Tesla: Arbiter (Step 3)

Date 2026-10-08. Inputs read in full: claims.md, physicist.md, mathematician.md, contrarian.md, sites.md (log and "Leads for review"), deep/voltage-gain-a17484.md. Facts.json, the other deep reads and the research notes were used through the reviewers' quotations of them (they were not re-read page by page). Tags: [read] from a file in the folder or research/; [web] with URL; [memory] with confidence; [calc] my own arithmetic (Python, numpy; steam properties from the IAPWS-IF97 formulation through the iapws library, run in a scratch environment).

Source gap note

Closed in part: the Colorado Springs Notes (1978 ed.), the 1904 Electrical World article and the 1916 interview were obtained by the Investigator (research/files/), so Tesla's own record of the 3 July 1899 storm night is now read, not quoted. Still not seen by anyone: the Allis-Chalmers (Dahlstrand) report, Richards's test log, John Trump's 1943 report on the papers, Eyring Corp.'s raw coil traces, any record of "teleforce" apparatus, any output measurement of a "fuelless" or "radiant energy" device, Lowenstein's own account of "similar observations", a published field measurement from the Viziv/CPG tower, FBI Vault Part 03. Conclusions that depend on these rest on the one secondary quotation named in each case. Several reviewer conclusions (loss figures for the earth-ionosphere guide, Schumann Q, steam tables, ground conductivity) rest on [web] or [memory] numbers; each is tagged where used.

Where the three reviewers differ, and my recomputation

Corrected claim list (final verdicts, Step 2 scale)

Claim 1. Power and signals sent to any distance through the earth, by resonance, with little loss, by conduction not radiation

Agreement: all three accept that (i) VLF/ELF signalling over global range works, (ii) the earth-ionosphere cavity resonates (Schumann modes 7.83, 14.1, 20.3 Hz [web: https://en.wikipedia.org/wiki/Schumann_resonances, via the Contrarian]), (iii) Tesla's diary entry of 4 July 1899 (about the night of 3 July) is real, more cautious than his 1904 article ("difficult to believe ... though it may be possible"), and was never repeated, (iv) "almost without loss" has no derivation anywhere in the folder. Disagreement: the Mathematician rates the narrow observation "Plausible, untested", the Physicist and Contrarian "Never properly tested"; the Mathematician and Contrarian each add a caveat to "Refuted by calculation". I side with "Never properly tested" for the observation (the data, a relay playing with a threshold adjusted several times, cannot be read as a stationary-wave amplitude pattern, and alternatives are not excluded) and with "Refuted by calculation" for the antipodal power claim through the earth-ionosphere guide, because the numbers above (40-60 dB loss at VLF; 3.6 mW reaching a 1 km^2 collector from 1 MW at 10 Mm) do not depend on any particular transmitter: the loss is in the guide and the geometry.

Split verdict: - Signals over global range at low frequency from elevated resonant transmitters: Established (by guided radiation in the earth-ionosphere guide, not by "conduction"). Tesla's scheme is the ancestor, not the mechanism. - Stationary waves seen on the night of 3 July 1899, as Tesla read them: Never properly tested (unrepeated; not dismissed by experiment; Wardenclyffe, the large test, was never run; Lowenstein's second-observer account not found). - Power delivered "almost without loss" to any terrestrial distance through the earth or the earth-ionosphere guide: Refuted by calculation. - The earth as an ideal conductor of limited dimensions at 10^4-10^5 Hz: Tested, not supported (skin depth about 10 m in 100 ohm-m ground at 250 kHz [calc, Mathematician]; ground-wave and guide attenuation measurements). - Power through ionised air at 30,000-35,000 ft with 15-20 MV (US 645,576): Never properly tested. The Mathematician shows the patent supplies no energy budget (about 0.19 A at 20 MV for 5,000 hp; no figure for the ionised volume or its upkeep) but no proof of impossibility is on record; I do not upgrade it. - The 18 million (or "ten million") horsepower figures: Tesla's own 1916 testimony defines them as peak capacitor-discharge "activity"; an undamped wave "gives you only ... 1 kilowatt and nothing more" [read: research/2026-10-08-archive-org.md §3]. They are not net power. At 13.4 GW a 28 J store (0.0225 uF at 50 kV) lasts about 2 ns, and the Corum 646 J pulse about 48 ns [calc]; the energy per pulse in Tesla's figures is unresolved (partly settled in revision F, F5: Tesla's 1916 antenna figures give a 4.4-5.3 kJ store, about 0.3-0.4 microseconds at 18 million hp).

• Added by the QE revision (section B): the 26-mile/200-lamp test, Tesla's 1904-1927 "around the globe" statements, the January 1900 short-range lamps, Golka's replication, cavity Q up to 1000 and the planetary signals each have a verdict line there.

What would settle it: (1) repeat the 3 July observation with a recording, non-threshold receiver and logged lightning positions; modern sferic-network archives (1999-2025) can ask whether half-hour modulation of far-storm signals exists at all. (2) Field strength versus distance (10-1000 km) from a Tesla-type grounded quarter-wave resonator at 100 kHz-class frequency, against 2-3 dB/Mm. (3) The receiver frequency in the 4 July entry, to compute node spacing against storm speed (node spacing = v x 0.5 h; 20-60 km/h gives 10-30 km and f of 5-15 kHz if the wave runs at c [calc]).

Claim 2. The transverse ether is wrong; the effects are electrostatic "electric sound waves"; later "longitudinal/scalar"

Agreement is complete on the substance; the reviewers differ only in labels. Tesla's observation that air near a high-potential high-frequency terminal dissipates energy (corona, ion wind, dielectric heating) is real and mainstream. His replacement of the transverse picture is contradicted by Hertz (1887-88) and all radio since. He never used "longitudinal" in the pages read (deep reads); the later authors did. In Bass's version (in Valone) E is parallel to B and the Poynting vector is zero, so it cannot carry power; Valone's Nichelson chapter has a unit slip (19 MeV per second read as an energy; 3.1e-14 V per electron, not 19-190 MV [calc, Mathematician]). - Energy lost to air by electrostatic and corona processes at high-voltage terminals: Established; his sigma^2 x frequency law: Never properly tested (Tesla: quantitative determinations "as yet wanting"). - Transverse electromagnetic waves do not exist or cannot carry energy to practical distances: Tested, not supported (Hertz and all radio). - "Electric sound waves" in a particle medium as the replacement picture: Historically superseded but instructive (reasonable with an unproven ether in 1891-92). - Later "longitudinal/scalar wave" energy transport (Bass, Valone, Vassilatos): Internally inconsistent as a power carrier (zero energy flow; the Valone/Nichelson unit error); otherwise Not testable as stated. Longitudinal field components in near fields, waveguides, plasmas and Zenneck/Sommerfeld surface waves are mainstream and are not what these authors mean [memory, high].

What would settle it: a defined, measured longitudinal energy flow in vacuum with the Poynting vector computed; none is in the folder.

Claim 3. High-frequency resonant circuits; wireless lighting; single-wire power; bifilar coil; "voltage gain"

Agreement: the resonant rise V_C = Q V_in is textbook and is not a power gain; Nichelson's "voltage gain" X_L/R with R silently 1 ohm is a reactance in ohms (24,822 and 14,173 [calc]) and the printed 929% should be 945.6% (5.39/0.57); 0.18 and 0.97 are measured transmission ratios of a passive part, below 1. Disagreement: why the peaks sit where they do (see table). The Coil patent (US 512,340) claims built-in capacitance, not gain [read: deep]. Tesla's 250,000 figure is a ratio of stored energies for adjacent-turn voltages ((50/0.1)^2); the lumped model N^2/4 predicts a 21-fold frequency drop for 43 turns but 1.73 was measured, a gap of about 150 in capacitance ratio (3.0 observed against 462 predicted) [calc, Mathematician]. That discrepancy is a real, unexplained observation and stays in the record. - Wireless lighting, single-wire (capacitive-return) power, large resonant voltage rise, self-capacitance of a coil: Established. - Tesla's stated purpose, a coil whose own capacitance neutralises its self-induction at self-resonance: Plausible, untested (consistent with self-resonance; no neutral test of the stated advantage). - A gain or "voltage gain" greater than theory in the bifilar coil: Tested, not supported (his own measured values are below 1); Nichelson's 25,000 / 14,000 and the 9.46 ratio: Internally inconsistent as a gain. The 1990 measurement itself (one afternoon, interested author, 4-inch coils per the voltage-gain paper; the Fuelless paper says 3-inch, so one is a misprint or they are two measurements) is a single unreplicated record.

What would settle it: remeasure with calibrated ports, S11 and S21 and phase, state wire length, pitch and resistance, and compare the peak frequencies with half-wave resonance of the wire (the [calc] hypothesis above). Cost: an afternoon and a network analyzer.

Claim 4. Energy drawn from the ambient medium ("fuelless", radiant energy)

Agreement (revised, see QE revision A): Tesla's 1900 Century argument carries his own caveat ("continuous pumping", "no known perfect converter") but then claims a partial net gain from one ambient reservoir; Nichelson's gloss is closer to the text than this paragraph first said. A closed single-reservoir engine cannot give net work (Carnot case: sink at 77 K in 300 K gives eta = 0.743 and the pump-out cost 0.257/0.345 = 0.743, exactly cancelling [calc, Mathematician]). Ambient gradients (sunlight, wind, thermal, atmospheric electric) are real but small: a 1 km^2 collector at 100 m height gives about 2e-6 A at 13 kV, 0.026 W; the whole global circuit dissipates 250-450 MW [calc/web above]. The Contrarian's "Plausible, untested" for the US 685,957 collector and the Mathematician's "Refuted by calculation in magnitude" are the same finding in two phrasings: it works at watt-scale-per-km^2 power and does not work at useful power. - Work from existing ambient gradients: Established (at densities far below "fuelless generator" claims). - A machine yielding net work from equilibrium surroundings with no gradient (Nichelson's reading): Refuted by calculation (second law; energy conservation for the closed unipolar dynamo, Elsasser). - Tesla's own 1900 sink text, partial-gain step ("what is converted is clear gain"): Refuted by calculation (revised 2026-10-08, see QE revision A; the earlier "not refuted; consistent with the second law" was wrong). His caveats ("no known perfect converter", "cannot hope to produce a self-acting machine ... for a long time yet") stay beside it. - The charged-particle collector (US 685,957) delivering useful power: Refuted by calculation (magnitude); as a charging device at tiny power: Established in principle, no output measurement on record. - Tesla's statements of 1902 and 1931-33 that he had harnessed "cosmic energy" (including the Pierce-Arrow car story, which is not in the Fuelless paper): Never properly tested (no device, no description, no named witness in the folder). Tesla was raising money in 1904 and was financially dependent in 1931-33 [memory, medium], which is stated as his stake, not as a reason to discard what he said. - Gray, Moray, Baumann "radiant energy" devices: Never properly tested: reports come from promoters and tabloid sources (Vassilatos, Lindemann); no instrumented neutral test in the folder. Silence of the folder is not contradiction.

