Meyers-Stan (Stanley A. Meyer, "Water Fuel Cell"): Physicist review (Step 2)

Reviewer: PhD physicist, 2026-10-09. Tags: [read] source text I read myself (page given); [calc] my own arithmetic (python, .venv); [web] URL; [memory] recalled, with confidence. Verdict scale from prompt.txt.

Source gap (repeated from claims.md)

• No digests exist for this project. The folder holds NO measurement, test report or witness statement; every patent, newsletter and the Technical Brief is Meyer's own characterization (stake: patent holder, seller of dealerships and "profit-sharing certificates" [read: claims.md; newsletters No. 10, 11a]).
• Unread by anyone so far: the 2 h 38 min New Zealand 1989 video, the 40 photographs, the scanned newsletters No. 4 and No. 8 (read only as page images by the fact-finder), and Birth pp.57-222 (I have now read Memo 424, Appendix A and scanned the rest for numbers; see below).
• Where a conclusion rests on [memory] alone I say so. The decisive government record (the NIST Office of Energy-Related Inventions evaluation, c.1990-91) was NOT obtained; I rely on a secondary quotation of it.

1. The mainstream view, stated fairly

• Splitting water. Liquid water splits into H2 and O2 when charge is passed through it. The minimum energy is the Gibbs free energy, 237 kJ/mol (reversible voltage 1.23 V). The thermoneutral figure is 286 kJ/mol (the heat of formation of liquid water, 1.48 V) [memory, high; NIST WebBook gives -241.8 kJ/mol for vapour, so liquid = 241.8 + 44 = 285.8; [web] https://webbook.nist.gov/cgi/cbook.cgi?ID=C7732185&Mask=1 confirmed the vapour value only]. Faraday's law ties gas volume to charge: 1 A for 1 h gives about 686 cc of H2+O2 mixture [calc]. Real cells run above 1.23 V, so 100% is a ceiling that practical cells approach but do not pass.
• Pulsed or AC drive. Many groups have studied it. A 2025 analysis in the International Journal of Hydrogen Energy (Puteanus et al., "Efficiency improvement by pulsed water electrolysis: An unjustified hope") argues by Jensen's inequality that pulsing cannot beat steady DC for a standard convex cell model [read: research/files/Puteanus-2025-IJHE-pulsed-electrolysis.pdf, abstract]. Its scope is a PEM-cell model, and it says Shimizu-type ultra-short-pulse papers report gains; "some studies suggest ... others report a decline". So pulsing is an open engineering topic with small effects, not a source of energy.
• Water as a circuit element. At 5 kHz, tap or well water is a resistor, not a capacitor (worked in section 3).
• Energy bookkeeping. Burning the H2 and O2 returns at most what splitting cost (286 kJ/mol). A vehicle fuelled "by water" needs an outside energy source to make the hydrogen.
• Patents. A US patent needs a credible utility and enablement, not a demonstration or a working model. USPTO (MPEP 2107): "Situations where an invention is found to be 'inoperative' and therefore lacking in utility are rare" and an asserted utility is presumed true unless doubted [web: https://www.uspto.gov/web/offices/pac/mpep/s2107.html]. So "granted under Section 101" is not independent confirmation, contrary to Hindley's claim (below).

