_e = 8e-6 W, about 8 microwatts. Metabolism is about 100 W, so the collar supplies about 1e-7 of the body's own power. Whatever the collar does (if anything), it is not supplying energy.
- 1/C(19,3) = 1/969, about 0.001. It is meaningful only if (a) the 3 plants were chosen at random, not for, say, smaller or older galls, and (b) the person judging necrosis did not know which plants were treated. Also unknown: plant temperature during exposure, distance and field strength, and how often untreated galls necrose on their own.
- Compare a radium-226 source in a surface laboratory and deep underground, where the cosmic flux is cut about a millionfold. At Gran Sasso, 226Ra and 40K showed no change in decay rate at the parts-per-million level [web: https://arxiv.org/abs/1802.09373]. The arbiter notes that paper tests time modulation underground; the direct surface-versus-underground comparison is the agreement of half-life values measured at many depths, plus Curie and Kamerlingh Onnes (1913), who found radium's decay unchanged at about 20 K (search summary, medium confidence). Alpha decay is explained by tunnelling with no external driver.