Arbiter (Step 3): Poincare-Henri, "The New Mechanics"

Role: weigh physicist.md, mathematician.md and contrarian.md, recompute where they differ, web-check remaining medium/low-confidence [memory] facts, and give a final verdict per claim on the Step 2 scale. Tags: [read] project document or project files; [web] URL; [calc] my own arithmetic (Python, .venv); [memory] with confidence.

Source gap note. None for the document: the one source, Halsted's English translation (The Monist 23(3), July 1913, pp. 385-395), has a full text layer and no equations [read: p.1-11]. But the project has only this one primary document. Everything about the essay's context (the 1909 Gottingen lecture, Lille lecture, the 1906 memoir, de Sitter, Bucherer, Lorentz) rests on the Investigator's web notes and the reviewers' web checks, not on files in the project folder. research/files/ is empty. Where a conclusion depends on those outside texts it says so below.

Terms for the reader. c = speed of light (about 300,000 km/s). Ether = the hypothetical light-carrying medium. gamma = 1/sqrt(1 - v^2/c^2), the Lorentz factor. Perihelion = the point of a planet's orbit nearest the Sun; its slow rotation is "perihelion motion". 1 arcsecond (") = 1/3600 degree. GR = general relativity.

Agreement across the three reviewers, and what I corrected

All three agree on the identity and date: the Gottingen Wolfskehl lecture of 28 April 1909, French in Sechs Vortrage (1910) pp. 49-58, English in The Monist, July 1913 (posthumous; Poincare died 1912). This corrects facts.json's "[memory, medium] about 1910". I re-checked the English text on Wikisource: header gives The Monist vol. 23, pp. 385-395, Halsted, lecture of April 1909, French published 1910 [web: https://en.wikisource.org/wiki/The_New_Mechanics_(G%C3%B6ttingen)]. That page also contains the "38"" and "32"" sentences and Kaufmann, and no Bucherer, which matches the project PDF.

All three agree on these corrections to Step 1, which I accept: - "38"" and "32"" carry no time unit in the essay [read: p.10]; the unit is per century (inference from Le Verrier/Newcomb and de Sitter). 6" = 38" - 32" appears only in the Lille text. - The essay's Mercury speed "about 100 km/s" is wrong: 47.9 km/s [calc: 2 pi a/T, a = 5.791e10 m, T = 87.969 d gives 47,873 m/s]. It does not affect any argument. - Michelson-Morley cannot "perfectly verify" the contraction: a null result is what contraction explains, and what several other pictures also give. - Kaufmann is cited in the essay for "constant mass negligible" while Kaufmann himself concluded his 1905-06 data disproved the Lorentz-Einstein formula [web: https://en.wikipedia.org/wiki/Kaufmann%E2%80%93Bucherer%E2%80%93Neumann_experiments, which says he believed his experiments "conclusively disproven the formula of Lorentz-Einstein"; also Bucherer's report via the Investigator].

Numbers recomputed where reviewers differ or stated slightly differently

Web checks done this run (budget 8; 5 used)

1. Wikisource Gottingen text (header, 38"/32", Kaufmann present, no Bucherer): [web: https://en.wikisource.org/wiki/The_New_Mechanics_(G%C3%B6ttingen)]. Settles the date/translator facts and the missing Bucherer passage in the Monist text.
2. Kaufmann-Bucherer-Neumann page: Kaufmann (1905-06) read his data against Lorentz-Einstein; Bucherer 1908 for it; Neumann 1914 "definitely proving" it for 0.4-0.7 c; Zahn and Spees 1938 argued the early experiments "would only show a qualitative increase of mass" and could not decide between theories; precision disputes lasted to Rogers et al. 1940 [web: https://en.wikipedia.org/wiki/Kaufmann%E2%80%93Bucherer%E2%80%93Neumann_experiments]. This retags the physicist's and mathematician's [memory] "tested and replicated" statements for claim 1, and sharpens them: the 1900s beta-ray data alone did not decide the matter.
3. Electromagnetic mass: the 4/3 factor, Poincare stress as a non-electromagnetic pressure carrying one third of the field energy, and the pure-electromagnetic-mass hypothesis "rejected in contemporary physics" [web: https://en.wikipedia.org/wiki/Electromagnetic_mass]. Retags the physicist's and mathematician's [memory, high] claim 4 points as [web].
4. Kennedy-Thorndike: Lorentz-FitzGerald contraction alone explains Michelson-Morley but not Kennedy-Thorndike, which needs time dilation too [web: https://en.wikipedia.org/wiki/Kennedy%E2%80%93Thorndike_experiment]. Retags claim 3.
5. Tried the Abraham-Lorentz force page for the same 4/3/Poincare stress points: it does not discuss them (no result; item 3 covers it).

