Mathematician review: Linerman-Peter (Peter A. Lindemann, with Edwin V. Gray's patents)

Reviewer role: PhD mathematician (Step 2). I check derivations, units, notation and hidden assumptions, and do the energy arithmetic that the claims turn on. Tags: [read] from the project documents (p.N = page of the compilation), [web] with URL, [memory] with confidence, [calc] my own arithmetic (inputs stated). Source stake: Lindemann is D.Sc. (honoris causa), sells the book and runs free-energy.cc (Clear Tech, Inc.); Gray's patent text is the inventor's own, attorney-drafted description. Neither document is a test report [read: claims.md].

Reader note: every symbol is defined where first used. E = energy (joule, J), P = power (watt, W = J/s), V = volt, C = capacitance (farad, F; uF = 10^-6 F), L = inductance (henry, H), COP = coefficient of performance (useful output divided by input).

What the "mathematics" of this project actually is

There are almost no equations in either work. Lindemann's argument is verbal ("the Electro-Radiant Event", "Light Ether", "propagates instantaneously") [read: pp.43-44, 60-62 of the PDF]. The only places where numbers can be checked are (i) the geometry of Gray's motor patent, (ii) a handful of stated power ratios (100:1, 1000:1, 16:1, 17:1, 7,460 W vs 26.8 W), and (iii) the energy bookkeeping implied by the circuits. So this review does three things: verifies the geometry, tests each stated ratio against energy conservation, and says which statements are too vague to have a mathematical content.

Corrections to Step 1

1. Confirmed: the "134 degree" discrepancy is an OCR misreading of 13 1/3 degrees. I re-derived the geometry independently. Stator pairs every 40 degrees, rotor and stator offsets of 13 1/3 degrees, give 40/3 per step, 27 steps per revolution, 27 x 13 1/3 = 360 [calc]. I checked Table I's pattern with a short script: step n has (rotor, stator) = (minor, minor), (major, major), (minor, major) for n = 1, 2, 0 mod 3; arm offsets +9 and +18 steps (= 120 and 240 degrees) reproduce every row the deep read transcribed (rows 1, 10, 19, 25, 26, 27) [calc; read: deep/free-energy-secrets, section 4].
2. Confirmed: the patent claims only partial recovery and needs a continuing supply ("a part of the energy consumed"; claim 1; "as long as d-c power is applied") [read: deep read sections 1, 8]. claims.md's "narrower point" is correct and important: the press and Lindemann claim more than the patent.
3. Added to the deep read's list of patent defects: the OCR-confusion aside, the patent's own "450 degrees" (for 240) and the FIG.7 vs Table I firing-pair mismatch stand. A further quantitative gap: the patent gives only voltage ratings (12 V batteries, 18 capacitors of 3,500 V), no capacitance, so energy per discharge cannot be computed from the patent; see claim 1 for what the press figures would require.
4. Correction to claim 5 and facts.json: facts.json gives Gray's years as 1925-1989 (Lindemann's date). The Investigator's government-derived records (VA-created Find a Grave memorial; FamilySearch tree) give 8 Jan 1925 - 20 April 1990, Sparks NV [web: https://www.findagrave.com/memorial/3813396/edwin-vincent-gray; https://ancestors.familysearch.org/en/LY5P-7C9/edwin-vincent-gray-1925-1990]. A third advocate account (KeelyNet 2000) says April 1993 [web: https://frienergi.alternativkanalen.com/Graymotor.htm]. Three different years; I keep all three in the record.
5. Correction to claims.md claim 3 wording: "Seven families ... proven beyond all reasonable doubt" is Lindemann's phrase [read: ebook p.1]. The same essay's own framing is "a small amount of one form of energy to control or release a large amount of a different kind" [read: ebook p.1], which is a weaker statement than energy from nothing. Several of the numbers he gives (below) go further than that framing.
6. Correction to the Investigator's lead on the "Crosby" test (not an error in sites.md, an addition): the Tattler's figures are internally inconsistent in a way that matters; see claim 2 section "The Crosby/Maize numbers".
7. Step 1 gap, now filled by me: the Gray circuit chapter (pp.56-60 of the PDF) rested on digests only. I read it in the extracted text. It contains one derivation-like passage that contains an error of reasoning (the "99% lost in the voltage drop" argument). See claim 1.

---

Claim 1. Gray's "cold electricity" is Tesla's radiant energy; the tube circuit discharges a capacitor through a spark gap; Gray's own explanation of the gain is wrong

a. Is the mathematics internally consistent? Split into three parts.