2. Verdicts at a glance

3. Claim 1: Water capacitor splitting

(a) Mathematics. Meyer's series-LC formulas (Eqs 1-4, 8, V_L and V_C) are textbook-correct in form [read: Birth p.8-10; deep read]. Problems: - Eq 9, "Z = R1 + Z2 + Z3 + R_E, where R_E is the dielectric constant of natural water", adds a dimensionless number (78.54) to ohms: a units error [read: page image p.10, per deep read]. - "At resonance the voltage ... is theoretically infinite" [read: p.9]. In a series RLC circuit the peak voltage gain is Q = (1/R)sqrt(L/C), set by the resistance, and Meyer deliberately builds the chokes from resistive wire. - Is the water a capacitor at these frequencies? [calc] Take Meyer's example cell: 0.5 in rod in a 0.75 in tube, 3 in long, water permittivity 78 [read: WO 92/07861 p.9, p.14]. Coaxial capacitance C = 2π ε0 εr l / ln(b/a) = 820 pF. Resonance at 5 kHz then needs L about 1.2 H, a plausible choke. But tap water of 500 µS/cm (0.05 S/m) gives a cell conductance path of about 17 ohm, while the capacitor's reactance at 5 kHz is 1/(2πfC) = 39 kohm. The current is 2000 times more resistive than capacitive. The voltage-boost resonance needs the water resistance to exceed about 39 kohm, i.e. conductivity below about 0.2 µS/cm, which is near theoretical pure water (0.055 µS/cm). Water's own dielectric relaxation is at tens of GHz, not kHz [memory, high]. So "tuning into the dielectric properties of water" at kHz has no basis for tap water; for very pure water a conventional LC resonance with the cell capacitance is possible, and that is just a resonant charging circuit (a well-known topology; the WO 92/07861 search report cites prior art Mishiro, US 4,275,363 against its resonance claims [read: p.30]). The 1991 Electronics World description (inductor plus cell capacitance, a rectifier diode, a "rising staircase DC potential" until the "water breaks down and a momentary high current flows") describes exactly that charge pump [read: Electronics World Jan 1991 p.10-11, pdftotext]. The word "breaks down" there means spark/arc breakdown, not a quiet molecular process. - Is the field strong enough to break bonds? [calc] Meyer's drive: 650 V at 5 kHz across a 3.2 mm gap (the coaxial gap is (0.75-0.5)/2 in) is 2 x 10^5 V/m; even his "20,000 V" is 6 x 10^6 V/m. A covalent O-H bond has a binding energy of about 4.8 eV over about 0.1 nm, i.e. an equivalent field of order 10^10 V/m [memory, medium; order-of-magnitude]. The applied field is about 10^3 to 10^5 times too small to tear the bond directly. At such fields water behaves as an ionic conductor; what happens at the electrode surface is ordinary electrochemistry (Faradaic reaction at about 1.5-2 V of surface overpotential). The statement that covalent bonding "ceases to exist" by "pulling away orbital electrons" [read: Birth p.13] has no quantitative support in the document. - Meyer's own energy table. In his 1995 "Independent Test-Evaluation Report" (pp.59-60, transcribed at stanslegacy.com [web: https://stanslegacy.com/documents/ and Investigator note research/2026-10-09-stanslegacy.md; unverified machine transcription]) he gives "12.5 VDC @ 40 amps = 500 watts", 9 tubes, "46,666 cc/hr per tube" at 12.5 V x 4.4 A = 55 W, and compares it with "electrolysis ... 10 cc gas per 1 amp hr", obtaining a "Differential Efficiency Factor" of 1,696 (later 301). Check [calc]: 1. Faraday's law at 4.4 A gives 3.0 L/h of gas per tube. Meyer's figure is 15.5 times that. If real, 46.7 L/h of mixture holds 1.27 mol H2 = 363 kJ/h = 101 W of combustion energy from a 55 W input: a violation of energy conservation, which Meyer does not discuss. 2. The gas rate came from "1 liter cavity @ 7 lbs/min" with "1 liter cavity = 1,000 cc of gas volume = 1 lb gas pressure" [read: transcribed p.60]. A pressure reading in lb (psi) is not a volume. If a 1 L cavity rises by 7 psi per minute, the gas amounts to 7/14.7 L at atmospheric pressure = 0.48 L/min = 28.6 L/h for 9 tubes = 3.2 L/h per tube. That equals the Faraday value of 3.0 L/h per tube within 6%, i.e. ordinary electrolysis at 4.4 A per tube, and not at "milliamps". This reading depends on the unverified transcription (OCR) and on assuming gauge psi; it is my reading, not Meyer's statement. The original scan should be checked. 3. The "prior art electrolysis" baseline of 10 cc per amp-hour is about 70 times below the true Faraday value (686 cc per amp-hour [calc]). That artificially small baseline produces the factor 1,696. 4. Also, 40 A on a 12.5 V supply is not "restricting amps"; and the meters were a clamp ammeter and a low-cost voltmeter, which misread pulsed spikes [memory, medium]. - Conclusion for (a): the circuit formulas are fine; the water-as-capacitor picture, the 20,000 V claim and the energy table are not internally consistent.

