Arbiter (Step 3): Frank W. Very, The Luminiferous Ether (1919)

Inputs weighed: facts.json, claims.md, deep/ read, physicist.md, mathematician.md, contrarian.md, sites.md and research/ notes. Source tags: [read] project document (PDF page = printed page + 4; text layer re-read by me for pp.19, 28, 36, 38, 40, 42, 44, 46, 50, 52); [read: page image] a rendered page; [web] URL; [memory] with confidence; [computed] my own arithmetic (Python, CGS units, scratch only).

Source gap note

There is no source gap for the central text: the only project document is the 1919 pamphlet (Google Books scan, 63 of 64 pages OCR). The exponents on PDF pp.37-38 are garbled in the text layer, so every reviewer used page images; they agree on the printed values. Be aware of what the verdicts rest on: - Pamphlet text: the reviewers read much of it themselves; the fact-finder's own coverage of pp.1-35 and 43-60 was digest-only. (Updated by the QE revision below: pp.11-35, 42-51 and 51-59 have since been deep-read, so no part of the pamphlet now rests on a digest alone.) - Fessenden's 1900, 1913 and 1914 notes and Very's 1909 paper are not in the project folder. They are known from the Internet Archive text (the Investigator located them; the reviewers read them). I checked the key Fessenden 1900 page myself as a page image (below). - Very's other works (New-Church Review 1913 and Oct 1918, "What becomes of the Light of the Stars?" 1913, the 1920 Nipher paper, "A Possible Limit to Gravitation") were not read by anyone. Conclusions about those are open, not negative. - Parts of the verdicts that rest on [memory] alone (high confidence unless stated): the post-1919 experiments (positron, neutron, Compton, Kennedy-Thorndike, Ives-Stilwell, electron-size bounds, mass-density of the cosmos), Barkla 1911, Trumpler 1930, Taylor 1909. See the web-check table at the end.

Corrected claim list

The claims.md list is kept, with corrections found by the reviewers and confirmed by me.

1. Light is made of discrete electron-sized "ether-particles" travelling at c, each carrying one Planck quantum, not a wave in a continuous medium. A separate universal "aura" fills space. (Correction: Very does not deny a space-filling medium; PDF p.36: "We need not abandon the idea of a universal interstellar atmosphere, but it is not the atmosphere which conveys light" [read]. Three layers: aura, universal; "bound ether", attached to every atom; ether-particles, light.)
2. The Planck quantum is the energy needed to gyroscopically overturn an electron and reverse its sign; the shed ethereal sheath becomes the ether-particle. (Correction: the book does link the quantum to frequency. PDF p.19: the "energy of formation" of the particle "is proportional to the frequency of the luminous vibration, but not to the radiant intensity" [read]; PDF p.40 speaks of "the fixed value of the energy-constant, equivalent to the light-quantum" [read]. Very's "energy-constant" is h, which is indeed fixed, so the two passages are not a plain contradiction, but he never states E = h nu for the overturn. Step 1 and the deep read, which said there is no link, were incomplete.)
3. The electron is a rotating vortical sphere of aura whose surface turns at c. Radius from Millikan, 1.9 x 10^-13 cm; "1.660 x 10^22 rev/s"; surface pressure 10^32-10^33 "dynes" (Fessenden's 2 x 10^32, "revised" by Very to "more nearly 10^33"); required volume elasticity 10^73. (Corrections: see section "Recomputed numbers".)
4. The Michelson-Morley null, the Rayleigh and Brace double-refraction nulls and stellar aberration are accounted for by a shell of bound ether carried with each atom, and so with the Earth, plus corpuscular light. (Correction: aberration is attributed to corpuscular light, not to the bound ether: "entirely independent of the bound ether", PDF p.52 [read, per the contrarian]. Rayleigh and Brace are cited as showing that "the ether concerned in refraction is bound to the earth". QE revision: the pamphlet never states the frame in which c holds; Very also relies on Eichenwald 1903 for an undragged aura (PDF p.23) and on Hirn for ether atmospheres confined near massive bodies (PDF p.35).)
5. There is one kind of electron and no positive electron; all atomic mass is electrical. Hydrogen = 1,844 electrons: a nucleus of 922 electrons of one rotation sense and 921 of the other, plus one satellite (PDF p.28) [read]. Uranium nucleus about 439,000 (238 x 1,844 = 438,872 [computed]), with 92 satellites.
6. Gravity is the synchronous pulsation of electrons in the aura, confined to a "galactic cell" around each galaxy; because neighbouring galaxies would otherwise collide "frequently", the cells must repel at intergalactic distances (PDF p.46) [read]. (QE revision: Very reaches repulsion by inverting Leahy's objection, taking the gravitational quarter-wave as about the radius of an aural cell, PDF p.51.)
7. Matter is gradually converted to radiation and re-originates from the aura, so entropy increase is not one-sided for the universe. Premise: Lodge's dilemma, "if the age of the universe is both ways infinite" (PDF p.42) [read].