(i) The motor patent (US 3,890,548). Geometry consistent (see Corrections 1 and 3). Energy statements are qualitative only. The one hidden-assumption point worth a reader's attention is torque at alignment. The coils are fired "at the instant of their alignment" [read: p.79]. Two like-pole electromagnets facing each other repel with an interaction energy U(theta) that is largest at exact alignment, so the tangential force F = -(1/r) dU/dtheta is zero there and nonzero only after the rotor has moved past. The patent handles this with +/- 6 2/3 degree "faster/slower" offsets [read: p.76], which is half a step, so the design is not self-contradictory, but the base 13 1/3 degree firing points as described give zero torque from the fired pair itself. Whether torque then comes from unfired neighbouring iron is not stated.

(ii) Lindemann's circuit "derivation" [read: pp.56-59]. His chain: battery raised to about 3,000 V, stored in a large capacitor, discharged through a spark gap clipped by a vacuum tube to under 50 microseconds, radiated as "Electro-Radiant" impulses to copper grids, which then discharge through an inductive load. At two points the argument fails as mathematics:

• "99% of the work potential of the energy is lost in the voltage drop" of the spark gap [read: p.59-60]. For a capacitor C charged to V0 discharging through a gap with an arc voltage drop Va (roughly constant once the arc forms), the energy dissipated in the gap is Va x Q, where Q = C V0 is the charge that passes. The fraction of the capacitor's energy (1/2) C V0^2 dissipated in the gap is therefore 2 Va / V0. With Va of 20 to 100 V (assumed typical arc drops, [memory] medium confidence) and V0 = 3,000 V, that is 1.3% to 6.7% [calc]. So a small gap voltage drop means a small loss in the gap, the opposite of "99% lost". The rest of the energy must go somewhere else (resistor #30, wire resistance, inductance, radiation, the load). His conclusion that the capacitor's energy is spent in the primary discharge may still hold, but the stated reason is a unit-confusion between voltage drop and energy loss.
• "This work output is 100 times greater than the energy lost to create it" [read: p.60]. This is an energy ratio and needs two measured energies. None is given. It is asserted, not derived.

Lindemann's rebuttal of Gray's patent explanation is sound arithmetic as far as it goes. Gray argues that "instantaneous peak apparent power" is very high because instantaneous V and I are both high. Peak power is not energy. Illustration with assumed values (not from the patent): C = 2.5 uF, V0 = 3,500 V gives E = (1/2) C V0^2 = 15.3 J. Into an assumed 10 mH coil the half-sine current peaks at V0 sqrt(C/L) = 55 A and the quarter period is (pi/2) sqrt(LC) = 248 microseconds, so peak power is about 190 kW, yet the energy is still 15.3 J [calc]. "Very high instantaneous power" and "small fraction of a millisecond" (Chalfin's phrase [web: rexresearch]) are both true of a 15 J pulse and imply no gain. Lindemann is right that you cannot multiply the voltage of one source by the current of another to get power [read: p.59]; the product V x I has the units of power only for the same circuit element at the same instant.

(iii) The "Electro-Radiant Event" list [read: pp.43-44]. Items such as "thousands of times higher voltage", "propagates instantaneously as a longitudinal electrostatic light-like ray", "behaves like an incompressible gas", "penetrates all materials" are verbal and carry no equation, no units and no parameter that could be measured. "Incompressible gas" is not a mathematical object. "Instantaneously" is testable but see claim 2.

b. Consistent with established experiment? The parts that are ordinary engineering are consistent: capacitor discharge across a spark gap, back e.m.f. (the voltage L di/dt developed when current in a coil is interrupted), and recharging a battery from an inductive kick are all standard. Lindemann's own build of Gray's motor circuit "did not produce cold electricity" and gave "dangerous discharges" [read: digest p.14-16; facts.json]. The 12 V / 3,500 V figures with no capacitance mean the patent cannot be compared with any measured result.