Still [memory], not web-checked, and no verdict depends on them: that the contraction cannot be seen directly in a comoving ruler (the essay itself says so [read: p.6]); Poincare's 1905 "gravific waves" (medium); the gravity-speed limit from pulsars/GW170817 (medium); GPS corrections; the Paris-Berlin distance; the Sun's 370 km/s CMB speed (high). Statements the Investigator's notes carry (Bucherer table, de Sitter 7".15, Lorentz 1900, Lille text, Palermo memoir "postulate ... may be confirmed or disproved") are [web] via the Investigator and I did not re-open them; the reviewers agree on them.

Corrected claim list (what the document actually claims, after review)

1. Mass (inertia) grows with velocity; light speed cannot be exceeded; radium-ray deflection experiments show the velocity dependence (no experimenter named in the essay; "Thus we have ascertained" overstated the contested 1909 evidence) [read: p.2-3].
2. The principle of relativity has absolute value; observers carried through the ether can know only "apparent" (local) time, not true time; at Lille Poincare is "absolument sur d'avance" that an aberration test of absolute motion will find nothing [read: p.4-6; web: fr.wikisource Lille text via the Investigator].
3. A body is deformed along its motion (Earth about 1/200,000,000, correct arithmetic); Michelson's experiments "perfectly verify" it (overstated: a null result is compatible with contraction but does not isolate it) [read: p.6-7].
4. Matter's constant mass vanishes: electron inertia is the ether's; Kaufmann's experiments show the real mass is negligible [read: p.7-8]. 5a. Lorentz's velocity-dependent attraction fixes the sense of Mercury's perihelion error; 38" observed, 32" left (per century by inference) [read: p.9-10]. 5b. Accelerated bodies radiate; planets would fall into the Sun in "millions of milliards of centuries" (1e17 years) [read: p.9].
5. Newtonian mechanics is not finally condemned; it is the mechanics of small velocities [read: p.10-11].

(claims.md bundled 5a and 5b; the mathematician split them and I keep the split. Claim 5 in claims.md stays as the heading.)

Verdicts per claim

Claim 1: velocity-dependent mass, c as limit

• Agreement: all three say Established, with the idiom caveat (modern usage: invariant mass, momentum gamma m v; longitudinal mass gamma^3 m, transverse gamma m) and the note that the limit is assumed in the essay, not derived.
• Disagreement: none on the verdict. Emphasis differs: the contrarian stresses that in April 1909 the beta-ray evidence was contested (Kaufmann's own reading; Bestelmeyer's error doubts); the physicist and mathematician mention it more briefly.
• Stronger argument: the contrarian's, on the 1909 state of play, which my check supports (Zahn-Spees 1938: early experiments could not decide; Rogers 1940). The verdict itself rests on much later evidence (Neumann 1914 on the same method, then accelerator and Bertozzi-type tests [memory, high; Bertozzi not web-checked]).
• Final verdict: Established (velocity limit and velocity-dependent inertia). Split: the essay's statement that the experiments had "ascertained" it in 1909 was an overstatement, since the case then was contested; that is a documented overstatement of the 1909 evidence, not a scale verdict (see QE revision).
• What would settle the remaining dispute (the 1909 data question): reread the Kaufmann plates with modern error analysis; the 2023 HSNS study ("Hazy Spots on Photographic Plates") is the record to obtain.

Claim 2: relativity principle; apparent time

• Agreement: all three: principle Established; ether plus unobservable "true time" Not testable as stated and historically superseded. The Paris-Berlin and 200,000 km/s examples reconstruct correctly [calc].
• Disagreement: none on verdicts. The mathematician and contrarian both note that the essay says "absolute value" while the 1906 memoir calls the postulate testable; the physicist reads this as lecture versus technical paper.
• Stronger argument: the mathematician's and contrarian's reading is a fact of the record (the two texts differ in claimed status); the physicist's "not a contradiction" is also right. They do not conflict: it is a difference of stress. I keep both.
• Final verdict: Established (no experiment detects absolute motion; Michelson-Morley, Kennedy-Thorndike, Ives-Stilwell and modern resonator limits). Ether-frame "true time": Not testable as stated, and Historically superseded but instructive (the Lorentz ether theory makes the same predictions as special relativity; it lost on parsimony, not on a refutation).
• Settle it: nothing measurable can; a detected preferred-frame effect (a Lorentz-violation signal) would be the only test. Modern limits: about 1e-17 for light-speed anisotropy [web via physicist: https://en.wikipedia.org/wiki/Tests_of_special_relativity].