Recomputed numbers (where reviewers differed or only one checked)

Claim-by-claim arbitration

For each claim the reviewers' positions (Phys = physicist, Math = mathematician, Con = contrarian) are compared, the stronger argument chosen, the single settling test named and a final verdict given on the Step 2 scale.

Claim 1: light as electron-sized ether-particles at c; the aura

Agree: all three say the particulate side of light (quanta, photoelectric effect, Compton) is Established but belongs to Planck and Einstein, not to Very's mechanism; all three say the strong form has no wave rule for interference, diffraction or polarisation; all three say the aura has no measurable property.

Disagree: the verdict label for the strong form. Phys and Con say Tested, not supported; Math adds Internally inconsistent on the p.52 energy point.

Weighing: the Math inconsistency claim is weaker than stated. Very separates the "energy of formation" (PDF p.19, not "included in the radiant energy which it conveys") from the "radiant energy" that falls as 1/r^2 by "spreading its field of force over a wider area" (PDF p.52) [read]. So the passage describes two quantities; it is undeveloped, not self-contradictory. The stronger points are Math's: Very gives no phase rule for fringes (the book says only that fields "interpenetrate and sometimes interfere", PDF p.59) and does not say whether speed c is relative to the source, the Earth or the aura. Con is right that no test was aimed at the model itself, and that no rejection of the pamphlet is documented. The ballistic reading is excluded by the double-star argument: de Sitter 1913, contested on extinction grounds by Fox in 1965, then replaced by X-ray binary data that bound the source-velocity coefficient k below 2 x 10^-9 (Brecher 1977) [web: https://en.wikipedia.org/wiki/De_Sitter_double_star_experiment]. Phys's figure for Alväger 1964 (pions at 0.99975 c) rests on a secondary web source; I did not open the primary paper.

What would settle it: the model must produce fringe spacing proportional to wavelength from particles of fixed size. A written particle rule giving the two-slit pattern, or a statement of the frame in which the speed is c, would settle whether it is a testable theory at all.

Final verdict (split). Quantised light: Established (by others). Light as sized elastic ether-particles with no wave rule: Tested, not supported (by interference and diffraction results that predate the book and are not addressed in it). Aura as a universal medium with no stated measurable property: Not testable as stated.

Claim 2: the Planck quantum as the energy to overturn an electron and reverse its sign

Agree: no equation; the "sign reversal" has no support (charge is conserved in atomic emission; sign reversal with pair physics begins at 1.02 MeV); h is an action, not an energy, so "least quantity of energy" is a slip; Fessenden's 1914 note supplies the missing "work proportional to angular velocity, so to frequency" step [web: https://archive.org/stream/jstor-1640236/1640236_djvu.txt, read by Math and Con].