What would settle it: metered input and output, with energy accounting, of any one device.

Claim 5. The Tesla disc turbine

Agreement is wide. The turbine is a working boundary-layer machine; Tesla/Lowenstein 1910-11 data (102.8 bhp at 53.4 lb/bhp-h; 77 hp at 35 lb; Prony brake, condensate weighed, record by Prof. C. D. Richards [read: Hayes deep]) are real measurements by interested associates; the Allis-Chalmers trial exists only as O'Neill's quotation of Dahlstrand (38% "mechanical efficiency", disk distortion, gearing); independent small rigs give Armstrong 1952 14%, Rice 1965 and others 20-35%, Hoya and Guha 2009 140 W at 25,000 rpm, about 25% with nozzle losses dominant [read: research/files/hoya-guha-2009-tesla-turbine-test-rig.pdf]. My steam recomputation (table, and QE revision E) gives about 25-52% of ideal expansion for Tesla's own tests. "10 hp/lb" has three mass bases in the sources (about 3 hp/lb for the 1906 model, a slogan of 10-20, and 0.5 hp/lb for the 400 lb Waterside machine); the Hayes mass bases differ by a factor of 20. - The disc turbine works as a viable machine, simple, robust, reversible: Established. - 80-90% efficiency: Tested, not supported. Open engineering question kept in the record: Guha's group attributes much of the gap to the nozzle, so rotor-only efficiency is not shown to be low or high by the data cited. - 10 hp per pound: Never properly tested (mass basis undefined; claim by Tesla and Lowenstein only). - Superiority at large scale: tried by Allis-Chalmers and not adopted (reasons as quoted by O'Neill); primary report missing.

What would settle it: the Allis-Chalmers test log; a rotor-only torque measurement with the nozzle losses separated.

Claim 6. Teleforce beam and the fate of Tesla's papers

Agreement (revised, see QE revision C): no apparatus or test is on record; written descriptions existed in 1943 (Trump's abstracts of Exhibits F, H, J, Q) and were judged unworkable by Trump. Physicist and Mathematician compute that a conventional charged or neutral beam cannot reach 250 miles through air (column density 1.6e4 g/cm^2 against about 90 g/cm^2 nuclear interaction length, so about 170 interaction lengths [calc, Mathematician]; about 100 MW delivered to melt 100 kg of aluminium in a second [calc]); the Contrarian marks it "never tested because there was nothing to test". These are consistent: the description as given is too vague to be tested, and the one concrete form (known particles in open air at that range) fails by calculation. - The beam as Tesla described it to the press: Not testable as stated; any form using known charged or neutral particles through about 400 km of low-altitude air: Refuted by calculation. - Papers sealed by the Alien Property Custodian on 9 January 1943: documented [read: FBI files]. "The FBI seized the papers" is contradicted by the file; "The FBI had no involvement" is overstated (New York inquired in January 1943; Washington told it to approach the State's Attorney and Surrogate Court) [read: claims.md corrections]. A 1981 FBI letter says the papers were examined by OAP, OSRD, ONI and NRL and released to Sava Kosanovic [read]. - Soviet or US military use of the papers ("Project Nick", 1983 airtel "allegedly"): Never properly tested (one 1983 airtel as primary item; the Project Nick chain is popular and secondary). - Trump's 1943 conclusion ("speculative, philosophical, and somewhat promotional ... did not include new, sound, workable principles or methods") is the only evaluation by a scientist who saw the papers [web: https://www.pbs.org/tesla/ll/ll_mispapers.html, via Investigator]; Trump was a government evaluator judging wartime value, and the full report is not online. It is kept beside the other accounts.

Claim 7. Priority claims made for Tesla

• Rotating field, induction motor, polyphase system: Established (patents, practice).
• High-frequency lighting, the Tesla coil: Established.
• Radio: Tesla's patents on tuned, grounded, elevated-terminal systems predate Marconi's, and a 1943 Supreme Court decision invalidated a Marconi four-circuit claim as anticipated; the decision did not decide who invented radio, and whether the Court relied on Tesla (versus Lodge and Stone) is not confirmed by the sources I could open. As a flat statement "Tesla invented radio": Not testable as stated (depends on the definition); as a priority in tuned coupled circuits: documented in the patents.
• X-rays: Tesla's known shadowgraphs are dated 1896, after Röntgen's announcement; earlier Crookes-tube work (1894) and a March 1895 laboratory fire are reported; discovery before Röntgen: Never properly tested (records lost or not available), and not established.
• Cosmic rays, artificial radioactivity, atom smasher, electron microscope: Never properly tested (O'Neill concedes the records "are not available"; the 1891-92 lectures describe electrode "disintegration", moving phosphorescent spots and a "very special radiation" only). If a record exists it may be in the Tesla Museum archive (163,911 call numbers).
• Corum and Corum's reading of the 1899 signals as planetary (Jovian/Io): a hypothesis by interested parties built on the diary; untested, not rejected.

Absence of evidence: where the record goes dark

Treated as anomalies, per SOURCE HONESTY; none is a neutral gap.

1. Wardenclyffe: no test of the full plant, and the tower's end. The record is dense to 1904 (Rankine letter of 19 April 1904: "could now be finished in three to four months", $100,000 needed; Tesla's covering mention of a Kerr, Page and Cooper statement, which he says shows his patents "controlling", and his own valuation of them at "not less than $5,000,000" (the opinion's text is not in the scan; corrected in revision F) [read: research/files/tesla-papers-1904-rankine-letters_djvu.txt lines 307-311]; Morgan and Astor interests), then goes thin: foreclosure (Waldorf-Astoria/Boldt, action 1915, judgment 20 April 1922) and demolition in July 1917. Competing explanations: (a) money ran out, no experiment ended it (supported by the Rankine letter and the foreclosure; against: nothing explains why no investor or later owner ran the test); (b) government wartime precaution (The Sun, 19 July 1917 says the Federal Government dynamited it "as a precaution"; the 5 August feature says the buyer is not known [read: research/2026-10-08-chronicling-america.md]); (c) commercial scrap or development after foreclosure (Anderson 1992; Wikipedia calls (b) a rumour [memory/web via Investigator]). (b) and (c) are not exclusive. Deciding records: the foreclosure-appeal transcript and Exhibit B inventory (Appendix III of the 1916 interview book; read in revision F, F6, which adds readings for the missing special apparatus), salvage records, 1917 Navy radio-station orders, NARA. The Contrarian finds no evidence that a failed experiment or suppressed data ended the programme; I agree that none is on record, and I add that the absence of any record of a completed large transmission test is itself what the claim lacks.
2. The 1943 papers. Sealed 9 January 1943; examined; released to Kosanovic (1952, per a 1981 FBI letter, which is the only statement of release in the file). Readings: (a) little of value, shipped to Belgrade (Trump's conclusion, the 1981 letter, the Museum's holdings); (b) some material copied for military interest ("Project Nick", Wright Field, Brig. Gen. Craigie; popular chain PBS 2000, probably Cheney 1981; 1983 airtel "allegedly"); (c) material removed or lost between sealing and release (no document says so, none excludes it). For (a): consistent, but the evaluator had a stake. Against (a)/for (b): an agency asked about the papers in 1983. Against (b): no primary record of Project Nick. Silence of online Air Force records is not contradiction. Deciding records: NARA RG 131 (OAP Tesla file, Trump report), RG 342 (Wright Field), an inventory of the Museum's archive against the 1943 inventory, FBI Vault Part 03 (not compared; the Vault has about 354 page objects against 265 pages in the folder).
3. Viziv/CPG tower. More than 100 patents citing US 787,412 (431 citing patents in all [read: research/2026-10-08-google-patents.md]) and a built tower, but no published field-strength measurement found. Readings: pre-licensing secrecy; negative or unimpressive results; published where the Investigator did not look (IEEE conferences, FCC experimental-licence filings). Interested parties (see the table). A single field-strength-versus-distance table would decide it.
4. Colorado Springs raw data. The diary has no trace record, no distance log and no sferic frequency; Lowenstein's "similar observations" have no account. This is the evidence the antipodal claim rests on. Readings: records lost (Tesla's papers went through the 1943 seizure and the Museum), never made, or not yet found in the Museum (the Notes were published in 1978 from the Museum's holdings). Deciding record: the Museum's original charts and call numbers.
5. Allis-Chalmers (Dahlstrand) report and Richards's log. Only O'Neill's quotation. Readings: report exists in the company archive (Wisconsin/Milwaukee); the firm dropped the product for engineering reasons; or the quotation was selective. Deciding record: the Wisconsin Historical Society holdings; the Investigator did not reach them.
6. Gray, Moray, Baumann devices. Promoters' reports; no instrumented test. Readings: no effect beyond an unmetered input; a real effect never measured neutrally; devices suppressed. Evidence for the first: Tesla's own 1916 definition of "activity" and the thermodynamic calculations; for the third: assertions by advocates, no document. Deciding record: any metered test of one device.