(b) Established experiments. Faraday's laws and the 1.23 V/1.48 V limits; the long literature on pulsed and AC electrolysis (the 2025 IJHE paper above is a recent review-style argument); water's electrical conductivity and relaxation [memory, high]. Water dissociation by the second Wien effect needs fields around 10^7 to 10^8 V/m, still far from the covalent scale [memory, medium].

(c) Tests. Evidence the Investigator found, each weighed: - Electronics World / Wireless World, Jan 1991, pp.10-12 (witnesses Prof. Laughton of Queen Mary College, Admiral Sir Anthony Griffin, Dr Keith Hindley, Oct 1990) [read: pdftotext of research/files/Electronics-World-1991-01.pdf]. They report gas at "no greater than a tenth of an amp on the ammeter, and claimed to be milliamps by Meyer", a "pale cream and dark brown precipitate", a cell that stayed cold, and an H2/O2 flame that melted a steel bar. My assessment: (i) the witnesses measured neither input energy nor gas volume; the magazine itself asks "what is the precise total cell energy input required to deliver a unit volume of dry gas mixture?" and calls the technique one with "more holes in it than a sieve". (ii) 0.1 A with no measured volume rate is compatible with ordinary electrolysis; the precipitate (Hindley attributes it to chlorine attacking the stainless steel) is direct evidence of Faradaic chemistry at the electrodes, which Meyer says is absent. (iii) Melting steel with an oxy-hydrogen flame needs no unusual gas: that flame is about 2800 C [memory, high] at any flow rate. (iv) A cold cell is expected if the power is small. (v) The magazine's statement that pure water has "a dielectric constant of about 5" is wrong (about 78-80 [memory, high]). (vi) Hindley's claim that patents "granted under Section 101" are "independent, critical, scientific and engineering confirmation" is his characterization and does not match USPTO practice (see section 1). This is genuine eyewitness testimony by credentialed people and I keep it in the record; it is not a measurement of efficiency. - Perkins Technology bench observation, 8 Oct 1992 ("dissociation observed to begin at about 10 volts ... maximum observed current less than 1 milliamp ... vessel at room temperature") [web, via Investigator note on stanslegacy.com, machine transcript of Meyer's own binder]. Again qualitative, no volume rate. The same Perkins note says "On none of the 3 occasions ... when UK experts visited Ohio, was there a practical demonstration of a running internal combustion engine fuelled by water" and "A substantial number of UK scientists and engineers regard WFC technology as in breach of the first and second laws of thermodynamics", and records that a Queen Mary and Westfield DTI-sponsored report "were unable to reconcile Meyer's WFC descriptions ... with currently established scientific doctrine ... However they left open the possibility" [web, same]. Source: a party that wanted to proceed and which had a $125k deposit on the table; reported as their characterization. - Nagypal (Vienna), 4 Jun 1991 [web, binder p.66-68]. This is the best quantified item. A 12 V supply raised to about 600 V bursts, 0.8 H chokes, about 10 nF cell, plates 3 x 2 in with a 1.6 mm gap; "consumes about 20 mA and produces about 0.5-1 cubic centimeter gas" per minute, "your theory works", "relative effectivity about 300%" against 10-12 W for a conventional cell; he "can't reproduce" Meyer's waveforms and "it works in another way". My check [calc]: 0.75 cc/min of mixed gas needs only 0.1 W to split (5.8 J/min). If his comparison cell used 10-12 W for that gas, his "300%" is about 3-4 W against a 0.1 W thermodynamic minimum: roughly 3% efficient, i.e. nothing above the mainstream limit. At 20 mA, Faraday's law predicts only 0.23 cc/min of mixed gas. His 0.5-1 cc/min is 2-4 times that, if the 20 mA is the current through the water; but it is unclear whether 20 mA is the supply-side average, the cell current or an rms value, and cc readings from a bubble count are crude (dissolved air release and heating also give gas). I cannot call this a violation; it is an unresolved discrepancy of a factor of 2-4 in a crude, uncalibrated measurement. He offered to be a distributor (interested party). - Dave Lawton / Dr Cramton builds (2006 onward), 6 tube pairs, currents about 1-2 A, no gas rate or energy figures [web, Investigator]. Builders' reports; not controlled tests. - NIST Office of Energy-Related Inventions (c.1990-91): Spy, Dec 1991, quotes NIST: the WFC "would be a net consumer of energy, not an energy producer"; Air & Space Oct/Nov 1991: "The verdict was clear"; Heard (2000) cites a NIST letter of 28 Jun 1991 [web, via Investigator's Google Books snippets; not obtained]. This is a US agency's characterization, with no stake in the outcome that I can find, but the evaluation itself (method, measurements) is unseen. - Court record (Fayette County CVH19930292, Willis and Harper v. Meyer, 1996, judge reportedly found fraud, $40,515.72 garnished) [web, stanslegacy docket transcription]. The decision transcript is not obtained, so I cannot say what the court found about the physics; Meyer's summary (court said "only electrolysis") is his own account. Exhibits and depositions were destroyed in 2014. The court is the plaintiffs' dealers' dispute over sales representations; it is weak evidence about the physics either way. - Not tested neutrally: no controlled test with measured input energy and measured gas volume, by anyone without a stake, is in the folder or the Investigator's finds. Meyer's patents say demonstrations were given, but the demonstrations were never recorded in the court (his claim) and no data exists. Absence of a record where one is expected is an anomaly; competing explanations: (1) the effect is ordinary electrolysis and nothing remarkable existed; (2) trade secrecy (Meyer withheld waveforms, per Electronics World and Nagypal); (3) records lost (court exhibits destroyed 2014; estate dispersed); (4) suppression (asserted in the Dispatch 2007 retelling by family, not documented). The records that would decide: the NIST file, the court's decision transcript, the 1998 police and coroner files, the probate valuation of "Water Fuel Cell: sole proprietorship" (15 Sep 1999).