c. Tested? By whom? Replicated? - Gray's 1973 demonstrations were seen by a Tattler reporter and physicists named only as "two knowledgeable physicists" [read: p.9-10 of PDF]. No instruments are described. - Crosby Research Institute / J. A. Maize test (1973), as reported by the National Tattler, 1 July 1973 [web: https://www.rexresearch.com/evgray/1gray.htm]. The test was for Pan World Enterprises, an investor with a stake; the reporter championed Gray. No Crosby document has been found. - Donald Cronin, TRW staff scientist, saw a test about mid-1974: "It ran for about 10 seconds and then everything blew" (AP wire 7 July 1975 via rexresearch). This is documented in one published wire story; no second source found yet. - Anonymous KeelyNet witness (2000): a 12-hour battery run ending with more charge than at the start; no data [web: https://frienergi.alternativkanalen.com/Graymotor.htm]. - Modern forum replication attempts of the tube circuit exist (groups.io EVGRAY; Reddit) but were not visited by the Investigator; I make no statement about their results. - No controlled, instrumented, neutral test of the tube circuit or the motor is on record in anything I have seen. Absence of such a test is itself an anomaly for the reader to weigh: Gray had a funded company, a US patent, the attention of the press, and a Japanese backer, and the record of any bench measurement stops at a tabloid account.

d. If true, what would have to change? For the strong form (output energy exceeds input energy in a closed system) the principle of energy conservation, which follows from time-translation invariance of the laws (Noether's theorem, [memory] high confidence), would have to be dropped. For the weak form (partial recovery of inductive energy) nothing changes: that is ordinary circuit theory, and it is what the patent says.

Verdict (split): - Weak form (the patent's claim: a capacitor-discharge motor that returns part of the inductive energy to a second battery and needs a continuing supply): Plausible, untested as a motor of unstated efficiency. Nothing in the geometry is inconsistent; no power measurement is on record. - Press/Lindemann strong form (output exceeds the energy supplied): Never properly tested (no neutral controlled test on record; the one witnessed trial in the record, Cronin's, was a failure; dismissed by the DA/press context rather than tested by a neutral party), and Refuted by calculation in so far as it is stated as net output exceeding the energy delivered by the batteries in a closed system (see claim 2 arithmetic). - "Gray's circuit is Tesla's radiant energy, and Gray's own explanation is wrong": Not testable as stated (the Electro-Radiant Event has no measurable definition); Lindemann's narrower argument that Gray's V x I explanation is wrong is correct arithmetic.

---

Claim 2. Radiant energy is longitudinal, instantaneous and ether-borne, and does not obey Ohm's law, thermodynamics or Maxwell's equations, so output can far exceed the energy lost in the primary discharge (about 100:1 for Gray, 1000:1 for Methernitha Testatika)

a. Internal consistency. Longitudinal waves vs Maxwell. Lindemann writes that Maxwell "rejected" longitudinal electric waves as impossible [read: p.15 of PDF, Chapter 2]. In vacuum Maxwell's equations include div E = 0. For a plane wave E = E0 exp[i(k.r - w t)] this gives i k.E0 = 0, so E0 is perpendicular to the direction of travel k. Longitudinal propagation in free space is excluded by the equations themselves. Longitudinal oscillations do exist in plasmas and as non-radiating near fields (a standard result, [memory] high confidence), and nothing in this project is a derivation of a longitudinal wave in vacuum. Lindemann then says the effect "does not obey Maxwell's equations" [read: p.62] while building on Tesla's Maxwell-era vocabulary; that is a consistent position only as an assertion that the equations fail, which requires a measurement, not a definition.

Instantaneous propagation. "Propagates instantaneously ... to any distance" [read: pp.43, 62] means infinite speed. Maxwell's equations give a finite speed c = 1/sqrt(mu0 eps0) for all waves in vacuum; infinite speed is a different, testable statement (send a pulse over L km; the delay should be L/c, about 3.3 microseconds per km). Measurement of the delay is the whole test, and none is cited.

Ohm's law. Ohm's law (V = I R) is not a law of nature but an empirical description of ohmic materials; a spark gap is non-ohmic anyway. Saying a medium is "not responsive to Ohm's law" says nothing until a V-I relation is measured.

The "gain" ratios, as arithmetic. "100 times greater than the energy lost to create it" and "must be close to 1000 to 1" for the Testatika [read: p.60]: - The Testatika figure rests on a photo caption, "The claim in this image is 1,000 watts of light" [read: Fig. 36 and p.60]. A 1,000 W claim needs a source for the input power. A Wimshurst machine (an electrostatic charge-separating generator) delivers a current I; power is V x I. At an assumed 30 kV and a generous 100 microamp (typical order for hand-cranked machines, [memory] low-medium confidence) that is 3 W; 1,000 W at 30 kV would be 33 mA, over 300 times larger [calc]. The statement that "the machine is turning itself" [read: p.60] is a claim about energy input; no measurement of the motor-drive input is given. This is a ratio with no measured numerator or denominator. - "Less than 2,000 watts from the wall" for a Tesla coil producing "4 amperes to ground" [read: p.46 photo caption] is an input bound, not a gain. Lindemann's bulb-lighting demonstration (bulb held in two hands near the coil) is the standard near-field capacitive coupling; the bulb's power is below the supply's 2,000 W, so it is not an energy anomaly.