Corrections to earlier documents

• physicist.md, Claim 1(a), "A receiver collecting from 1 km^2 would get about 4 W of 1,000,000 W": wrong by a factor of about 1000; correct is 3.6 mW after 20 dB (0.36 W with no attenuation) [calc]. The conclusion is unchanged and stronger.
• contrarian.md, Claim 5(a), "176 psi and 21 inches of vacuum, a drop of roughly 280 Btu/lb ... so 35 lb implies about 26 percent": IAPWS-IF97 gives 252 Btu/lb (ideal 10.1 lb), 29%. Mathematician and Physicist steam figures confirmed within 2 points.
• mathematician.md and contrarian.md, Claim 7: "the Court noted Tesla's 645,576 disclosed the four-circuit system": not confirmed; summaries state the anticipation rested on Lodge and Stone and that the Court did not decide who invented radio. The wording should be "Tesla's patents are part of the prior-art argument; the sources I could open say Lodge and Stone were decisive".
• physicist.md, Claim 7: "X-ray images c.1895": the known shadowgraphs are 1896, after Röntgen's announcement; pre-Röntgen work is reported only as Crookes-tube experiments (1894) interrupted by the 13 March 1895 fire.
• physicist.md, mathematician.md, contrarian.md: the Corum stake was [memory, medium]; now supported by web (founders of Viziv, co-inventors for CPG), with the caveat that "Tesla-based" is the firms' own framing. The tower is placed in Milford (Hackaday) and Waxahachie (Dealroom) in different sources.
• physicist.md, Claim 3: the S21 peak is attributed to a parallel resonance of a series element; a parallel resonance gives a notch. The Contrarian's check (0.007 for a series inductor) is correct and the lumped-C values in the Mathematician's review (0.34 pF, 1.02 pF) depend on a model the data do not support.
• Claim list from claims.md carried forward: claim 1 should be split into 1a (stationary waves, 3 July) and 1b (antipodal and "almost without loss" generalisation; first printed in 1904, and in US 787,412 filed 16 May 1900); the observation is under the heading "Colorado Springs, July 4, 1899: Observations made last night", while the 3 July entry holds circuit data (Physicist, Contrarian). The Source gap note in claims.md is out of date as described above. claims.md claim 5 should name which efficiency definition each figure uses and the three mass bases of "10 hp/lb" (Mathematician). Apply these in primer and entry.
• Unresolved discrepancy kept open: Nichelson's coil is 4-inch (voltage-gain.pdf, Thermodynamics2) or 3-inch (Fuelless footnote 15). Do not merge.

Open questions

1. What did the 3 July 1899 receiver respond to every half hour: a spatial standing wave, ground-wave/sky-wave or mode fading, successive storm cells, sunset changes in the D layer, or Tesla's re-adjusted relay threshold? Raw charts, if any survive, and sferic-network climatology would discriminate.
2. Do the 11 and 19 MHz peaks in the 1990 coil measurement correspond to half-wave resonance of the wire, and why does the lumped inter-turn model overstate capacitance 150-fold?
3. What did the 1943 papers contain, and who saw what between 1943 and 1952?
4. Why did the Viziv/CPG tower publish no field data?
5. What did Lowenstein observe, and what records of Allis-Chalmers, Richards and Wardenclyffe survive?
6. Whether and how the 1943 Supreme Court opinion used Tesla's patents (read the opinion text).
7. What are the pulse energy and duration behind the "10 million" and "18 million" horsepower "activity" figures?

Summary of final verdicts

1. Global signalling by resonant elevated transmitters Established (as guided radiation); 3 July 1899 stationary waves Never properly tested; "almost without loss" power Refuted by calculation; earth as ideal conductor Tested, not supported; ionised-air conduction Never properly tested.
2. Air dissipation Established; transverse ether wrong Tested, not supported; "electric sound waves" Historically superseded but instructive; scalar/longitudinal power Internally inconsistent / Not testable as stated.
3. Resonant lighting and voltage rise Established; bifilar purpose Plausible, untested; claimed gain Tested, not supported; Nichelson's gain comparison Internally inconsistent.
4. Ambient gradients Established but tiny; equilibrium generator Refuted by calculation; Tesla's 1900 partial-gain step Refuted by calculation (caveats kept); Tesla's 1902/1931-33 device and Gray-type devices Never properly tested; collector at useful power Refuted by calculation (magnitude).
5. Disc turbine Established; 80-90% Tested, not supported; 10 hp/lb Never properly tested.
6. Teleforce Not testable as stated (known-particle form Refuted by calculation); papers fate documented in part, unsettled; Soviet/US use Never properly tested.
7. Polyphase and coil Established; "invented radio" Not testable as stated; pre-Röntgen X-rays and cosmic-ray, radioactivity, atom-smasher, electron-microscope priorities Never properly tested.

Web checks used 6 of 8 (global circuit; Marconi case; Corum/Viziv; Tesla X-ray; plus two failed fetches of the opinion text). Steam properties computed with IAPWS-IF97 rather than looked up.

QE revision (2026-10-08, after qe.md, verdict MINOR, and the QE deep reads)

Inputs read for this revision: qe.md in full; the new deep-read sections (Prodigal Genius pp.153-185 and 209-221; Tesla Said planetary and globe claims; Complete Patents pp.404-448; Valone: Van Voorhies, Grotz, Flowers, Corum ch.12, Rauscher and Van Bise, Tutt; FBI 01 pp.70-133; FBI 02 pp.43-60; Free Energy Secrets pp.1-19 and 141-156) read by their own summaries and the passages cited below; claims.md (updated); and Tesla's 1900 Century text myself [read: _text/tesla-the-problem-of-increasing-human-energy-c47de7.txt, lines ~1280-1296 and ~1350: "what is converted is clear gain"; "cannot hope to produce a self-acting machine"]. Not re-read: Rauscher/Van Bise and Corum derivations in detail (their deep reads say what they show). Tags: [read], [web], [memory], [calc] (Python, numpy; steam by IAPWS-IF97 through iapws in a scratch venv). Web checks this revision: 1 (Schumann width), after 6 used earlier. Steps 2-4 do not need redoing: QE's MINOR verdict stands. One verdict changes (A); four gain new split lines (B, C, D, F); the rest is added record.

A. Claim 4: Tesla's 1900 "sink" argument. VERDICT CHANGES: Not refuted -> Refuted by calculation (partial-gain step)

What was wrong: this arbiter, the primer (4.4 and evidence table) and the Mathematician's framing quoted Tesla's caveat and stopped. The paragraph continues: "necessitating continuous pumping out. But evidently there will be less to pump out than flows in ... some energy will be gained from the medium. What is not converted in flowing down can just be raised up with its own energy, and what is converted is clear gain" [read: Century reprint p.21]. Thermodynamics2's deep read had flagged exactly this; I did not carry it forward because I took the deep reads through the reviewers' quotations. That was my error.

Calculation [calc]: take a heat engine between ambient T_h = 300 K and a sink cooled to T_c = 77 K. Per unit heat Q_h taken from ambient the best engine gives work Q_h(1 - T_c/T_h) = 0.743 Q_h and rejects Q_c = 0.257 Q_h to the sink. "Raising up what is not converted" means refrigerating Q_c back to ambient; the minimum work is Q_c(T_h/T_c - 1) = 0.257 Q_h x 2.896 = 0.743 Q_h. Gain minus cost = 0.000 (exact, to rounding). No part of the downward conversion is "clear gain" once the sink must be restored. Tesla's error is to treat heat as a fluid that keeps its capacity to do work after being lifted.

What stays: Tesla did not claim a working machine in 1900 ("cannot hope to produce a self-acting machine deriving energy from the ambient medium for a long time yet"); he reports the liquid-air consumer's gain offset by the producer's loss; the Linde and Dewar work he cites is real refrigeration. The claim is a hypothesis stated by Tesla in good faith, from a heat-as-fluid picture; it is not a hoax.

Consequences: (i) Nichelson's reading (Tesla meant net work from one ambient reservoir) is closer to the text than the primer said. The primer lesson "later readers drop the caveat" must say instead that Tesla argued the partial-gain version himself. (ii) Nichelson's further reading, a closed unipolar dynamo or bifilar coil doing this, remains Refuted by calculation on its own grounds (claim 4 above); that is not Tesla's text. (iii) Lindemann's 1995 "Self-Acting Engine" article builds on this paragraph, claims working models (Doczekal 1930s, Schaeffer and Serogodski DE 42 44 016, Wiseman) and gives no power-in/power-out numbers [read: Free Energy Secrets deep read §4b]; unmeasured, so Never properly tested as to the devices, and the thermodynamic principle they invoke is the one just refuted.

B. Claim 1: new lines for Tesla's own claims of tests, and for replications

Added to the split verdict (claim 1 above), each with its stake (Tesla raising money 1899-1904 and 1927; O'Neill an admirer; Grotz, Golka, the Corums and Van Voorhies hold or promote related projects):