Disagree: Phys calls the claim "Refuted by calculation in its only numerical reading" (69 MeV). Con and Math call the fixed-energy reading Internally inconsistent. Math notes the idea is dimensionally respectable (E ~ L x Omega), but Very's electron has an L about 500 times too small to be hbar.

Weighing: the 69/104 MeV reading is a construction by the reviewers; Very does not equate the quantum with h times the surface rotation rate and says explicitly that the radiant period depends on "some function of the radius of the electronic orbit" (PDF pp.39-40) [read]. So I do not give "Refuted by calculation" to Very's claim. I do record that if the idea were pinned to the only frequency he supplies, it fails by 10^4 to 10^7. I also drop "Internally inconsistent": the p.19 and p.40 passages can be read together (fixed constant h, energy of formation proportional to frequency). What remains is that Very never writes the equation that would make the overturn energy equal h nu, and Fessenden, whose gyroscope body rotates at the radiated frequency, is not Very's rotating electron whose rate is fixed.

What would settle it: derive, from the gyroscope picture with the stated angular momentum, an energy proportional to frequency with the observed constant (this needs L = hbar-scale; Very's electron gives about 1/500 of that), and show an observable (for example, a connection between photon emission and a charge-sign change, absent in all measured atomic transitions).

Final verdict: Not testable as stated (qualitative, no stated measurable consequence). The sign-reversal mechanism is contradicted by charge conservation in atomic emission (Tested, not supported, by every emission experiment). The frequency-proportional form is the Planck-Einstein relation (Established, but not derived here).

Claim 3: the rotating vortical electron

Agree: the printed rate is wrong (2.512e22); "dynes" for a stress and the diameter/radius slip originate in Fessenden; the "revised 10^33" goes the wrong way; Very's volume-elasticity formula fails the unit check and mixes radii; constant surface speed c on a sphere requires omega = c/(r cos lambda), which diverges at the poles (the mathematician adds that a nowhere-zero tangent field on a sphere cannot exist, the "hairy ball theorem"; correct, I accept it); the rotating-c electron carries 1/150 to 1/250 of the measured spin; no electron size has been measured, the bound being about 10^-22 m, versus 1.9e-15 m.

Disagree: (i) origin of 10^73 (settled above: Fessenden's); (ii) the final label. Phys: Internally inconsistent and Refuted by calculation (spin). Math: Internally inconsistent; the Fessenden estimate itself is dimensionally consistent. Con: numbers Internally inconsistent, the idea Never properly tested, the size Tested, not supported.

Weighing: Math's separation of Fessenden's estimate from Very's formula is the strongest and most useful finding: Fessenden's constants cohere (rigidity/density gives c, bulk modulus/density gives 5e36 cm/s, so the compressional wave is about 10^26 c), so what fails is Very's arithmetic and units, not necessarily the existence of a coherent (if exotic, untested) model. On spin Phys and Con disagree on the label; Con's caution is right (spin postdates 1919, so it is a later test, not a flaw Very could have seen), and Phys's calculation is right as a test of the model today. Con's note that the rotational energy (0.2 m c^2) disturbs the Thomson mass bookkeeping is an additional, uncontested check. Sign of the stress (Math): a charged surface is pulled outward by its own field, while Very says it would "collapse"; this is unresolved in the text (Fessenden's "tension and pressure" and the "electric condensation" of the aura may be the reason), and I flag it rather than decide.

What would settle it: (1) a single re-derivation of the elasticity with one radius and stated units (Math's P^2 V/2U route is one candidate); (2) the second scan (https://archive.org/details/luminiferousethe00veryrich) to confirm that 1.660 is printed in a second copy, and the Oct 1918 New-Church Review printing to see whether the figure and the radius differ there; (3) any scattering or g-2 result showing electron structure at 10^-13 cm (none exists; the bound is far smaller).