(d) If true. Energy would be released from water without a matching input: a violation of the first law (or a new energy source), since burning the gas returns 286 kJ/mol. For the weak form (some Faradaic gas from pulsed drive) nothing changes. For non-Faradaic bond breaking at 10^5 V/m the textbook picture of water's bonding and of field-induced dissociation (which needs >10^7 V/m) would have to change.

Verdict: split as in the table. Reasons for the strong-form calculation verdict: field 10^3 to 10^5 below bond scale; Meyer's own table, read with correct units, matches ordinary electrolysis; measured quantities in the record (Hindley, Perkins, Nagypal) do not exceed Faraday's limit by more than a factor of 2-4 in the best (unreliable) case.

4. Claim 2: "Hydrogen fracturing", energy beyond combustion

(a) The claim (2.5 million barrels per gallon, Birth pp.6, 17, 117; 44,000 to 108,000 barrels in 1987-88 newsletters; "over two million" 1996) is stated with Eq E = M_d c^2 and "nuclei decay" in the figure [read: Birth p.22 page image per deep read]. Memo 424, "Atomic Energy Balance of Water", which Step 1 did not read, is qualitative: "Energy Aperture", "Universal Energy", "Gyroscopic Action" of nuclear particles, "Energy Pumping Action"; it contains no derivation of 2.5 million barrels, no equation beyond E = hv and p = hv/c, and no measurement [read: Birth pp.117-123]. Appendix A (Tab 36-39) gives engine arithmetic but not an energy balance. The documents say both that the process "does not create energy ... thus complying with the Laws of Thermodynamics" and that it releases "Universal or Zero Point energy" through an "energy aperture" in the nucleus [read: Newsletter 11a pp.2, 9]. Those two statements are in tension. Calculation [calc]: one US gallon of water is 3.785 kg. Chemical value of its hydrogen burned back to water: 0.01 barrel of oil (at 6.1 GJ per barrel [memory, high]). 2.5 million barrels = 1.5 x 10^16 J, which is 2.5 x 10^8 times the chemical value and 4.5% of the entire rest-mass energy of the gallon (3.4 x 10^17 J). Even complete fusion of all the hydrogen in the gallon to helium (0.42 kg of hydrogen, about 0.7% mass defect [memory, high]) gives 2.7 x 10^14 J = 44,000 barrels. That equals Meyer's lowest quoted figure by coincidence (I do not claim he derived it that way, and nothing in the documents says so). 2.5 million exceeds the maximum from fusing every hydrogen nucleus by a factor of 57 and would require converting oxygen's mass too. No chemical or electrical process (energies of eV per atom) can touch nuclear binding energies (MeV) at room temperature; this is the accepted account of why "cold" nuclear release needs a nuclear mechanism, and none is described. (b) Consistent with no experiment. Ordinary hydrogen/oxygen combustion energy is well measured; Meyer's flame temperatures (">5,000 F" [read: Birth p.34]; adiabatic flame temperature of H2/O2 is about 3100 K = 5000 F [memory, medium], so this is the ordinary oxyhydrogen limit) are not in conflict with it. (c) No test of the energy yield exists in the folder. The Kansas State and Northeastern "water explosion" reports that Meyer cites in his binder (Appendix A "Mode of Operability") describe electrically exploded water, with the energy that was put in; I have not seen them (cited only) and so cannot say what they measure [read: Birth Appx A p.04]. Kucherov, US 5,632,870 (cites Meyer; an ambient-heat cell) [web via Investigator] is a different claim, not a test. (d) The mass-energy balance of ordinary matter, and the conservation of baryon number at the energies involved, would be violated if true. Verdict: Refuted by calculation (strong form); also Internally inconsistent (thermodynamics statement vs "Zero Point" statement).