Claim 3: contraction and Michelson

• Agreement: the number is right; "perfectly verified" is overstated; the ether-deformation mechanism is superseded.
• Disagreement: wording of the first verdict. The physicist says "Established (size, null-consistency)"; the mathematician splits into four parts and calls the "Michelson verified" reading "Tested in the weak sense"; the contrarian says "Established as the kinematic effect".
• Stronger argument: the mathematician's four-part split is the most precise. The check on Kennedy-Thorndike [web, item 4] confirms that contraction alone is not enough: with time dilation it is the Lorentz transformation that is tested, not a freestanding contraction. Extra hidden assumption from the mathematician (kept): the 1/200,000,000 assumes the Earth's speed through the ether is its orbital speed; with the Sun's 370 km/s the number would be about 7.6e-7 (about 150 times larger), though by the relativity principle neither is measurable.
• Final verdict: the arithmetic 1/200,000,000 is correct [calc]. Length contraction as part of the Lorentz transformation: Established (tested indirectly, as part of the whole). "Michelson's experiments verify the deformation": overstated, no scale verdict (the null result is compatible with it but does not isolate it; revised, see QE revision). Reading it as a real deformation of matter by the ether: Historically superseded but instructive.
• Settle it: already settled in the standard theory. The decisive pair is Michelson-Morley (contraction) with Kennedy-Thorndike (time dilation) and Ives-Stilwell (direct dilation).

Claim 4: all mass electromagnetic; ether alone inert

• Agreement: velocity dependence is Established (claim 1); the "no constant mass; only the ether is inert" part is rejected; the essay cites Kaufmann for something Kaufmann did not conclude; Poincare's own 1906 memoir needs a non-electromagnetic stress (confirmed in outline by [web, item 3]), which sits badly with "only the ether is inert".
• Disagreement: how to label the strong form. The physicist and mathematician say Historically superseded but instructive; the contrarian says Tested, not supported.
• Stronger argument: the mathematician's point decides: the experiments measure the shape of m(v), not the share of mass that is "constant", so no experiment cited isolates "real mass is null". The contrarian's label overstates the directness of the test. But the contrarian is right that the idea was tested indirectly and lost: Abraham's rigid purely electromagnetic electron fitted worse than Lorentz's in Bucherer's 1908 data [web via Investigator], and the pure-electromagnetic-mass hypothesis is rejected today [web, item 3]. I therefore combine: tested indirectly, not supported, and also superseded.
• Final verdict: velocity- and direction-dependent inertia: Established. "All mass is electromagnetic; the ether alone is inert; Kaufmann shows constant mass is null": Tested, not supported (indirect only; the cited experiment does not discriminate it, and the 4/3 problem needs a non-electromagnetic stress) and Historically superseded but instructive. Kept in the record: part of the idea survives (field energy contributes inertia; most nucleon mass is strong-field energy [memory, high]).
• Settle it: already settled by the Standard Model's mass sources. For the historical question: the Kaufmann 1902-06 papers (his own interpretation) next to Bucherer's discussion.

Claim 5a: Lorentz/Poincare gravity and Mercury

• Agreement: the law gives about 6-7"/century, the right sign, scaling with planetary speed; it explains about one sixth; GR (1915) gives the full 42.98"; the essay is itself modest ("hardly decisive").
• Disagreement: labels. Physicist: Tested, not supported plus historically superseded. Mathematician: Historically superseded (with "direction is right" Established). Contrarian: Tested, not supported plus superseded.
• Stronger argument: the contrarian's and physicist's. De Sitter (1911), Wacker and Lorentz (1910) calculated it independently [web via Investigator: https://en.wikisource.org/wiki/On_the_bearing_of_the_Principle_of_Relativity_on_Gravitational_Astronomy]; the shortfall persisted until 1915. No data were hidden. The mathematician's partial verdict ("the sense is right") is correct and is included below.
• Important distinction I keep (all three note it): "Lorentz's law" is not one number. Lorentz's own 1900 law gives about 1.4"/century and he said it "cannot account for the observed inequality" [web via Investigator: https://en.wikisource.org/wiki/Considerations_on_Gravitation]. The 6-7" belongs to later Lorentz-invariant (Poincare/Minkowski) laws. The essay's 38" is Le Verrier's 1859 figure; Newcomb (1882) gave 43" [web, physicist and contrarian: https://en.wikipedia.org/wiki/Tests_of_general_relativity]; so with the modern value the unexplained part is about 36-37", not 32".
• Final verdict: "direction of the correction is right, and about 6-7"/century": Established (as a calculation reproduced by independent authors). "This law explains Mercury": Tested, not supported (about 16% of 43"/century), and Historically superseded but instructive (first Lorentz-covariant gravity; replaced by GR).
• Settle it: done by ephemerides plus GR. Open historical item: whether the Gottingen French (Sechs Vortrage, 1910) contains the "6"" (see absence of evidence below).