• "26 miles, 200 lamps, more than 95 per cent" (O'Neill 1944, pp.171-172): Never properly tested. O'Neill says "by piecing together the fragmentary material published in a number of publications" and names none. The deep read traces each element: "200 incandescent lamps" and "a loss not exceeding a few per cent" are in the World System prospectus (a current "around the Globe"), 96-97 per cent is the 1917 Edison Medal speech [read: Prodigal Genius deep read §15b]; the documented Colorado lamp tests are one to five lamps, at about 60-100 ft by induction (Century 1900; Notes, 1 January 1900) or fed from a grounded coil at a distance the Notes do not state (2 January 1900; corrected in revision F); Tesla in 1916, asked the Colorado receiver distance, said "Oh, 10 miles or so"; Marincic reports no great-distance measurement in the diary; Tesla's June 1900 caveat ("not as yet actually effected a transmission of a considerable amount of energy ... of industrial importance, to a great distance") sits on the next page of O'Neill. Valone's two repetitions cite O'Neill and are not independent. The diary's silence is a gap to explain (diary incomplete; test reported elsewhere; no test), not a refutation. Energy check of the form O'Neill states (about 13 hp = 10 kW of lamps at 26 miles = 42 km from a roughly 300 hp = 224 kW plant) [calc]: even if all 224 kW were radiated into a hemisphere with no loss, the flux at 42 km is 2.0e-5 W/m^2; a single resonant receiving antenna has an effective area of about 0.12 lambda^2 (4.8e5 m^2 at 150 kHz, 5.3e6 m^2 at 45 kHz, the Colorado range quoted by Van Voorhies), so it can collect at most about 10 W (150 kHz) to 100 W (45 kHz), 100 to 1000 times less than 10 kW. So the specific form "10 kW of lamps lit at 42 km" is Refuted by calculation for any radiative (far-field) coupling; the claim as a historical event remains Never properly tested (a different, smaller test could have been meant; the 200 lamps are O'Neill's arithmetic, "if two hundred were used").
• Tesla's statements that he sent current around the globe: 1904 (Carpenter interview, via Corum) "I actually sent electrical vibrations around the world and received them again"; 1907 "passed a current of one hundred amperes around the globe" [read: Tesla Said, June 21 1907 letter p.111; no voltage, frequency or method given]; 1915 voice "clear across the globe ... Colorado in 1899"; 1927 "undiminished strength" with a 1,500 kW generator (the Corums elsewhere give 250 kW) [read: Valone/Corum ch.12]. Verdict: Never properly tested as an event (no record, no witness, no signal data; the 3-4 July 1899 entry does not describe one; the 1916 interview with his counsel (not sworn; revision F) gives 10 miles for the receivers; Tesla's 1904 article describes a thunderstorm to about 300 km, not a global test). "Undiminished strength" at the antipodes at his working frequencies: Refuted by calculation (40-60 dB at VLF, above). The wording "current around the globe" is also loose: 100 A is an antenna current, not a delivered power, and 1916 Tesla himself split energy "5 percent in electromagnetic waves and 95 percent in the current through the earth", a split of the source's output, not an efficiency of delivery.
• January 1900 diary: lamps lit "through the ground" by a synchronised coil; "the existance of stationary waves proves the feasibility of the project almost beyond any doubt" [read: Notes, January 1900]: the lamp lighting is Established as short-range near-field and ground-current coupling (corrected in revision F: the 1 January field lamps were in an ungrounded loop about 60 ft from the primary, coupled magnetically; the 2 January table lamps were fed through the ground at an in-laboratory distance the Notes do not state); the inference from stationary waves to global feasibility is Never properly tested. This sentence shows the diary was not simply the cautious record; the caution of 3-4 July concerned only where the waves were reflected.
• Golka's replication (Wendover, Utah): reported by Van Voorhies and Grotz (both interested in the project): a 51 ft coil and 130 ft tower built 1973-74, used by the US Air Force for lightning simulation to 1982, "twice Tesla's original power levels" [read: Valone deep read C1]. That is a replication of the transmitter. No result is published in the sources of exciting the earth-ionosphere cavity, measuring insertion loss, or receiving at a distance; Golka and Grotz's 1986 proposal predicts 46 pT at 7.8 Hz and "90-94 per cent efficient" without a calculation shown. Verdict on the transmission claim: Never properly tested (a built apparatus whose test result is not on record: an absence-of-evidence item, F below). The sources conflict on power ("150 Kwh", "over 600 kilowatts", "150 kW") and start date (1968 project; coil 1973-74); Van Voorhies writes "Tesla's worldwide wireless power concept remains unverified".
• Planetary signals (new claim 8 in claims.md): Tesla (Collier's, 1901): periodic changes with "a clear suggestion of number and order", "could not decipher their meaning", "the feeling is constantly growing on me that I had been the first to hear the greeting of one planet to another"; 1919 "might have come from Mars"; 1921 "a regular repetition of numbers" that "must have emanated from Mars" (the 1901 text does not say numbers) [read: Tesla Said deep read §A]. No laboratory record of such signals has been found in the pages read; the Notes' passage on "intelligent beings on a neighbouring planet" is the copy of a press statement of 8 December 1899 (revision F), not a laboratory record. In 1921 Tesla attributed other people's 1919-21 planetary signals to transmitter undertones. Verdict: Never properly tested (his receiver records are not in the folder; lightning, magnetospheric whistlers and local interference are among the ordinary candidates, none excluded or confirmed by the record). The Corums' Jovian reading remains an interested-party hypothesis, untested (map: Never properly tested).
• Flowers 1925: John B. Flowers drafted a 10-page proposal (16 July 1925) to test Tesla's system, including a 300 m ring electrical scale model, and a balloon demonstration with standing waves; the Bureau of Standards (H. L. Curtis, J. H. Dillinger) declined [read: Valone deep read D; known only from a 1999 newsletter via Valone]. Arithmetic in the document is internally consistent [calc by the deep reader]; it shows standing waves in a rubber balloon, which proves nothing about the earth. Added to "Absence of evidence" (item 7).

Cavity Q "as high as 1000": weighed. Source: Van Voorhies (Valone ch. 7, p.140): "Q ... generally reported ... about 6-8", but Sutton and Spaniol (1988, magnetic measurement, 3-30 Hz) "found that the previously measured Q values were limited by instrument noise ... measured levels as high as 1000, which they say were also confirmed by others"; "others" is a personal correspondence (ref. 50). I have not seen the paper; no instrument, site or definition is given [read: Valone deep read A1]. Against it in the same book: the Corums' own table (Fig. 4) gives accepted Q 3.8-7.8 and Tesla-derived 3.2-6.4, attenuation 0.20-0.30 dB/Mm; Grotz's "6 per cent per round trip" implies Q about 100 and an attenuation about 37 times below the 0.25 dB/Mm both chapters quote [calc by the deep reader, reproduced: Q = 2 pi/(-ln 0.94) = 101.5]. Wikipedia: Schumann peaks "exhibit a spectral width of approximately 20%" [web: https://en.wikipedia.org/wiki/Schumann_resonances], which is Q = f/delta f of about 5. Three incompatible pictures: Q about 5, Q 50-100, Q up to 1000. What each gives for a 7.8 Hz pulse over the 0.067 s round trip, the worked example of the primer [calc]: Q = 3.3: 37 per cent of power returns; Q = 6: 58; Q = 100: 97; Q = 1000: 99.7 per cent. So if Q = 1000 were a property of the cavity as a power store, the ELF row of the primer's Step 3 table would be close to lossless, and the primer should say so. What it does not change: (1) the VLF/LF conclusion rests on measured attenuation of 2-3 dB/Mm, not on Q; (2) a narrow spectral line (high Q of a resonance seen in noise) says little about how much power a small transmitter can inject: Sanguine delivered about 1 W of radiated ELF from 2.6 MW [web, via the Mathematician], a coupling loss of order 1e-7 that cavity Q does not remove; (3) Tesla's 20 kHz-and-above scheme is not an 8 Hz scheme. Verdict stays: ELF cavity storage with a high Q is Never properly tested as a power-transfer path (the 1000 figure is one reported measurement of a spectral property, second-hand to me, kept in the record with the 4-6 figure beside it).

Gray-type devices (also claim 4): added from the Free Energy Secrets deep read. The 1973 demonstrations (National Tattler, 1 July 1973, Tom Valentine; echoes in Probe and NewsReal) were verbal and unmetered, with company officers (Hackenberger, Lens) as witnesses and one outsider (C. V. Wood Jr., McCulloch Oil) quoted as amazed; no input or output energy was recorded. Scale [calc, assumptions flagged]: a 1.25 lb magnet thrown 2 ft needs about 3.4 J, which a capacitor of only about 0.75 microfarad at 3,000 V stores; and a "small 15-amp" motorcycle battery (assumed 12 V, 15 Ah = 650 kJ) feeding a 150 W load for 10 minutes uses about 90 kJ (14 per cent), a change a voltmeter would not clearly show. So neither demonstration, as reported, distinguishes a net-gain device from an ordinary one; this is an observation about the evidence, not a ruling on the device. Lindemann, an advocate, built the 1975 patent (US 3,890,548) and reports it "doesn't perform anything like what was described", "only a negligible amount of energy" recovered, with no bench data [read: Free Energy Secrets deep read §2]; he places the real effect in the 1986 and 1987 patents. The legal record is the book's account (raid 22 July 1974 by the Los Angeles District Attorney, a grand-theft charge dropped, a March 1976 guilty plea to two minor SEC violations; "unprovoked" and "bogus" are Lindemann's words; NewsReal 1977 independently reports the plea); no court or SEC document is in the folder. Verdict: 1973 demonstrations Never properly tested; the patent as built by Lindemann: one unquantified negative replication by an advocate, which weighs against the patent's described performance without being a controlled test, so I do not change the verdict to Tested, not supported.

C. Claim 6: teleforce wording and the Wright Field trail. WORDING CHANGES; verdicts unchanged

• "No apparatus, drawing, parameters or test exist" is overstated. Trump's 30 January 1943 abstracts list 17 exhibits A-Q [read: FBI 02 deep read, addendum]. Exhibit F: an electrostatic generator, an "open-ended vacuum tube" accelerating "presumably electrons" into a "concentrated non-dispersive" beam through air; Trump: "incapable of the transmission of large amounts of power in non-dispersed beams over long distances", and the disclosures "would not enable the construction of workable combinations of generator and tube even of limited power". Exhibit H: letters to British representatives (28 Aug 1936, 26 Oct 1937, 15 Dec 1937, 5 Apr 1938; reply 7 Jan 1938, "a polite expression of disinterest") offering a "death ray" for a fee. Exhibit J (Tesla's hand): the only description is a jet of fluid creating a vacuous space round a terminal. Exhibit Q: a signed agreement of 20 April 1935 with the Soviet trading body Amtorg, $25,000 received, for plans for 50 million volts, "very small particles in a tube open to air", range "a hundred miles or more", speed "not less than 350 miles per second" [read]. Trump's inference that Q equals F and "subsequently proved unworkable" is his; the file has no Soviet reply. The full Tesla texts are not in the folder. So: no apparatus, test data or full specification on record; a written outline existed and was sold or offered to two governments; the examiner who saw it had an institutional stake and judged it unworkable.
• Check of the figures [calc, the deep reader's, checked]: 350 mi/s is 563 km/s, 0.19 per cent of c; electrons at 50 MV travel at essentially c, so the Amtorg particles are heavy ones, which Tesla's 1940 press account ("microscopic particles") also implies. The Refuted-by-calculation verdict for known particles through 400 km of air (170 interaction lengths) stands for either case.
• Verdicts: beam as described to the press Not testable as stated; known-particle form Refuted by calculation. Added evidence for the record: the Amtorg agreement is the strongest documentary Soviet link in the file and a paid disclosure of a "directed energy" method; the 1983 airtel's Soviet-beam concern has this as background.