5. Claim 3: Water-fuel injection and vehicles

(a/b) Meyer's own Appendix A [read: Birth pp.220-221, transcribed]: a 50 hp VW engine at 65 mph uses "44.4 ml/min" of water = 0.74 ml/s; 100 injection cycles/s (3,000 rpm, 4 injectors) gives 7.4 µl per cycle; the voltage circuit draws "40,000 volts @ 1 ma = 40 watts"; "water is 2.5 times more powerful than gasoline (US National Bureau of Standards)". My checks [calc]: - Energy per injection: 50 hp = 37.3 kW, so 373 J of shaft work per injection cycle. 7.4 µl of water = 7.4 mg = 0.41 mmol; its absolute maximum chemical energy (if all burned as H2) is 117 J, and 117 J was needed to make it. With an engine efficiency of 25-30% the engine needs 1.2-1.5 kJ of fuel energy per cycle. The shortfall is a factor of about 10 to 13 even at 100% conversion, before counting the splitting energy. - At the stated rate, 0.74 g/s of water contains at most 11.7 kW of combustible energy, versus 37 kW of shaft power; and supplying that energy to split the water costs the same 11.7 kW. - "Water holds 2.5 times more hydrogen than gasoline" [read: Newsletter 10 p.15, 11a p.8] counts hydrogen atoms; gasoline's energy is carried by its carbon-hydrogen bonds' oxidation (about 44 MJ/kg [memory, high]) and water has none left to burn. Compare: hydrogen's energy in a kilogram of water is 0.111 kg x 142 MJ/kg = 16 MJ, but the water is spent hydrogen. - WO 92/07861 states "about 5 gallons of water per hour" converted at 0.65 W of drive power (1 mA x 650 V) [read: p.13-14, deep read]; splitting 5 gal/h (0.29 mol/s) needs 83 kW at 286 kJ/mol [calc]. A factor of about 10^5 between the claim and the thermodynamic minimum, with a patent text written in the future tense ("will occur") and no measurement. (c) Demonstrations: ABC television news on the dune buggy (1985, claimed; not in the folder). Charles Hughes, a family acquaintance, told the Columbus Dispatch (2007) that Meyer's tractor ran 15 minutes on well water [web via Investigator, https://www.dispatch.com/story/news/2007/07/08/the-car-that-ran-on/987361007/; read only through the Investigator note]; that is a recollection, a single published source; no second witness found. The Perkins note records three UK visits (1990, 1992 x2) with no running engine on water; Newsletter 10 (1994) says the buggy was being retrofitted "to match the acceleration performance of gasoline-driven cars", which implies it did not yet. No dyno, fuel-consumption or gas-supply measurement exists. The 1993 IECEC review (Bailey and Grotz) repeats Meyer's own figures ("over $1.6 million" spent) [web via Investigator, snippet]. (d) If an engine ran on net water energy, a perpetual-motion-class violation of the first law, i.e. the central principle of thermodynamics, would have to be amended. Verdict: Refuted by calculation for water-as-net-fuel, using Meyer's own flow and power numbers. Supplementary hydrogen from an on-board battery-powered electrolyser to improve combustion is known (Plausible, untested here), but it gives no free energy. Meyer's later US 5,293,857 (the US member of the WO 92/08046 family; downloaded by the Investigator) is a fuel-management system; I have not studied its drawings, and it contains no data to change this verdict [web: https://patents.google.com/patent/US5293857A/en, existence only].