Final verdict (split). As written, the quantitative chain (1.660e22; units; the 1.5-factor and radius mismatches; constant surface speed on a sphere): Internally inconsistent. As a model of the real electron (size 1.9e-13 cm, spin): Refuted by calculation (spin deficit 150-250; no size down to about 10^-22 m), noting that spin was unknown in 1919. The underlying idea of an extended electron held together by an elastic medium: Historically superseded but instructive (the Abraham-Lorentz-Fessenden line), and Fessenden's breaking-strain estimate: Never properly tested.

Claim 4: bound-ether shell, corpuscular light, Michelson-Morley, Rayleigh/Brace and aberration

Agree: the null results are real and Established; no test of the shell hypothesis as distinct from Lorentz-FitzGerald contraction or relativity is on record; the book does not mention Fizeau, Airy, de Sitter or Sagnac (Math searched the text, no hits; also no mention of relativity).

Disagree: Phys: Historically superseded but instructive, full-drag Tested, not supported. Con: Never properly tested; ballistic aberration Tested, not supported. Math: the explanation Tested, not supported, because under either reading of "speed c" it fails (Reading A: 0.37 fringe predicted; Reading B: emission theory excluded).

Weighing: Math's dilemma is correct as logic but Very's text does not commit to either reading; he says the bound ether "is not the ether which conveys light from outside sources" (PDF p.36) [read], so the shell cannot by itself remove a wind from the light path, which is Math's Reading A failure, and nothing in the book derives the null. That makes "Tested, not supported" too strong for an undetermined mechanism, and Con's "no distinct prediction" fits better. On evidence against full drag: Michelson-Gale-Pearson (1925) detected the Earth's rotation by a Sagnac effect (measured 230 parts per 1000 against 237 predicted), which directly contradicts complete drag of the ether by the Earth [web: https://en.wikipedia.org/wiki/Michelson%E2%80%93Gale%E2%80%93Pearson_experiment]. I give Sagnac 1913 less weight than Phys and Con did: the Sagnac effect "could not prove this type of aether wind" against a universal aether, and von Laue showed in 1911 that it is consistent with special relativity [web: https://en.wikipedia.org/wiki/Sagnac_effect]. So the argument against the shell rests on Michelson-Gale for rotation and on the aberration/Fizeau record for translation. Fizeau 1851 [memory, high] is the cleanest test against zero drag by moving matter, which Very's p.53 statement ("independently of its state of relative motion") implies. Con's strongest honest point stands: Dayton Miller's Mt Wilson result is the one respectable positive on record; reanalysis by Shankland et al. 1955 attributed it to statistics and temperature [web via Con: https://en.wikipedia.org/wiki/Dayton_Miller]; this is a critic's reading of the data and I record it as such. I did not recheck Miller's data.

What would settle it: Very's text would have to state the frame in which light moves at c and show, for a stated interferometer, the predicted fringe shift. Experimentally, the existing record already tests the strong forms: Kennedy-Thorndike (1932), Ives-Stilwell (1938) and later tests of Lorentz invariance [memory, high; not rechecked], and Michelson-Gale for drag.

Final verdict (split). The nulls: Established. The shell hypothesis as Very states it (no frame, no formula, no prediction): Never properly tested (it also fails to derive the null; it is not shown to explain it). Its full-drag form: Tested, not supported (Michelson-Gale; Fizeau). Its emission-theory (source-dependent) form: Tested, not supported (de Sitter 1913, Brecher 1977). As a reading of 1900-1920 evidence close to Stokes and Lodge: Historically superseded but instructive.

Claim 5: one kind of electron; all mass electrical

Agree: all three give the same verdict family. The counting closes (922 + 921 + 1 = 1,844; the satellite rings 6+11+15+18+20+22 = 92); nuclear sizes, the positron (Anderson 1932), the neutron (1932) and quark structure rule it out; the idea is shared with Fessenden (1900: about 1,000 corpuscles in hydrogen; 1914: nucleus "can be formed of a number of negative electrons"); and the all-electromagnetic-mass program was serious in 1901-1915.