Claim 5b: planets radiate and spiral into the Sun

• Agreement: no derivation is given; the essay's 1e17 years is not reproduced by any mechanism the reviewers can compute; actual gravitational-wave inspiral is about 1e23 years for the Earth (196 W) [calc]; the conclusion "no cause for alarm" holds under either number.
• Disagreement: three different labels. Physicist: Refuted by calculation (neutral bodies do not radiate electromagnetically). Mathematician: Plausible, untested. Contrarian: Not testable as stated (and the EM mechanism unsupported).
• Stronger argument: none of the three is complete. The physicist's refutation assumes the essay means radiation from the planets' ordinary electric charge; but the essay sits inside Lorentz's picture in which gravity is itself a residual electromagnetic force, and the Lille text (Investigator, via fr.wikisource) puts the planetary acceleration at a billion times smaller than at a Crookes-tube anticathode. In that picture the "neutral" objection is not decisive, so "refuted" is too strong for the claim as Poincare meant it. The mathematician's "Plausible, untested" is too generous: the number has no stated basis and the picture behind it (gravity as a residual electromagnetic force) is superseded. The contrarian's label fits the number and the underlying picture best. The physicist is right for the literal reading (ordinary charge on neutral planets).
• Final verdict, split: (i) electromagnetic radiation from the planets' actual (near-zero) net charge: Refuted by calculation. (ii) The essay's specific time of 1e17 years within Lorentz's residual-force gravity: Not testable as stated (no derivation; the Lille ratio cannot be turned into a time without the missing model) and Historically superseded but instructive. (iii) Accelerated masses lose orbital energy by radiation: Established in the general-relativistic form (Hulse-Taylor, LIGO [memory, high]), with a rate about 1e6 times slower than the essay's number (1e21 vs 1e15 centuries). Poincare's idea that accelerated bodies radiate was right in kind and wrong in mechanism and magnitude.
• Settle it: a calculation: write Lorentz's residual-force model with a stated charge asymmetry and compute the dipole/quadrupole loss; the Lille text gives the only hint of Poincare's own estimate.

Claim 5 (finite speed of gravity; reply to Laplace, from the Lille text)

Not in the Monist essay but weighed by the physicist. Plausible, untested in 1909 and historically superseded. Poincare's reply to Laplace (the force-direction aberration argument "proves nothing against Lorentz's theory") is correct in outline [web via Investigator, fr.wikisource]. Kept in the record as a real physics point the essay itself does not carry.

Claim 6: Newtonian mechanics survives at low speed

• Agreement: all three: Established; correspondence limit; gamma - 1 = 4.9e-9 at Earth's speed, 1.3e-8 at Mercury's, 4.3e-15 at 100 km/h [calc].
• Final verdict: Established. The "pedagogic danger" remark is advice, not a testable claim. No settling experiment needed.

Where the reviewers disagreed on the inventor, and my ruling

The contrarian and mathematician both flag overstatements in the author (data "ascertained" in 1909; "perfectly verified"; Kaufmann misread; Mercury speed doubled; "38"" without "per century"). The physicist makes the same points more gently. I keep all of them; they are true. The three also agree there is no sign of blocked funding, hidden data or dismissal without testing: every testable claim was calculated or measured (Mercury within two years). This is not an outsider's rejected theory; most of the essay was absorbed into standard physics, and the rest was replaced by calculation and later experiment.

What would settle it (summary table)

Open questions

1. Whether the "6"" and Bucherer passages were cut by Halsted or the Monist editor, absent from the Gottingen French, or never spoken.
2. What exactly Poincare computed for planetary radiation, and whether he used Lorentz's residual-force gravity.
3. Whether 38" in 1909 (Le Verrier) rather than Newcomb's 43" was a choice or a lag in the literature; the essay says nothing.
4. How decisive the 1908 Bucherer data really were given the 10% off-nominal-speed rays Bestelmeyer raised.
5. Which of Poincare's two gravity laws (the 1905 "mainly compensated" version versus the one de Sitter used) the essay's "Lorentz's law" corresponds to; Walter's study is the lead.
6. The Truman "approximately 1902" Halsted offprint: a different translation, a cataloguer's estimate, or a mistitle.