6. Claim 4: Electrical Particle Generator (EPG)

(a) CA 1,213,671 / EP 0 098 897 / US 4,613,304 describe a loop of non-magnetic tube containing magnetized gas particles accelerated by a primary coil and picked up in a secondary with more turns. The patents give no voltages, currents, masses or velocities [read: deep read, CA1213671 short answer]. Mathematically there is nothing to check. A back-of-envelope: induced EMF is the rate of change of flux. A gas of "magnetized particles" has a negligible magnetic moment per unit volume compared with iron; at a gas density of about 1 kg/m^3 even if every molecule were fully magnetized its moment density would be of order 10^3 A/m at most; the flux linkage is tiny. The power the secondary can deliver is at most the kinetic power imparted by the accelerator coil (energy conservation), so "voltage amplification" is current-for-voltage trading at best, like a transformer. The deep read of US 4,613,304 correctly notes that it is a combined cell-and-accelerator patent without pulsed resonance; US 4,613,779 is a rotating generator (not a cell circuit), a Step 1 mislabel avoided by the fact-finder. Later newsletters' "gas battery" "producing electric energy directly from fuel cell gases without chemical interaction" is stated without any description of a mechanism [read: claims.md]. (b) Faraday induction and transformer action (established) account for any "voltage amplification" without power amplification. (c) No test. (d) An output exceeding input would violate energy conservation; no known mechanism exists for it. Verdict: Split. Voltage step-up as a transformer-like effect: Established. Net power gain: Not testable as stated (no numbers), and would be refuted by energy conservation if asserted.

7. Claim 5: Resonance in the electrode gap (1982 form)

(a) "the distance between plates ... will have ... a wavelength ... related to the motion of the water molecule in traveling from the one plate to the other" with velocity "increasing geometrically" [read: EP 0 103 656 pp.9-10]. No equation, wave speed, frequency or data exist [read: deep read]. Water molecules in a liquid have a mean free path of about 0.1 nm and cannot travel 3 mm ballistically. The only moving species in a field are ions: the H3O+ mobility is about 3.6 x 10^-7 m^2/(V s) [memory, high], so in 2 x 10^5 V/m (650 V over 3.2 mm) the drift speed is 0.07 m/s and the transit time across 3.2 mm is about 45 ms, i.e. a "cavity frequency" of tens of Hz, not kHz [calc]. Since steady drift is viscous (no inertia), no growth of speed "geometrically" can occur. Meyer's later patents (US 4,936,961; WO 92/07861) quietly replace this mechanical resonance by an electrical LC resonance, so the idea changed. (b) Ion transport in electrolytes is well characterized. (c) No test. The 1982 patent is a pre-electrical-resonance form. (d) If true, molecular-scale transport in liquids would need new physics. Verdict: Not testable as stated. The partial calculation above doesn't match it. This is a historical stage of Meyer's thinking, superseded by his own later electrical-resonance description.