b. Consistency with established experiment. Everything checkable here is contradicted by or silent against established results: - Energy conservation has been verified in electrical machines to a precision of a fraction of a percent in routine calibrations for over a century ([memory] high confidence); no accepted measurement shows output above input. - "Impulses shorter than 100 microseconds are completely safe to handle and will not cause shock" [read: p.43]. This is a safety claim, not an energy claim. Nerve-stimulation thresholds follow a strength-duration curve, in which shorter pulses need higher current to stimulate; they do not become harmless at a sharp 100 microsecond edge, and high-voltage pulses can still injure by burns and arcs ([memory] medium confidence; not checked on the web). I flag it as a numerical claim with no source.

The Crosby/Maize numbers (a check of the single most specific test claim on record). As reported by the Tattler [web: https://www.rexresearch.com/evgray/1gray.htm]: 10 hp load, 1,100 rpm, "7,460 watts"; batteries "capable of 5,454 watts per hour"; "total battery power consumed is only 26.8 watts per hour"; run time 203 hours; "measured overall system efficiency exceeding 99%". Taking each at face value [calc]: - Units: "watts per hour" is not an energy or power unit. The reading that makes the arithmetic work is battery capacity 5,454 Wh and net draw 26.8 W, since 5,454 / 26.8 = 203.5 hours, matching the reported 203 h. - 10 hp = 7,457 W, matching 7,460 W. At 1,100 rpm that is a torque of 64.8 N m and 407 J of shaft work per revolution. - If output is 7,460 W and net battery consumption is 26.8 W, then output/input = 278. That is not "efficiency exceeding 99%". The 99% figure is only consistent with the Tattler's other statement of recovery: 26.8/7,460 = 0.36%, so "99.64% returned to the battery". Either reading fails energy conservation: if 99.64% of the energy is returned to the battery, the energy converted into shaft work (7.46 kW) cannot also be part of what is returned, so the net draw would be at least the shaft work, 7.46 kW, not 26.8 W. With a 5,454 Wh battery and 7.46 kW net draw the run lasts 0.73 hour, not 203 hours. - At 12 V, 7,460 W of gross draw is 622 A; 26.8 W is 2.2 A. The EMA-4 data sheet (12 V, 2,550 rpm, 32.05 brake hp, "Fuel consumption: none", [web: rexresearch, Probe the Unknown June 1973]) is 23.9 kW, which at 12 V would be about 1,990 A [calc]. These are the inventor's company's and a tabloid's figures, not a measured data set. The mismatch above does not show the test did not happen; it shows that the report as published cannot describe a measurement made on a closed energy ledger. The right measurement would be energy into the battery bank over a full discharge-recharge cycle, which is what the Maize report would have needed to give.

c. Tested? Replicated? Same record as claim 1. Methernitha/Testatika: the only evidence offered is that "5 or 6 working models" exist [read: ebook p.1]; no measurement is in the folder. Whether Testatika has ever been measured by an independent party I could not establish; I did not search for it and state nothing about it.

d. If true, what would change? Maxwell's equations would need a non-causal, infinite-speed longitudinal mode; special relativity would be violated by any signal faster than c; energy conservation would fail locally. Lindemann himself writes elsewhere that his over-unity cases are not violations of the First Law (see claim 4), which is in tension with this claim.

Verdict (split): - "Longitudinal / instantaneous / outside Maxwell" as physics statements: Not testable as stated in the vague parts ("light ether", "incompressible gas"); the instantaneity sub-claim is testable and contradicts the finite c of Maxwell's equations; it has Never properly tested status in this record (Tesla's own statements are quoted but no timing measurement). - "Output far exceeds the energy lost, 100:1 / 1000:1": Refuted by calculation for the numbers as stated in a closed system (the Crosby arithmetic above is internally inconsistent; the 1000 W / Wimshurst figure requires about 300 times the current such a machine delivers), with the caveat that no measurement from the inventor's side has been seen. Never properly tested in the sense that no neutral measurement exists on record.