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

A. The "6"" and the Bucherer passage missing from the Monist text, present at Lille. The Monist text states 38" and 32" but not their difference, and cites Kaufmann but not Bucherer [read; web check item 1]. Competing explanations: 1. The Gottingen French never had them (it was spoken in April, the Lille lecture in August; Poincare may have added the Mercury figure and Bucherer's 1908 data later). For: Lille is a different, longer lecture (3 Aug 1909) and Poincare reworked his lectures. Against: Bucherer's 1908 paper existed before April 1909. 2. Halsted or the Monist editor condensed the text (translation 1913). For: it was a popular journal and translation shortened lectures [memory, low]. Against: nothing found either way. 3. The French printed text differs from the Monist's source (a different French version). Evidence: none found. 4. Poincare omitted them deliberately for a general audience. For: the essay is non-technical and has no equations. Against: it does give the 38" and 32". No suppression is indicated by any record. The deciding record is Sechs Vortrage (1910) pp. 49-58; the Investigator did not reach it.

B. No Halsted, Kaufmann or Bucherer letters in the Poincare Papers calendar. The Investigator searched the correspondence calendar and found Hilbert's and Lorentz's letters arranging the visit, but no letters to Halsted, Kaufmann or Bucherer [web: https://henripoincarepapers.univ-nantes.fr/chp/ via research/2026-10-06-henripoincarepapers.md]; letter texts were not opened. Explanations: the calendar is a partial index; Halsted translated after Poincare's death (1913) with the publisher (Open Court/Carus), so no letter would exist; the exchange was not by letter. Silence here says nothing about the claims. Deciding record: Open Court Publishing and Monist editorial files, and Halsted's papers (the Truman State "Greenwood Papers" holding is a lead).

C. Einstein, Minkowski and FitzGerald are not named anywhere in the essay, and the Lille text three months later is the same on Einstein [read: deep read; web via Investigator]. The 1906 memoir does credit Lorentz and FitzGerald by name for the contraction [web via Investigator]. Explanations: Poincare habitually attributed relativity to Lorentz's school (the "new mechanics" is Lorentz's) and cited Einstein rarely [memory, medium; not checked]; the Lille text gives credit to Lorentz; priority was not a topic of the lecture; or a deliberate omission. For and against: only that the pattern recurs across the two 1909 texts. The record cannot decide between them, and no evidence of motive exists; the omission is stated and left to the reader. Deciding records: Poincare's other 1909-12 papers and letters on relativity, and his remarks in the Gottingen discussion.

D. The relativistic-gravity programme (Lorentz/Poincare/Minkowski/Wacker) goes quiet after 1911. Explanations: GR (1915) explained the full 43", making the 6-7" laws moot; Poincare died in 1912; Minkowski died in 1909. Evidence for each is dated and public [memory, high on dates; web for de Sitter]. Not an anomaly of suppression; the successor theory explains the end of the line.

E. The Truman "approximately 1902" offprint. Not a gap in the physics record; the explanations are listed under the open questions, and the printed volume line decides.

Final verdicts, in brief

Remaining [memory]: none carries a verdict. Corrections made by the arbiter: the physicist's "1e18 years" should read 1e17 years; the physicist's and mathematician's "Established" for claim 1 now carries the web-checked caveat that the 1900s data alone were inconclusive (Zahn-Spees 1938).

QE revision (Step 4b remediation, 2026-10-06)

Inputs: qe.md (verdict MINOR; no deep reads ordered, because the one document was deep-read in full and QE re-checked every passage at issue against the extracted text and page image 6). I re-ran the one new calculation myself. No central verdict reverses; two labels change and one calculation is added. The verdict table and Corrected claim list above were updated in place. Tags as above.

Verdicts that change

Unchanged: claims 2, 4, 5a, 6, the finite-speed-of-gravity item, and parts (ii) and (iii) of 5b.