D. "Absence of evidence" item 2 (the 1943 papers): added documents; reading (b) strengthened but not decided

Corrects "Against (b): no primary record of Project Nick". True of the name only. The FBI file records: - 21 Aug, 5 Sep and 11 Sep 1945: Air Technical Service Command asks for a trip (Pvt Bloyce D. Fitzgerald); Col. Holliday, Equipment Laboratory, Wright Field, asks for photostats of the Trump-annotated exhibits; OAP forwards them to Lt Robert E. Houle, "please return"; Cheney says "never returned" [read: FBI 02 deep read §6, via Cheney's quotation of OAP letters; the originals are not in the file]. - 17 Oct 1945: some 75 packing boxes and trunks still at Manhattan Storage, sealed by the New York Department of Taxation, rent about $15 a month paid for Sava Kosanovic; 26 Oct 1945: Tesla's protege and two Wright Field research men asked the FBI for access and were referred to the Custodian; the FBI file does not say whether they ever saw the effects [read: FBI 01 deep read §F]. In the 1950s Manhattan Storage staff told FBI agents the only recorded visit was by "Federal authorities" in January 1943, who "had taken numerous photographs" (possibly microfilm) [read: same]. - 6 Dec 1960 (not 1959): a caller "making a research study for U.S. Air Force on subject of geophysics" wanted the papers; the FBI told him O'Neill's statement was "untrue" and referred him to OAP [read: FBI 01 deep read §B]. - 9 Feb 1981: Lt Col Allan J. MacLaren, Office of the Under Secretary of Defense (Research and Engineering), memo "Papers Recovered on the Death of Nicola Tesla": "certain of Tesla's papers may contain basic principles which would be of considerable value to certain ongoing research within the DoD"; the research is not named. FBI replies of 10 and 20 March 1981 say the FBI never held the papers and relay the examiners' opinion that there was nothing of "significant value ... or which would constitute a hazard in unfriendly hands" [read: FBI 01 deep read §E]. The same FBI file repeats (1961-1973) the internal note that the papers "were believed dangerous to the country's security if they fell into unauthorized hands"; the file does not reconcile the two. - Bloyce Fitzgerald's January 1943 statements to the FBI: Tesla said his wireless-power experiments "had been completed and perfected", a working model "which cost more than ten thousand [dollars]" lay in a Governor Clinton Hotel deposit box, and there were "some eighty trunks" of plans; the FBI confirmed the hotel held property as security for a debt of $400 since 1932 and did not examine it. The Trump party later reported that "several items of scientific apparatus ... in a deposit box at the Governor Clinton Hotel" were "standard electrical measuring instruments in common use several decades ago" [read: FBI 01 deep read §2; FBI 02 addendum §A]. Fitzgerald is an interested informant with an unclear stake; Trump had one too.

Effect: readings (a) little of value, (b) copied for military interest, (c) removal or loss. Documentary support now exists for the 1945 Wright Field interest and photostats (reading b), and for the 1981 DoD interest; nothing in the folder shows what Wright Field did with them. The FBI's own 1981 statements disagree on whether the papers were released to Kosanovic in February 1943 (letter of 10 March 1981) or were still sealed in storage in October 1945 (the 20 March letter); the "1952 release" in my item 2 above came from the 1981 letter's loose chronology and is not established. Deciding records: NARA RG 131, RG 342 (Wright Field), the Houle correspondence, Manhattan Storage and Taxation Department records.

E. Claim 5: turbine efficiency band corrected to about 25-52 per cent

QE's check of Lowenstein's first test (102.8 bhp at 53.4 lb/bhp-h, 100 lb gauge, 15 in. vacuum) used memory steam values (27 per cent). IAPWS-IF97 [calc], saturated steam at 114.7 psia expanded isentropically to 7.33 psia (15 in. vacuum) drops 193 Btu/lb: ideal rate 13.2 lb/hp-h, so 53.4 lb/bhp-h is 24.6 per cent of ideal expansion; if the steam had 100 F superheat (not stated for this test), 207 Btu/lb and 23.0 per cent. The earlier cases stand (35 lb at 176 psia to 21 in. vacuum: 29 per cent; Tesla's own 130 Btu gives 52 per cent). The band is therefore about 25-52 per cent, with the low end resting on an assumed steam condition. Brake horse-power includes gear loss in that test (Hayes: "No allowance was made in these tests for loss of power in gear"), so turbine-alone efficiency is somewhat higher. Verdict (80-90 per cent: Tested, not supported) is unchanged. The Richards name differs in the sources ("C.D." in Lowenstein, "C.B., Professor Emeritus of Mechanics, Yale" in Hayes); unresolved.

F. New "Absence of evidence" items (added to the list above)

7. Flowers 1925 and the Bureau of Standards. A written proposal to test Tesla's system, shown to two Bureau of Standards officials, declined; known only from Valone quoting a 1999 newsletter. Readings: declined on engineering judgement (the system's physics, as above); declined for lack of funds or interest; the text is partial or reshaped in the newsletter. Deciding record: Bureau of Standards (NIST) and National Archives files, 1925; Flowers's own document.
8. Golka/Grotz Project Tesla. A built full-size coil (51 ft) and a 1986 proposal to excite the cavity, with no published measurement in the sources. Readings: no funding; run with null or unpublished results; used only for lightning simulation. Deciding records: Golka's Wendover contract reports, Radio-Electronics (Feb. 1981), the 28 Oct 1986 proposal, Third International New Energy Technology Symposium (1988) proceedings, Sutton and Spaniol (1988).
9. Gray's prototypes and the 1974 raid. The book says the DA never returned the prototypes and Gray died in April 1989 "under mysterious circumstances"; no document is in the folder. Readings: devices were unremarkable and the dispute was a securities matter; suppression; business failure after Zetech. Deciding records: Los Angeles County court and DA records, SEC file, Zetech records, the Sparks, Nevada death record.
10. Tesla's 1930s Amtorg and British dealings. Documented in abstracts only (Exhibits H, Q); full texts and any British Government or Soviet files are not seen. Deciding records: NARA, UK National Archives (War Office reference "84/T/3458" quoted by Valone), Tesla Museum.

G. Verdict-scale mapping (QE process point)

Labels used earlier outside the Step 2 scale are mapped: "Not refuted" (1900 sink text) -> Refuted by calculation for the partial-gain step (A); "documented in part, unsettled" (fate of the papers) -> a historical finding, not a verdict: what happened to the papers is Never properly tested only in the sense that no inventory exists; the sealing, the examination and the 1945 requests are documented (D); "untested, not rejected" (Corum planetary reading) -> Never properly tested; "tried by Allis-Chalmers and not adopted" (large-scale turbine) -> a historical finding; the efficiency claim there is Tested, not supported as quoted (O'Neill quoting Dahlstrand, 38 per cent mechanical efficiency, definition unstated).

H. Retention and smaller items now in the record

• Marincic (editor of the Notes) lists Tesla's errors: what Tesla expected near 10 kHz occurs at much lower frequencies, the waves are in the earth-ionosphere capacitor and not in the earth, and high voltage was not needed [read: research/2026-10-08-archive-org.md §1]. This is evidence against Tesla's mechanism by the Museum's own editor and supports the worked example.
• Rauscher and Van Bise (Valone ch. 14; authors hold the '390 patent for the detector): ELF magnetic spectra at Portland and the Bay Area (1979-1987), a claimed 3.16 Hz earthquake precursor "approximately 84% accuracy", a pi/2 velocity ratio taken from Tesla's patent 787,412 (which is exactly (pi/2) x 300,000 km/s = 471,240 km/s, as the patents deep read confirms) and a 3.575e12 W energy estimate. No Mt. St. Helens data appear in the chapter, no statistics for the 84 per cent, no calibration method; three different coil wire lengths are given. Kept as an interested-party measurement with unreplicated precursor claims; it bears on cavity signals, not on power transfer. The Valone editor's note about the USGS is an endorsement, not a finding.
• US 787,412 numbers (QE spot check confirmed): the three figures (6 Hz, 0.08484 s, 471,240 km/s) follow from one model, a disturbance crossing the earth's diameter at c with the surface pattern moving at (pi/2)c [calc, patents deep read §5.5]; internal consistency does not make the model physically right. No patent in pp.404-448 gives a measured loss or efficiency. US 723,188 is only partly in the folder (PDF p.415 duplicates p.414).
• Lindemann's admitted failed build, Tutt's "cosmic ray" newspaper quotes (Brooklyn Eagle 10 July 1931; New York American 1 Nov 1933: no device, no description) and the Brooklyn Eagle/American texts are kept under claim 4's "Never properly tested" line for Tesla's 1902 and 1931-33 statements.
• 1970 letter on "junk men": second hand, 44 years after, does not name Wardenclyffe; kept as a local tradition of lost notes at a scrapped plant.
• Golka and Bass's ball-lightning model, the 1957 "wall of force" newsletter story, and the 1976 letter's own view that "Tesla's claim to setting up standing waves ... was erroneous and involved techniques similar to Project Sanguine" (the writer's view, writer redacted) are recorded; none changes a verdict.