---

Claim 3. Seven families of "free energy" devices are proven beyond reasonable doubt (no test report given)

Lindemann's "proven beyond all reasonable doubt" gives no tester, instrument or report [read: ebook deep read, section 7]. I check each stated number.

Worked example: the 350 hp solar figure [read: ebook p.2-3]. 100 sq ft = 9.29 m^2. Peak sunlight at the ground is about 1,000 W/m^2 (standard value, [memory] high confidence), so at most 9,290 W = 12.5 hp of sunlight falls on the collector. 350 hp = 261 kW, or 28 times the incident solar power at peak, even at 100% conversion [calc]. A heat engine between a hot collector at, say, 100 C (373 K) and the ground at 55 F = 286.5 K has a Carnot (ideal reversible) efficiency of 1 - 286.5/373 = 23%, so even an ideal engine delivers at most 2.2 kW = 2.9 hp from that collector [calc]. Lindemann's "17 times" divides 350 by the compressor's 20 hp and leaves the solar heat uncounted [read: ebook deep read, section 1.7]. Using the ground as extra heat source does not rescue it, because a system working from the ground alone is a single-reservoir engine, which the Second Law forbids. By contrast his claim that heat pumps reach COP 8 to 10 at small temperature lifts is plausible: the ideal heating COP is T_hot/(T_hot - T_cold), which is 15.6 for a 20 K lift at 313 K, and a real machine at half of ideal gives about 8 [calc]. That COP is for moving heat, not for producing 17 times more work than the compressor takes.

b. Consistency with experiment. The claims on Meyer, cold fusion and the MEG have been examined by others, as follows (all [web], from search-result summaries; I did not open the primary documents): - Meyer: an Ohio court in 1996 found "gross and egregious fraud" and ordered repayment to two investors; three expert witnesses reported the cell was conventional electrolysis; Meyer did not allow a car test by Prof. Michael Laughton [web: https://en.wikipedia.org/wiki/Water_fuel_cell]. Meyer is an inventor with a stake; the investors and court are the other side. Lindemann's text names "Meyers" and lists patent 4,936,961 without mentioning this case. - Cold fusion: the 2004 US Department of Energy review of 18 experts split about evenly on whether the evidence for excess power was compelling; two thirds found evidence for nuclear reactions unconvincing; the panel recommended against a major program, but identified areas for further study [web: https://www.scientificamerican.com/article/back-to-square-one/ and DOE report at https://www.newenergytimes.com/v2/government/DOE2004/DOE-CF-Final-120104.pdf (search summary)]. Lindemann's "debunked within 6 months" omits that split. - MEG: search summaries report that Naudin's replication gave output equal to input rather than COP above 1 [web: https://www.energeticambiente.it/forum/energie-controverse/free-energy-z-p-e/apparati-immobili/speciale-m-e-g/375-bearden-si-risveglia , second-hand summary; I did not read Naudin's own page]. Lindemann cites Naudin as having "duplicated" the device [read: ebook p.2]; "duplicated" and "verified over-unity" are different claims and the second-hand report points to the first only. The Naudin primary page (jnaudin.free.fr) should be read by a later step.

c. Tested? Mixed. Meyer: tested in court, not supported. Cold fusion: examined by a government panel with a split result and the claimed effect not repeatable by most reviewers' account [web above]. MEG: one replication reported, with output equal to input [web, second hand]. Others: no test in the record. Mechanical heaters of the Frenette or Perkins type are real fluid-friction heaters whose output equals the shaft work; the physics there is conventional.

d. If true: each strong form would break the First or Second Law; cold fusion (if nuclear) would need no change in thermodynamics but a new low-energy reaction mechanism that overcomes the Coulomb barrier (the electrostatic repulsion between nuclei), for which the DOE panel found most reviewers unconvinced [web above].

Verdict (split by family): Radiant, permanent-magnet over-unity, "ten times" heaters, and 350 hp solar: Refuted by calculation (as stated); Freedman alloy: Refuted by calculation; Meyer electrolysis: Tested, not supported (court-appointed experts, 1996); MEG: Never properly tested as a neutral controlled measurement (one second-hand replication report with COP 1); cold fusion: Never properly tested in the sense that no accepted, repeatable result exists and a panel split, which is also the sense in which it was not dismissed without testing; implosion/vortex: Not testable as stated; heat pumps at COP 8 to 10: Established (ordinary engineering) but not what the 17:1 claim says. The umbrella statement "proven beyond all reasonable doubt": Not supported by anything in the folder (no report, tester or measurement).