Calculation for claim 5b(i) [calc, .venv Python; agrees with qe.md]

Give the Earth a generous net charge: the surface charge a fair-weather field of 100 V/m implies (100 V/m is [memory, medium]; the result is robust to many orders of magnitude, so no verdict depends on it). - q = eps0 x 100 V/m x 4 pi R_E^2 = 4.5e5 C (R_E = 6.371e6 m). - Orbital acceleration a = v^2/r = 5.93e-3 m/s^2 (r = 1.496e11 m). - Larmor power (radiated by an accelerated charge) P = q^2 a^2 / (6 pi eps0 c^3) = 1.6e-9 W. - Orbital energy G M m / (2 r) = 2.65e33 J, so infall takes about 5e34 years. That is about 5e17 times the essay's 1e17 years (qe.md rounds to 1e17; my figure is 5.3e17) and about 2e11 times the gravitational-wave time (1e23 years). Ordinary charge on a planet cannot give the essay's number. This covers the literal reading only; in Lorentz's residual-force picture of gravity the claim stays Not testable as stated, as ruled above.

Material QE found missing, now in the record

1. Claim 5: an open question Poincare himself recorded (Lille 1909). He separates mass as "coefficient of inertia" from mass as "coefficient of attraction" and says whether the second grows with velocity "we have no means of deciding" [web: fr.wikisource La_Mecanique_nouvelle_(Lille), via research/2026-10-06-fr-wikisource.md item 6; I did not re-open it]. It is a caution in his favour and anticipates the equivalence principle (inertial = gravitational mass) [memory, medium]. Verdict for the 1909 question: Never properly tested; later tested by Eotvos-type experiments [memory, high; not web-checked; no verdict depends on it].
2. Related work for 5a. Lorentz (1900) says Weber's velocity-dependent force law can account "to some extent" for Mercury's perihelion motion [web: en.wikisource Considerations_on_Gravitation, via research/2026-10-06-en-wikisource.md section 4]. A velocity-dependent gravity law of the same family, set aside by mainstream physics; kept as a lead, not yet researched (sites.md lists it for the next run). It does not change the 1.4"/century figure for Lorentz's own law.
3. Claim 2: Poincare's own worked example of the compensation argument. At Lille, comparing the annual aberration of two opposite stars would reveal the Sun's absolute speed; Poincare is "absolument sur d'avance" no difference will appear, because the graduated circles also deform [web: fr-wikisource, Investigator item 5]. It supports the contrarian's and mathematician's point that the lecture declares the principle certain in advance while the 1906 memoir calls the postulate open to test. No verdict changes: principle Established; ether "true time" Not testable as stated.
4. Leverrier's undiscovered planet and "only a bird" [read: p.10]. Poincare recalls that Leverrier explained Mercury's anomaly by an unseen planet, an amateur thought he saw it cross the Sun, and "it is unhappily certain that this planet perceived was only a bird." (Vulcan and Lescarbault, 1859, are [memory, high], not in the text.) It is the 19th-century ad hoc fix for the same anomaly, instructive beside claim 5a and beside the contraction hypothesis, which also rescued a null result. No separate verdict.
5. Other passages not judged by the reviewers [read: p.7-9]: a 1909 atom model with a "positive central mass" and circulating negative electrons (before Rutherford's nuclear atom of 1911 [memory, high]); X-rays as the "electromagnetic agitation" when cathode rays strike the anticathode (the later bremsstrahlung picture [memory, high]); Lorentz's free-electron account of incandescence, "a swarm of gnats in a jar" (the Drude-Lorentz picture [memory, high]). None is a central claim. As stated in 1909 they are Plausible, untested pictures; in their modern forms they are Established [memory, high; not web-checked].
6. Credit for the transformation (absence-of-evidence item C). Voigt had the transformation in 1887; Minkowski called the rigid electron "a monster"; rival electron models were Abraham's rigid, Lorentz's deformable and Langevin's constant-volume [web via the Investigator, sites.md Leads item 3; not re-opened]. This widens the candidate explanations for the essay not naming Einstein or Minkowski without deciding between them.

Housekeeping and still open

• facts.json still says "about 1910" for the date; the correct record is the opening section of this file and qe.md section 7. A facts.json fix, not a verdict change.
• physicist.md's "1e18 years" is wrong (1e17); corrected above.
• physicist.md's dismissal of Lorentz's charge-asymmetry gravity as "refuted by calculation" ([memory, medium], no calculation) is not adopted. My calculation covers ordinary charge only.
• Records for the next Investigator run, unchanged: Sechs Vortrage (1910) pp. 49-58, the HSNS 2023 study, Walter on Poincare's two gravity laws, Wacker, Weber's law and Mercury, the Truman offprint's volume line.