I. Corrections to earlier documents (additional to the list above)

• arbiter.md claim 4 and summary item 4, and primer.md 4.4/evidence table: "Tesla's 1900 sink text: not refuted; consistent with the second law" is wrong (A). Fixed above in this file; the primer must be changed by the orchestrator.
• arbiter.md claim 6 and primer.md 4.6: "no apparatus, drawing, parameters or test exist" overstated (C). Fixed above in this file.
• arbiter.md "Absence of evidence" item 2: "no primary record of Project Nick" is true of the name only; "released to Kosanovic (1952, per a 1981 FBI letter)" is not established (D).
• arbiter.md claim 5 and primer.md 4.5 and 6: "29-52 per cent" (and "14-52") should read about 25-52 per cent (E); the QE figure of 27 per cent used memory values.
• primer.md 7: "his diary was more cautious than his press articles" is incomplete (B, January 1900).
• qe.md item 3.9 lists the Wright Field letters as 1945 with "Pvt Bloyce D. Fitzgerald"; the deep read says the 1945 FBI informant letter places Fitzgerald at Wright Field and Cheney quotes the OAP letters, so the chain is Cheney's quotation, not the FBI file's own copies of those letters (the FBI file holds Cheney's book pages).
• qe.md item 3.11 and the FBI 01 deep read: the "1959" Air Force caller is dated 6 December 1960.
• Summary table row "Radio priority" (earlier in this file) and everything else: unchanged.

J. Updated summary of final verdicts (replaces the changed lines of the summary above)

1. Unchanged, plus: 26-mile/200-lamp test Never properly tested (10 kW at 42 km by radiation Refuted by calculation); Tesla's "around the globe" statements Never properly tested ("undiminished" at VLF Refuted by calculation); January 1900 short-range lamps Established, feasibility inference Never properly tested; Golka replication of the transmitter reported, transmission Never properly tested; cavity Q up to 1000 Never properly tested as a power path; planetary signals Never properly tested.
2. Tesla's 1900 partial-gain step Refuted by calculation (changed from not refuted); Gray's 1973 demonstrations Never properly tested; Tesla's 1902/1931-33 devices Never properly tested; equilibrium generator and useful-power collector Refuted by calculation.
3. 80-90 per cent Tested, not supported (band about 25-52 per cent).
4. Teleforce Not testable as stated (written outlines existed; known-particle form Refuted by calculation); papers: sealing and examination documented, later fate unsettled; Wright Field and DoD interest documented, outcome unknown.

QE verdict MINOR stands. Remaining unread by anyone (as of this revision; revision F below records what has since been read): the Colorado Springs Notes entries of 30 July, 6 August 1899 and lines ~8018-8128, Martin pp.417-430 and 505-520, the 1916 interview chapter VI and Appendix III, the full Kerr, Page and Cooper opinion; the verdicts above do not turn on them but claim 1a could.

QE revision F (2026-10-08 further reads)

Inputs read in full for this revision: the retry sections of deep/tesla-the-problem-of-increasing-human-energy-c47de7.md ("Retry 15-24") and deep/ebook-Free-Energy-Secrets-with-Tesla-patents-344906.md ("Retry 137-146", "Retry 147-156"); deep/further-R1.md (Colorado Springs Notes entries and Marinčić's commentary), deep/further-R2.md (1916 interview chapter VI and Appendix III, the Wardenclyffe foreclosure appeal) and deep/further-R3.md (1904 Rankine file); qe.md sections 6-7, Corrections, Recheck and Orchestrator fix. Passages checked against the source where a correction turns on them: research/files/tesla-papers-1904-rankine-letters_djvu.txt context via R3 and research/2026-10-08-archive-org.md line 43-45; _text/tesla-the-problem-of-increasing-human-energy-c47de7.txt page markers (pp.15-24 and the p.32 caveat). Web checks this revision: 1 (Kelvin's 1851 wording), 8 of 8 used in all. Tags: [read], [web], [memory], [calc] (Python).

Summary: no verdict changes. The reads add record, sharpen wording, settle one open question in part (the pulse energy behind the "activity" figures), and correct several statements in earlier files (section F9). Each affected claim is weighed below.

Source gap after these reads. The content-filter gap of the first run is closed for Free Energy Secrets (pp.137-146 and 147-156 now deep read) and for Century pp.15-24. Still without a deep read: Century pp.1-14 and pp.25-33. Those pages are not unread: the fact-finder read pp.15-33 from the text layer in full and facts.json quotes them, and I checked the p.32 caveat myself ("While I have not, as yet, actually effected a transmission ...", _text line ~1937). They lack only the second, independent reading a deep read gives. Free Energy Secrets p.148 has no text layer. The Kerr, Page and Cooper opinion is not in the saved scan (R3), so it remains a record to obtain.

F1. Claim 4 (energy from the medium): Century pp.15-24 read in full. Verdicts unchanged; one split line added

What the read adds: - Tesla's second law is a paraphrase that drops one word. Tesla: it is "impossible for an inanimate mechanism or self-acting machine to cool a portion of the medium below the temperature of the surrounding, and operate by the heat abstracted" [read: Century reprint p.19]. Kelvin (1851): "It is impossible, by means of inanimate material agency, to derive mechanical effect from any portion of matter by cooling it below the temperature of the coldest of the surrounding objects" [web: https://collection.sciencemuseumgroup.org.uk/people/cp37484/william-baron-kelvin-thomson; https://quotepark.com/hu/idezetek/1795452-william-thomson-it-is-impossible-by-means-of-inanimate-material-ag/ (one quote collection prints "coldest surrounding bodies"; the 1851 paper itself was not opened)]. The same sources report that Kelvin's footnote to this axiom already speaks of "a self-acting machine" producing mechanical effect "by cooling the sea or earth" if the axiom were denied [web, same; footnote wording from the search summary, not seen in the original]. So Tesla's "self-acting machine" is Kelvin's own phrase, and the difference between them is "coldest". Tesla's thermopile from the earth to outer space (pp.19-20), which he offers as an inanimate engine that cools the medium below its surroundings, uses outer space, which is the coldest of the surrounding objects: it is an ordinary heat engine between two reservoirs, and Kelvin's wording allows it. - New split line: the thermopile-to-outer-space thought experiment is consistent with the second law (a two-reservoir engine; as thermodynamics, Established). It is not an exception to Kelvin. Whether such bars could be built or give useful power is not addressed by Tesla and is not tested; I give no device verdict. - The partial-gain step stands as before. The deep reader reproduces the cancellation independently with other numbers (T_h = 300 K, T_c = 80 K, Q_h = 100 J: W = 73.3 J; Q_c = 26.7 J; minimum pumping work 26.7 x (300/80 - 1) = 73.3 J) [calc, deep reader; checked]. The quotation "What is not converted in flowing down can just be raised up with its own energy, and what is converted is clear gain" is confirmed on p.21, with Tesla's remark that the plan "seemed to be contrary to the statements of Carnot and Lord Kelvin" [read: p.21]. Verdict unchanged: Refuted by calculation (partial-gain step), caveats kept. - No test of the self-acting engine in pp.15-24. Tesla's chronology (sink scheme conceived in Paris, late 1883; resumed 1889; Dewar's experiments seen in London, early 1892; the mechanical oscillator shown in Chicago, summer 1893; oscillator and compressor working by the close of 1894; a refrigerating element begun when the laboratory burned; elements four and five never named) is a programme, with no measured output or efficiency [read: pp.21-22]. Tesla's own admissions sit beside it: liquid air is uneconomical because "it takes coal to keep air cold", and a 200 lb casting warmed by the air gives "about one effective horsepower during one hour" but "this gain of the consumer is offset by an equal loss of the producer" [read: p.22]. The deep reader's check (2.68 MJ per hp-hour; 200 lb of iron would have to cool by about 66 K to supply it from its own heat) makes the figure a plausible ambient heat inflow, as Tesla says, not a gain [calc, deep reader]. Verdict for Tesla's 1902 and 1931-33 statements is unchanged (Never properly tested); the 1900 text itself claims no working machine. - Other items now in the record (none central; kept for completeness): a limestone engine "operated satisfactorily" and a cold-coal battery whose "lack of success has been complete" (both chemical energy, no figures) [read: p.17]; two "novel facts", a current in a wire from the ground to a pointed terminal at height and the upper air charged opposite to the earth, earth-air-upper strata forming "a highly charged electrical condenser" [read: pp.18-19] (the second agrees in outline with the mainstream global circuit [memory, high]; the first gives no figures); solar input "over four million horsepower per square mile", which is 1,150 W/m^2 and sound [calc, deep reader, checked: 4e6 x 745.7 / 2.59e6 = 1,152]; "thirty times as much aluminium as iron in bulk", unexplained (crustal ratio about 1.4 [memory, high]); diagram b's enclosure T, where the word "cylindrical" is printed in square brackets in the reprint [read: page image p.20], so it should be called Tesla's "enclosure T".

F2. Claim 4 (later readers): Lindemann's 1994 and 1995 articles, and US 787,412 as reprinted. Verdicts unchanged; three lines added