---

Claim 4. The First and Second Laws of Thermodynamics are not universal, so machines that draw heat or ether energy without matching input are possible (Tesla's self-acting engine, Constable's etheric projectors, DePalma's rotating-magnet generator)

a. Internal consistency. Lindemann's own text undercuts the claim. - He writes "The line of logic embodied in the 'Laws of Thermodynamics' is flawless" and that the problem is in "certain interpretations" [read: Borderlands Q3 1994, PDF p.144-145]. That concedes the mathematics and disputes the application. - For Constable's projectors, he says "There is nothing over-unity about this situation ... the energy tapped by the system is the atmospheric ether and all of the energy gain in the system occurs outside of the equipment" [read: same]. Then he gives a gain "on the order of 100,000 to 1" and says "While energy is probably not being created or destroyed in the universal context, it is apparently appearing and disappearing from the machine ... in the local space" [read: same]. These are two different claims: an open system with an external energy source (consistent with the First Law) and local non-conservation (not consistent). The text slides between them. - Arithmetic check of the 100,000:1 rain example [calc]: 1 million gallons is about 3,785 m^3 or 3.8 x 10^6 kg. Falling 100 m through a turbine it releases m g h = 1.03 MWh (the height is an assumption, he gives none). A 300 W motor running one hour uses 0.3 kWh, giving a ratio of about 3,400:1; 100,000:1 needs an input of about 1.03 MWh / 100,000 = 10 Wh, i.e. a 300 W motor running about 2 minutes (the text says the rain falls "for hours", so the ratio depends on the assumed head and volume). The ratio is arithmetically reachable but is meaningless as an efficiency, because the energy stored in the water vapour is the source: condensing 3.8 x 10^6 kg of vapour releases 2.4 GWh of latent heat [calc, with latent heat 2.26 MJ/kg, [memory] high confidence]. So the example is a trigger-and-reservoir claim (like cloud seeding), which is consistent with thermodynamics when the energy is supplied by the atmosphere, and the 100,000:1 number then says nothing about violating a law. Whether the projectors cause rain is a separate, experimental claim that I have no test report for (Constable's work is cited by Lindemann only; none of it is in the folder). - The local-conservation statement has a precise mathematical form. From Maxwell's equations, Poynting's theorem says d(u)/dt + div S = -J.E, where u is electromagnetic energy density, S is the energy flux (Poynting vector) and J.E is the rate of work on charges. Energy "appearing and disappearing locally" would require this identity to fail, which is a failure of Maxwell's equations, not of an interpretation. A flow of energy that is not tracked by S is possible only if the accounting is incomplete. Lindemann's DePalma example ("charges appearing at the periphery and disappearing at the center") is also the normal behaviour of a current in a homopolar (unipolar) generator, a conducting disc turning in a magnetic field, in which current flows radially through the disc and returns through external contacts; the back-torque is the Lorentz force on that current. His statement that "the generated current produces a motoring effect that opposes the input torque should not be interpreted as a vindication of the conservation rule, but as an admission that this is an inefficient way to generate electricity" [read: same] is a non sequitur: the torque times angular speed equals the electrical power plus losses, which is the conservation rule.

b. Consistency with experiment. The Second Law's statements (no net work from a single reservoir; no cyclic process converting heat entirely to work) are consistent with every measured engine and refrigerator ([memory] high confidence). The "Joule paddle wheel" and "Carnot" historical passages Lindemann gives [read: PDF pp.143-146] are fairly summarised in outline, but he moves from "was an interpretation made from experiments" to "need not hold", which does not follow.

c. Tested? Tesla's "self-acting engine" and Constable's projectors: no test report is in the folder. DePalma's N-machine: described, no data given; I did not search for the test history of the N-machine (Kincheloe/Stanford, if any) and make no statement about it. Absence of cited tests is the finding, not a negative result.

d. If true: the Second Law's statistical foundation (that for a macroscopic system the probability of heat flowing spontaneously from cold to hot is vanishingly small) and the First Law's origin in time-translation invariance would both fail; a reader could ask what mechanism, with what equation, replaces them. Lindemann offers none.

Verdict (split): - Strong form (the laws fail; local non-conservation): Internally inconsistent (the same text calls the logic of the laws "flawless" and says there is nothing over-unity in the main example, then asserts local non-conservation) and Refuted by calculation for the specific numbers where they can be tested (350 hp solar, above). - Weak form (open systems drawing on an ambient reservoir are possible, e.g. a heat pump or a rain-making trigger): Plausible, untested for the specific Constable claim (no test in the record), and Established as general thermodynamics, which it does not contradict.