8. Claim 6: Puharich's resonant splitting

[read: deep read of US 4,394,230.] Claims: carrier 200 Hz-100 kHz with 20-200 Hz amplitude modulation, "resonant frequencies of tetrahedral water molecules", 80-100% efficiency, "114.9%" exergonic/endergonic, 100 mW cell. (a) Section 7 of the deep read already finds the arithmetic divided by 22,400 cc rather than 33,600 cc per mole of water. I agree: the correct gas volume per mole of liquid water is 1.5 mol x 22.4 L = 33.6 L at STP, so the measured 16.2 cc is 63% of ideal, not 91% [calc by deep reader; I accept it]. The 114.9% is the ratio of combustion enthalpy (286 kJ/mol) to the free energy of splitting (237 kJ/mol) = 1.21, i.e. an accounting ratio of a kind that does not by itself violate energy conservation (thermoneutral electrolysis can reach about 120% efficiency of the free-energy figure when the cell draws heat from the surroundings [memory, high]). Reading it as "over unity" is the reader's trap. The parameters stated in different places differ (resonance 600, 3,980, 1,272-6,474 Hz), and the power is measured as rms V x rms I in a reactive load with no phase factor, even though the patent says the reactance is nearly 180 degrees out of phase. (b) The bond angle of liquid water is not a resonance parameter; the "tetrahedral" angle 109.5 vs 104.5 degrees is a gas-phase molecular property [memory, high] and its use as a driving resonance has no support in standard spectroscopy (water's molecular modes lie in the THz-infrared range, far above 100 kHz). (c) The GC/MS analyses were by "two laboratories" not named; no replication is documented. (d) If true, a resonance at audio frequencies would imply water molecule modes at kHz. Verdict: Never properly tested (unreplicated; the patent's own bookkeeping errors reduce its force). Relation to Meyer: his US 4,798,661 cites US 4,394,230 [read: deep read] and he adopted pulsed-voltage resonance; both inventors claim "resonance", but with different frequencies, mechanisms (molecular vs LC) and waveforms.

9. Meyer's claims of confirmation and finance (not a numbered claim)

• "Laboratories around the world ... confirming them" (1994) [read: Newsletter 10 p.1]. What I can verify: Nagypal's letter (a 20 mA, 600 V plate cell, 0.5-1 cc/min; he distributes interest); Perkins's bench observation (<1 mA; no engine); Sussex University "breadboard rig" producing hydrogen (reported by Perkins; report not found); a Queen Mary and Westfield DTI report that could not reconcile the claims with science but "left open the possibility"; NIST's "net consumer of energy". None of these is a measurement of gas volume per unit energy that beats Faraday. The negative government evaluation was not mentioned in the newsletters I read; I note that omission as a fact, not as proof of intent.
• Sales: dealerships and profit-sharing certificates were sold on these statements; the court dispute (1993-96) was about the dealers' purchases [web, stanslegacy; Meyer's own account]. The deep reads show Meyer's costs figures changed from issue to issue.
• Meyer's death on 20 or 21 March 1998 (coroner: cerebral aneurysm) and the assertions of poisoning appear in family retellings reported by the Dispatch [web via Investigator]; this is outside what a physicist can judge. The documented facts: estate inventory about $306,608, patents assigned to the widow (22 Mar 2001), fees lapsed 2004-2006, no licensee found: the technology went quiet. If the cell produced net energy, it is hard to see why no replication in 28 years has put it on the market, while the simplest competing explanation (no net energy) needs no secrecy. I note that this is an inference, and that suppression cannot be excluded without the missing records; it is not a test.

10. The Investigator's leads: what I keep and how I weigh them

Kept in the record whatever the verdict: Electronics World eyewitness account (sec. 3c); Perkins observation and the Perkins reservation note; Nagypal letter with numbers; Sussex breadboard report (unseen); Queen Mary/Westfield DTI report (unseen); Hans Nieper letter 14 Dec 1996 ("the tested overunity effect"; interested party, unseen) and Griffin's Warsash lecture (Sep 1993: 10 W in, gas at 10 psi within 10 s, cold vessel; Meyer's binder, unseen by me); NIST ERIP evaluation (second-hand); Lewett letter 28 Jun 1991; Kucherov US 5,632,870 (cites Meyer); Lawton builds; the Puteanus 2025 paper (mainstream context); the Fayette County docket; the probate record and patents' lapse; the Columbus Dispatch 2007 witnesses; the missing police tapes of the two Belgian investors; Laughton's role (favourable observer 1990, named as the expert in the Sunday Times of 1 Dec 1996 but absent from the docket transcription: an unresolved discrepancy). Of the Investigator's items, the ones that bear most on physics are the NIST evaluation (unobtained, highest value), Nagypal's numbers, and Meyer's own energy table and Appendix A, which let me test claims 1 and 3 on his own numbers.