• US 787,412 as reprinted in the book (pp.137-139) describes collecting synchronised impulses in a condenser to work an insensitive receiver; it claims no gain and gives no received-energy figure [read: Free Energy Secrets deep read, Retry 137-146 A]. It is not evidence for "radiant energy", and the book makes no claim about it on those pages.
• Lindemann 1994 ("Thermodynamics and Free Energy" and its "Ethericity" remainder): an "Etheric Energy Field" moves energy from low to high concentration without work, so the second law is broken; the First Law is "an outmoded, intellectual MODEL"; "over-unity" is an oxymoron. His numbers are quotations, not measurements: a "100,000 to 1" weather-engineering gain that "has never been done"; Tewari's "twice as much hydrogen" from a rotating-magnet generator (no setup, no citation); unnamed patents heating water with "less than half" the mechanical work; "millions of gallons" of rain he says he saw, with no measured volume [read: Retry 137-146 B-C; Retry 147-156 A]. Lindemann and T. J. Constable's group are interested parties (Lindemann a BSRF board member writing in its journal).
• Etheric field breaking the second law: Not testable as stated (no defined quantity, no measurement).
• "The exact equivalence of ft-lb, watt-hours and BTU is not known": Tested, not supported. The mechanical equivalent of heat has been measured repeatedly since Joule (1840s) and Rowland (1879) by paddle, electrical and other methods, all agreeing [memory, high]; Lindemann cites unnamed patents and no data. The Joule figures he himself quotes (772.5 and 778.26 ft-lb per Btu) are close to the modern 778.17 ft-lb per Btu [memory, high].
• Tewari's hydrogen doubling and the DePalma/Tewari rotating-magnet generator: Never properly tested in this record (one second-hand sentence; no measurement in the folder). This is close to Nichelson's closed unipolar dynamo (claim 4, Refuted by calculation as a net-work machine from equilibrium surroundings), but the electrolysis claim is a different, testable statement and is kept separate.
• Lindemann 1995 ("Tesla's Self-Acting Engine"): his Figure 3 cycle has a compressor bringing a two-phase fluid to liquid, an ambient heat exchanger, a throttle that flashes part of the liquid, and a turbine on the wet stream, with no condenser; turbine output is said to exceed compressor input [read: Retry 147-156 B]. A cycle that takes heat from one reservoir (the ambient) and rejects none cannot give net work (Kelvin-Planck); the throttle is also irreversible, which can only lower the output. Refuted by calculation as a net-work cycle. The named machines (Doczekal, Austrian patent "NR. 155744", 1939, "well above" Carnot; Schaeffer and Serogodski, DE 42 44 016 A1; Wiseman's HEAT books, for which Lindemann prints Eagle Research's address and price) carry no test data, efficiency or witness: Never properly tested (as in QE revision A(iii)). The Doczekal claim of exceeding Carnot "with or without condenser" is a claim the patent office record could check; the Austrian patent has not been read.
• Book's back cover (seller's copy, an interested party): Tesla found the "cold electricity" effect in 1889; blurbs by Constable, Desborough ("well over 1000 amps", no set-up) and Rauen (Infinite Energy, January 2001) [read: p.156]. Recorded, no verdict. The retry found no account of Lindemann's failed Gray build in pp.147-156; it is on p.16 (QE deep read), as facts.json already says.

F3. Claim 1a (Colorado observations): R1 adds the arrester experiments and fixes the January 1900 set-ups. Verdicts unchanged; two lines added, one reworded

• 24 July 1899. With the secondary moved from the water pipe to an isolated ground plate, strong sparks still crossed the lightning arresters; Tesla: "the earth behaves simply as an ordinary conductor and ... it will be possible, with powerful apparatus, to produce the stationary waves which I have already observed in the displays of atmospheric electricity" [read: Notes line 5077-5088].
• 30 July. A 240,000-per-second system on a 30-inch ball "hoped" to produce stationary waves; no result is recorded in the entry. Arrester sparks appeared only with a heavy-cable connection, which Tesla read as "short waves are more effective giving higher e.m.f. per unit of length in the ground" [read: lines 5768-5799]. Marinčić: "the standing waves in terms of which he tried to explain the arcing over the lightning arresters cannot be significant at these frequencies" [read: lines 21457-21460]. The wavelength and turn figures on pp.115-116 do not agree with each other (2000 ft against about 4,100 ft at 240 kHz; 22,500 ft of wire) and the formula is garbled in the OCR [calc, R1]; nothing is built on them here.
• 26 August. Thick arcs on both arresters with waves Tesla puts at "3-4 miles long" (about 47-62 kHz [calc, R1]); whether the ground connection is a node or a crest is "to be decided by further observations"; Tesla asks himself "Could the sparks be produced by static induction upon wire through the air and not chiefly by conduction through earth?" and finds 3/8-inch sparks in a coil on the arrester ground wire, showing current in that wire [read: lines 7972-8081].
• New split line: arrester sparks between grounds tens of feet apart are Established as local ground-potential differences and induction from a nearby high-frequency transmitter (standard physics; Tesla's own measurements). As evidence of stationary waves in the earth they are Never properly tested: Tesla's own alternative (induction through the air) was raised and not excluded, and the editor of the Notes judges standing waves insignificant at these frequencies.
• 6 August (water pipe). The diary lists the experiment among items "partly worked out" and omitted; the only datum is "Distance from groundplate to End of main exactly 550 feet" (if a quarter wave, about 450 kHz [calc, R1; inference only]); line 6993 is a bare heading [read: lines 6320-6327, 6993]. No measurement exists to weigh; no verdict line. Earlier files that list it as an experiment to read are corrected in F7.
• 1-2 January 1900 (reworded line). Two set-ups, not one: on 1 January three 100 V lamps in parallel in an ungrounded 62 ft 5 in square loop centred "a little over 60 feet" from the primary loop, tuned to about 42,640 per second (Tesla's figure; 1/(2 pi sqrt(1.39e-4 H x 1e-7 F)) = 42,600 Hz [calc, R1, checked]), magnetic coupling; on 2 January five lamps in series between ground and the lower end of a grounded coil (98 turns, matched to a quarter wave at 93,110 per second, wire 2,660 ft) with the oscillator's own windings short-circuited, so the energy came through the ground, over a distance "Tesla gives no data about" (Marinčić) [read: lines 18198-18560, 22520-22521]. The verdict stays Established for both as short-range coupling (one magnetic, one through the ground in or beside the laboratory); the feasibility inference stays Never properly tested.
• Tesla's 2 January energy estimate. From the coil's free-end capacity (about 162 pF) at about 141,400 V and 93,000 Hz he finds 300,000 W "set in movement", of which the lamps take 1 per cent; he calls it "merely an example" [read: lines 18575-18636]; C V^2 f = 1.62e-10 x (1.414e5)^2 x 9.3e4 = 3.0e5 [calc, R1, checked]. Marinčić: this is the energy circulating in the resonator; "the active power dissipated in the circuit is much less than this and is inversely proportional to the Q-factor" [read: lines 22512-22515]. So the 300 kW is a circulating (reactive) rate, like the "activity" figures, not power that could be sent to a distant circuit. As an estimate of transferable power: Refuted by calculation (the standard energy bookkeeping of a resonator, as the editor states), with Tesla's own "merely an example" beside it. As arithmetic on his own assumptions it is consistent.
• Tesla's magnifying factor on 2 January, 10,530 for the coil alone and "approximately only 10.5" with the lamps, is recorded with his argument that hot filaments made the true figure "very much larger" [read: lines 18510-18545]; no verdict.

F4. Claim 1b (any distance, and "I did it"): R2 and R1 strengthen the existing lines. Verdicts unchanged

• In the 1916 interview Tesla himself bounds the Colorado distances: they "were small", meant "to give me quantitative data", "10 miles or so"; his greatest signalling distance ever was about 30 miles (Houston Street to West Point, before 1897); asked where the grounded no-antenna coil (Fig. 64, "at the close of 1899") stood, "That was out in the field. I cannot tell you"; asked the greatest distance from Long Island, the plant "was not yet quite ready"; Wardenclyffe experiments were "simply preparatory" [read: R2, lines 5677-5685, 9338-9392, 2230-2243]. His statement that the coil would light "no difference" at the antipodes or 10 feet away is a design statement, not a result [read: lines 5598-5648].
• Effect: the 26-mile/200-lamp line and the "around the globe" line stay Never properly tested, and the weight against O'Neill's account increases: in a long technical interview with his own counsel, in 1916, Tesla claims no Colorado test beyond about 10 miles. This is Tesla's own account and Tesla had a stake (the interview served patent litigation); it is not neutral, but its stake runs towards claiming more, not less.
• The 1904 "Transmission of Electric Energy Without Wires" (in the Rankine file, R3) repeats the 1904 claims already weighed: the earth "behaved like a conductor of limited dimensions", power "to any terrestrial distance and almost without any loss", a transmitter of "ten million horse-power" maximum activity, a 10,000 hp first plant at "one hundred million volts" [read: R3]. No change.
• Wardenclyffe in 1922 testimony: power transmitted "with a loss of not more than five per cent" [read: R2, lines 11001-11050]. Covered by the existing line: "almost without loss" to any distance Refuted by calculation.
• Tesla's "law which I discovered in Colorado" (propagation through the earth with a surface velocity varying as the cosecant of an angle; light speed at 6,000 miles) is described with no Colorado data in those pages [read: R2, lines 7784-8010]. It is the same geometric model as US 787,412's 471,240 km/s ((pi/2)c), already checked as self-consistent (revision H); no new verdict.
• Marinčić's introduction adds two points for the reader: Tesla predicted earth resonances at "6, 18 and 30 Hz" against the measured 8, 14 and 20 Hz; and Tesla "could not have known" the effects would show only at very low frequencies "because it seems he was never able to carry out the experiments" [read: R1 §6]. He also credits Tesla with one correct observation, that which kind of wave dominates depends on how the earth is excited [read: lines 21673-21677]. These support the existing "Established (ancestor, not mechanism)" and "Refuted by calculation" lines; no change.

F5. "Activity" figures (claim 1, open question 7): partly settled by Tesla's own 1916 numbers

Tesla in 1916: Colorado antenna current 1,000 A; tension "3 1/2 or 4 million volts"; antenna capacity 500-600 cm (550-660 pF); "18 million horsepower activities" from condenser energy discharged in a tiny interval; peak "7 or 8 million volts" in the ball experiment [read: R2 A2]. Calculation [calc]: - Stored energy 1/2 C V^2 at 4 MV: 4.4 kJ (550 pF) to 5.3 kJ (660 pF). - 18 million hp = 1.34e10 W, which that store sustains for 0.33-0.39 microseconds. - Consistency: a 1,000 A peak current into 550-660 pF at 4 MV needs I = 2 pi f C V, so f = 60-72 kHz, inside the 50-150 kHz range his Colorado circuits used. Tesla's three 1916 numbers are mutually consistent. So the "18 million horsepower" is the whole antenna store released in about a third of a microsecond, about 2-3 per cent of a cycle at the derived 60-72 kHz (period 14-17 microseconds). The 250 kHz figure belongs to Tesla's separate rate argument (R2 A2), not to this check. That is a peak rate, as Tesla said. The earlier figure of "about 2 ns" (claim 1 above; primer Step 6) used the 0.0225 uF primary condenser at an assumed 50 kV, a different store; both are peak rates, and neither is delivered power. Open question 7 is answered in order of magnitude for the antenna store; the energy per pulse of the "ten million horse-power" Wardenclyffe design figure is still not given anywhere.