---

Claim 5. Gray's patents and demonstrations (1973-76) were real but his venture was raided, prosecuted and ended; LA DA raid July 1974, charge dropped, SEC plea 1976, death in 1989 "under mysterious circumstances"

This is a historical claim; mathematics enters only in checking dates and counts.

a. Internal consistency. The dates are not self-consistent across sources, which matters for the reader: - Charges: Lindemann has one grand-theft charge, dropped, and two minor SEC violations in a 1976 plea [read: claims.md]. The AP wire of 7 July 1975 gives 12 counts filed May 1975: one conspiracy, eight grand theft, three securities violations; about 800 shareholders from 1971; the DA said it was "hard to find investors willing to press charges" [web: https://www.rexresearch.com/evgray/1gray.htm]. The AP source is a mainstream wire report of the prosecution's position; the DA's office is the prosecutor and has a stake. 1 + 8 + 3 = 12, so the AP's own counts add up. Lindemann's count (1 charge) and the AP count (12 counts) cannot both describe the same filing. The outcome (plea, dismissal) is not in the AP piece. - Death: Lindemann gives April 1989 (Sparks shop, "mysterious circumstances"); KeelyNet gives April 1993 in a Riverside trailer; the VA-based burial record and FamilySearch give 20 April 1990 at Sparks NV, Riverside National Cemetery plot 27-1479 [web, Find a Grave and FamilySearch links in Correction 4]. US 4,595,975 lapsed for non-payment of fees on 17 June 1990, two months after 20 April 1990 [web: https://patents.google.com/patent/US4595975A/en], which is consistent with the 1990 date. Month and Sparks/Riverside match parts of both advocate accounts; the year matches neither. - The US 4,661,747 priority date of 1983-12-16 (earlier than the 1984 application named in the book) [web: Google Patents list, per sites.md] and the unlisted 10 April 1980 "Regenerative Energy Recovery System" application [web: rexresearch] are items the book does not mention.

b. Consistent with documents? The raid and prosecution are independently confirmed by AP; the details differ. "Mysterious circumstances" is Lindemann's characterisation; the one account with detail (KeelyNet, anonymous informant: police gave three different stories, no autopsy record found) is documented in one published source and no second source has been found. I cannot verify it and I do not dismiss it.

c. Tested? Not a testable physical claim; records that would decide: LA Superior Court file for the 1975 case, Washoe County death certificate, USPTO assignment records, SEC news digests for 1975-76 (Investigator did not reach these).

d. If true: nothing in mainstream physics changes; it bears on the "absence of evidence" question. The record goes quiet where a funded company with a patent and a hardware prototype would be expected to continue. Competing explanations: commercial failure after prosecution (supported by the prosecution record and the lapse of the '975 patent); dispersal of hardware (KeelyNet: six motors survive, rights held by "Western States", an oil-drilling company, and Zetech earlier); suppression (Lindemann's and KeelyNet's assertion, with the FBI mentioned in KeelyNet; no document). The records named above would decide between them.

Verdict: Never properly tested (a historical claim, not an experimental one; the prosecution is partly confirmed by a mainstream source, and the "mysterious death" and suppression statements rest on single advocate sources with no document either way). Dates for Gray's death are in conflict across three sources.

---

Claim 6. Free energy is withheld from the public by four forces (wealthy families and central banks, governments, charlatans, the public), which are really one force of "unspiritually motivated" human behaviour; the public must build devices itself

a. Internal consistency. The text states "To say that there is and has been a 'conspiracy' ... only leads to a superficial understanding" and then lists "murder and attempted murder of the inventors ... arson" as weapons of the First Force, and "arson, theft of property during shipment, phone taps" for the Second [read: ebook deep read, section 3]. That is an internal tension between denying a conspiracy and describing coordinated violence. The text also says the Fourth Force "all four are just different aspects of the same process", which makes the claim unfalsifiable: any outcome (suppression, public indifference, fraud by inventors) is explained by the one force. His own chain of premises is: (1) free energy devices exist (unsupported, claims 1 to 4); (2) powerful parties would suppress them (assertion); (3) therefore absence from the market confirms (2). Step (3) is circular unless (1) is established independently.

b. Consistent with the record? The one documented example in the folder, Gray, has a prosecution record (AP 1975) that is also consistent with an ordinary securities-fraud investigation, and the same record shows 800 shareholders and DA investigators who "invested money of their own for a while, but later got it back". Whether the raid was suppression or enforcement cannot be decided from the folder. No victim, date, document or case is named for murder, arson or patent denial [read: deep read section 3, "Evidence offered: none"]. Patent secrecy orders on national-security grounds exist as a general practice [memory] medium-high confidence; I did not check on the web and the essay does not cite one for any of these inventors.

c. Tested? Not applicable in the experimental sense. Absence of cited evidence is the finding.

d. If true: nothing in physics changes; it would require documents of the kind listed under claim 5.