11. What would settle it

A calibrated test: a cell and drive from Meyer's patents on a bench, with a true-RMS power meter (V x I integrated over time, with a fast scope on the cell terminals), a gas-volume rate meter (gas burette), the same cell run as DC electrolysis for comparison, and an ice-calorimeter or flow calorimeter for heat. If gas energy (286 kJ/mol H2 x measured mol) exceeds measured input, the claim stands; if not, it fails. The test is cheap and has been possible since 1990. The NIST file and the court transcript would show whether it was done.

Corrections to Step 1

1. Birth Memo 424 / "unread" (claims.md "requested deep reads"). I read Birth pp.117-123 and scanned Appendix A. Memo 424 gives no derivation or calculation of "2.5 million barrels"; it is a qualitative text on the "Energy Aperture". The deep read of claim 2 can therefore say that the figure is asserted without a calculation. The Opus deep read of pp.57-222 is still useful for Memos 425-430 (not read in detail by me).
2. claims.md says "No gas rate, current, input power or energy balance is given anywhere" (claim 1). Wrong for the wider record: Meyer's 1995 binder (pp.59-60, via the Investigator) gives 12.5 V at 40 A, 46,666 cc/h per tube and a "Differential Efficiency Factor" of 1,696. It also says "No engine test data" (claim 3), while Appendix A gives engine flow arithmetic (44.4 ml/min, 7.4 µl/cycle, 40 kV at 1 mA). These are calculations, not measurements, but the energy-balance statement should be corrected.
3. Water's dielectric constant in Electronics World. The magazine says pure water has "a dielectric constant of about 5". That is wrong (about 78-80); the Investigator marked it for checking. Meyer's own 78.54 at 25 C is correct; his 78.54 at 20 C (Birth p.52) is not (about 80.1 [memory, high]).
4. Investigator correction 3 (Nagypal context). Confirmed. In addition, 0.5-1 cc/min is far below any practical rate, and "300% effectivity" is relative to a 10-12 W reference; the actual energy ratio to the thermodynamic minimum is about 3% (sec. 3c).
5. Claims.md claim 3 text "droplets of 7.4 to 148 µl per cycle for engines from 50 to 1,000 hp". Appendix A scales the droplet by "148 µl / 7.4 µl = 20" times 50 hp = 1,000 hp. The droplet scaling is linear, so the same energy shortfall (factor of about 10) applies at every size.
6. Deep read (Birth) note on "Eq 6 and 7". The deep reader's check is right; in addition, V_L and V_C "theoretically infinite" is only for R = 0.
7. Step 1 says the Fayette case is dealers' fraud suit (Investigator correction 1) and "$25,000". Agree with the Investigator; I use $40,515.72 garnished (docket) and note the $25,000 figure comes from the Sunday Times retelling.
8. Death date. Dispatch 20 March vs probate 21 March 1998: both stand (Investigator correction 2).
9. Unverified transcriptions. Several of my numbers (sec. 3a points 1-3, Nagypal, Perkins) come from stanslegacy.com's machine OCR of Meyer's binder ("Unverified transcription"); I did not view the original scans. The "7 lbs/min" reading should be checked against the scan image before the psi interpretation is cited.

Web checks done (budget 6)

1. USPTO MPEP 2107 (utility, inoperability, perpetual motion): confirmed, https://www.uspto.gov/web/offices/pac/mpep/s2107.html.
2. NIST WebBook water enthalpy: vapour value confirmed (-241.8 kJ/mol); liquid value (-285.8) retained as [memory, high] since the page did not show it.
3. Puteanus 2025 abstract read from the saved PDF. Not done: Shimizu pulse papers, NIST ERIP file (not online as far as the Investigator found), H3O+ mobility (memory, high), O-H bond field estimate (memory, medium; order of magnitude only).