F6. Wardenclyffe and the 1904 money (Absence of evidence item 1): Exhibit B and the Rankine file read. No verdicts; item amended

• The Kerr, Page and Cooper opinion is not in the saved file. The only trace is Tesla's letter to Rankine of 19 April 1904, "enclosed short statement of Kerr, Page & Cooper, relative to some of my patents ... you will see that they are controlling", followed by "These patents have an absolutely assured value of certainly not less than $5,000,000" [read: research/files/tesla-papers-1904-rankine-letters_djvu.txt lines 307-311, via R3]. "Controlling" is Tesla's reading of the enclosure, and the valuation is Tesla's sentence. The catalogue entry lists the firm's reply as part of the item (research/2026-10-08-archive-org.md line 43), so a fuller scan or the Smithsonian original may hold it. Corrected inline in item 1 above.
• Tesla's own financial claims in the letter (interested party, soliciting money): the Nikola Tesla Company "has no liabilities" and its assets are his patents; Wardenclyffe "could now be finished in three to four months" for no more than $100,000 and would earn "ten thousand dollars a day"; investors would get "about seven times" their money in the worst case [read: R3]. The file records nothing after 1904. These are forward claims; the record shows the plant was never completed.
• The 1922 appeal record (Appendix III; parties: the Waldorf-Astoria Hotel Company and the Boldt estate as creditors, their engineer Bingham, and Tesla as defendant):
• Tesla describes a 100 x 100 ft brick plant with a 400 hp Westinghouse engine and special dynamo, two 300 hp boilers, eight lathes, a 100 hp motor feeding condensers, four 60 kV oil-tank transformers, four condenser tanks (two nearly built), his own generating apparatus of 1894-95 "used in Colorado", a thousand bulbs and tubes; a 187 ft tower with a 55-ton terminal never plated; a 120 ft shaft with sixteen pipes driven 300 ft [read: R2 B2].
• Exhibit B (a bill of sale of 8 April 1915 from William N. Hallock to the hotel company, "including but not limited to") lists conventional power-plant and machine-shop items: an auto-compound engine with a 200 kW alternator, a 25 kW set, two Babcock and Wilcox boilers, four lathes, motors of 15 hp and under, four 15 kW transformers, "4 high-tension transformers in tanks", two "submarine boat[s]". It lists no oscillator, condenser bank or coil apparatus [read: R2 B4].
• Bingham, the plaintiffs' engineer, describes visits from about 1913 to July 1915: windows open, desks rifled, "four truckloads" of books and papers on the floor, brass and fittings stripped, some machine tools removed (some by himself to the Waldorf), a lathe taken by a garage owner with Tesla's permission according to the station agent (hearsay in the record), and a tower "badly rotted" and "absolutely unsafe" in his opinion; the referee struck some of his characterisations [read: R2 B3].
• Weighing: the plant's counts differ between Tesla's 1922 memory and the schedule (400 hp against 200 kW; eight lathes against four; a 100 hp motor against 15 hp). Exhibit B is a 1915 bill of sale to the hotel company (Hallock to Waldorf-Astoria), entered as a defendants' exhibit (the page is headed "Defendants' Exhibit B" and Tesla testified for the defendant); who drew up the schedule is not stated. [Correction 2026-10-08, Recheck 2: earlier wording called it "the plaintiff-side schedule"; that was wrong.] The record does not settle which count is right, and both are kept. The absence of Tesla's special apparatus from the schedule shows it was not listed, not that it was absent (the schedule says "including but not limited to").
• New reading added to item 1, for the missing special apparatus: (d) removed earlier by Tesla or his staff (for example to storage in New York; the 1943 effects in Manhattan Storage would then be where to look); (e) lost to scavenging during 1913-15, as Bingham's account of an open, rifled building allows; (f) included under generic schedule lines ("4 high-tension transformers in tanks"). Deciding records: Exhibit A's schedule (the Tesla-to-Holmes bill of sale of 30 March 1915), the Manhattan Storage inventory against Tesla's 1922 list, and the Tesla Museum's holdings of Wardenclyffe apparatus.
• What the appeal record adds to "no test of the full plant": Tesla's own testimony that the plant could have served "as a receiving station and also as a transmitting station, but not to the extent that it could in the fully developed plant", with no distance or result [read: R2 B2], and his 1916 "not yet quite ready". No transmission measurement from Wardenclyffe exists in any document read. Readings (a)-(c) for the tower's end are unchanged; Tesla dates the destruction "about in 1917".

F7. Claim 8 (planetary signals): wording only

The Notes' "intelligent beings on a neighbouring planet" sentence is the text of a statement Tesla sent on 8 December 1899, at I. Hawthorne's request, to the North American (Philadelphia) on the century's achievements; the same day's laboratory work measures a coil inductance and has nothing on signals [read: R1 §4]. It is a press statement copied into the diary, not a laboratory draft. R1 did not search the 3 July "signals" entries that the planetary claim is usually tied to (that link is [memory, medium], R1). Verdict unchanged: Never properly tested.

F8. Updated summary of final verdicts (adds to J; nothing replaced)

1. Unchanged, plus: arrester sparks between nearby grounds Established as local coupling, as evidence of earth stationary waves Never properly tested; 1 January 1900 field lamps (magnetic, about 60 ft) and 2 January lamps (through the ground, distance unstated) Established as short-range coupling; Tesla's 2 January "300,000 W set in movement" as transferable power Refuted by calculation (circulating energy; his own "merely an example" kept). "18 million horsepower" = the 4.4-5.3 kJ antenna store over about 0.3-0.4 microseconds [calc]: a peak rate, as before.
2. Unchanged, plus: Tesla's thermopile-to-outer-space example consistent with the second law (Established as thermodynamics; not an exception to Kelvin); Lindemann's condenser-less cycle Refuted by calculation; Lindemann's etheric-field breach of the second law Not testable as stated; "the mechanical equivalent of heat is not known" Tested, not supported; Tewari's hydrogen doubling Never properly tested.
3. Unchanged (Never properly tested); the Notes passage is a press statement.

What would settle the new lines: for the arrester sparks, a modern repeat with a high-frequency transmitter, two grounds 60 ft apart and a loop probe in the air beside the ground wire, which separates conduction through the ground from induction through the air; for Lindemann's cycle, one metered run (shaft work out, compressor work in, ambient heat in) of any built version; for the Wardenclyffe apparatus, Exhibit A and the Manhattan Storage inventory.

F9. Corrections to earlier documents

1. arbiter.md "Absence of evidence" item 1, primer.md section 6 ("Where the record goes dark"), and contrarian.md line 206: "Kerr, Page and Cooper opinion that the patents were 'controlling' and worth 'not less than $5,000,000'" (primer: "patent counsel Kerr, Page and Cooper called his patents 'controlling'"). Wrong. The saved file holds only Tesla's covering letter: "controlling" is Tesla's reading of an enclosed statement whose text is not in the scan, and the $5,000,000 is Tesla's own sentence, written to a prospective backer [read: research/files/tesla-papers-1904-rankine-letters_djvu.txt lines 307-311; deep/further-R3.md]. Fixed in item 1 above and in primer.md; contrarian.md and research/2026-10-08-archive-org.md §4 (line 45, "whose text is in the same scan") need the orchestrator: the opinion's text is not in this scan.
2. arbiter.md QE revision B (26-mile and January 1900 lines), primer.md section 4.1 (lines 82 and 95) and the evidence-table row "January 1900 diary": "lamps sit at tens of feet from the primary" and "one to five lamps at about 60-100 ft by near-field induction" merge two set-ups. The 1 January field lamps were in an ungrounded loop centred about 60 ft from the primary (magnetic coupling); the 2 January table lamps, which carry the "through the ground" and "feasibility" text, were fed by a grounded coil at a ground-path distance the Notes do not give [read: Notes lines 18209-18215, 18391-18459; 22520-22521; deep/further-R1.md Corrections 2]. Verdict unchanged. Fixed above and in primer.md.
3. qe.md lines 42 and 237, research/2026-10-08-archive-org.md line 18, contrarian.md line 203, mathematician.md line 151, and claims.md (Remaining reads, before this revision): "6 August (water-pipe stationary waves)" is listed as if the diary records the experiment. It does not: it is listed among items omitted for want of time; the only datum is the 550 ft distance [read: Notes lines 6320-6327, 6993]. claims.md is fixed; the others need the orchestrator.
4. claims.md claim 8, arbiter.md QE revision B (planetary line) and primer.md section 4.7 and the Source gap note: "a draft note"/"draft passage" at line ~17149. It is the copy of an 8 December 1899 press statement to the North American [read: Notes lines 17127-17149]. Fixed in claims.md, here and in primer.md.
5. qe.md line 61: places the feasibility sentence at "lines ~18535-18560"; it is at lines 18559-18560 of the 2 January entry [read: R1 Corrections 3]. Small; for the orchestrator.
6. claims.md claim 5 (stale copy, after the orchestrator fix of line 40): still said "about 27%" for Lowenstein's first test; it should be about 25% (QE revision E). Fixed in claims.md; primer.md's correction note that flagged it is updated.
7. arbiter.md claim 1 ("the energy per pulse in Tesla's figures is unresolved"), primer.md Step 6 and section 7 ("a peak rate during a few nanoseconds of discharge"): incomplete rather than wrong. The nanoseconds figure comes from the 3 July primary condenser at an assumed 50 kV; Tesla's own 1916 antenna figures give a store of 4.4-5.3 kJ and about 0.3-0.4 microseconds at 18 million hp (F5). Primer updated.
8. physicist.md line 70 and the Nichelson entry of facts.json ("a cylinder", "Tesla's cylinder"): the word "cylindrical" is printed in square brackets in this reprint, probably an editor's insertion; Tesla's text says "an enclosure T" [read: page image p.20]. Not material to any verdict; recorded so later files quote Tesla's word. Whether the 1900 Century printing has the word is unknown from this file.
9. arbiter.md QE revision B (the "around the globe" line): "the 1916 sworn-style interview" is loose: R2 describes it as an interview with questions by counsel, prepared for patent litigation [read: R2 header]; it should not be called sworn testimony. The 1922 appeal transcript (Appendix III) is testimony.