Verdict: Not testable as stated (an unfalsifiable framework with an internal tension between denying and describing conspiracy). The documented sub-case (Gray's prosecution) is addressed under claim 5.

---

Items from the Investigator that I keep in the record, whatever the verdict

• AP wire of 7 July 1975 (Progress Bulletin, Pomona): 12 counts, 800 shareholders, Cronin's TRW account ("ran about 10 seconds and then everything blew"), Japanese team left after the test. Source: Rex Research reprint; the DA's office has a prosecutorial stake; Cronin is a witness with no stated stake [web: https://www.rexresearch.com/evgray/1gray.htm].
• Crosby/Maize test report via the Tattler, with its mixed units and its contradictory "99%" and "26.8 W" figures (claim 2). Pan World Enterprises had an investor's stake; the Tattler championed Gray.
• Chalfin and Wester (Caltech per the Tattler) endorsements. Chalfin's quoted mechanism, "energy is used only during a small fraction of a millisecond; energy not used is returned to an accessory battery", is the partial-recovery mechanism of the patent, and is consistent with my 15 J / 248 microsecond example. It does not claim output above input.
• EMA-4 data sheet: 12 V, 2,550 rpm, 32.05 brake hp, "fuel consumption none" (inventor's company). 23.9 kW at 12 V implies about 1,990 A if drawn from the battery; no battery or current figure appears.
• Anonymous KeelyNet witness: 12-hour run with batteries ending higher (no instruments or data); six motors said to survive; rights said to be held by Western States. Single anonymous source, no second source found yet; not dismissed.
• Gray's foreign patent family (CA 1030600, JP S50-74712) and the unlisted 1980 application "Regenerative Energy Recovery System", whose abstract describes capacitor discharge through an inductive load with a diode sending the inductive current to a battery. That is the partial-recovery mechanism, which Lindemann sets aside for the "radiant" explanation. The 1980 application's text should be obtained: if it contains measured results it could change claim 1.
• Death records (VA/Find a Grave, FamilySearch: 20 April 1990) vs the advocate accounts (1989, 1993); '975 lapse 17 June 1990.
• Lindemann's background (D.Sc. honoris causa in complementary medicine; VP of Borderland Sciences 1987-99; Clear Tech co-founder) as interested-party context. His own build of Gray's motor circuit did not work as he hoped [read: digest p.14-16].
• Lindemann's explicit statement that "Gray's explanation of the gain mechanism cannot be true" [read: PDF p.59]: an inventor-adjacent author agreeing with the critics on that one point.

Summary of what holds up mathematically

• Solid: the patent geometry (40/13 1/3/27 numbers), the partial-recovery mechanism in the patent, and ordinary heat pump COP of 8 to 10 at small temperature lift.
• Wrong or unsupported: the "99% lost in the voltage drop" reasoning; every quoted gain ratio (none comes with a measured numerator and denominator; the ones that can be checked break conservation); the 350 hp figure (28 times the sunlight on the collector); the Tattler's mixed-up "26.8 watts per hour" and "99% efficiency".
• Missing: no capacitance, inductance, coil or core data in the patent; no measurement report for any device.
• Open: whether the 1980 regenerative application or any Crosby document contains measured data; whether independent measurement of Testatika or the N-machine exists (not searched).

Web checks done (budget 6; 3 used)

1. Meyer 1996 Ohio court [web: https://en.wikipedia.org/wiki/Water_fuel_cell]: confirmed by search summary.
2. DOE 2004 cold fusion review [web: Scientific American, DOE PDF]: confirmed by search summary; split on excess power.
3. Naudin MEG [web: energeticambiente.it]: second-hand summary only; primary page not read. Not checked on the web, marked [memory]: typical arc voltage drops (medium), Wimshurst machine currents (low-medium), nerve strength-duration behaviour (medium), patent secrecy orders (medium-high).