Disagree: Phys: Tested, not supported; Math: Refuted by calculation on size and binding energy; Con: Tested, not supported, plus Barkla 1911 (an electron count about half the atomic weight) as tension available in 1918.

Weighing: Math's packing numbers are robust: a cluster of 1,844 electrons of radius 1.9e-13 cm cannot be smaller than 28 proton radii; one of 439,000 not smaller than 19 uranium nuclear radii; and the pair-interaction energy is of the order of the rest energy, so "mass equals the sum of electron masses" fails by order one. These use measured nuclear radii (modern) and Rutherford's 1911 nuclear-size inference (available then). Barkla 1911 is [memory, medium], unchecked; it is not needed for the verdict. Con's remark that two opposite orientations of one object echo the Wheeler-Feynman idea of the positron as an electron moving backward in time [memory, high] is a resonance in spirit only; Very's model has no particles of different mass for the proton.

What would settle it: already settled experimentally; the one open item for the reader is to run the packing arithmetic above (done here).

Final verdict: Refuted by calculation (size and binding-energy budget, using measured nuclear radii) and Tested, not supported (positron, neutron, quark structure). For a reader in 1918: Historically superseded but instructive.

Claim 6: pulsation gravity in galactic cells; galactic repulsion

Agree: no equation; the Bjerknes analogy needs perfect synchrony and Very admits no means of producing it; Andromeda approaches the Milky Way (-301 km/s, measured by Slipher in 1912 [web via Con: https://en.wikipedia.org/wiki/Andromeda_Galaxy]), mergers and bound clusters are observed; the only repulsion now accepted is the accelerating expansion, a different thing; his Andromeda distance is far too small though his claim that it is of "truly galactic dimensions" (PDF p.45) was a live question until 1923-24.

Disagree: (i) what the evidence for repulsion is. Phys says Very reads the "relative crowding of the galaxies" as repulsion. The text (PDF p.46) is different: Very argues that without repulsion or a limited gravitational field the fast galaxies would collide often, since neighbours are "only a few centimeters apart" in a millimetre-dot scale. Phys's remark that clustering fits attraction is true but does not answer Very's argument. (ii) Whether the collision argument is "Refuted by calculation" (Con), "fails on its own numbers" (Math), or not mentioned (Phys).

Weighing: both reviewers' arithmetic reproduces (2.1e11 yr; 1.4% in 10 Gyr). The result depends on the finite age, which Very did not use (infinite age is his premise on p.42). So the calculation shows the inference "frequent collisions inevitable" is overstated for a universe of age 1.4e10 yr, not that Very made an arithmetic slip by the standards of his own premise. I therefore do not call it Refuted by calculation; I say it is Not supported by the numbers once the finite age and the actual spacing of field galaxies are used. The observed mergers show that no universal repulsion is needed to avoid them.

What would settle it: a quantified force law with the Newtonian constant G derived from the pulsation (none given); and a galaxy-scale test of a gravity cut-off at about the galactic radius: orbital dynamics of group and cluster members, lensing, and the Local Group timing argument all use attraction beyond one galaxy's extent [memory, high]. The paper Very cites, "On a Possible Limit to Gravitation" (Publ. AAS 3, p.335, per Math), and his 1920 note on Nipher's electrical-gravity experiment, are unread; Nipher's claim is an anomaly claim of the time; whether it was replicated is unknown (not checked; no source either way).

Final verdict (split). Pulsation mechanism: Not testable as stated (no equations, no distinct prediction). Gravity cut off at a galactic cell with galaxy-to-galaxy repulsion: Tested, not supported (mergers, Andromeda blue shift, bound clusters). Collision argument: revised by the QE revision below: valid on Very's 1918 distance scale (collision time about 1e8-1e9 yr); it fails because that scale is about 300 times too small, not because of his arithmetic. (The wording originally here, "weak on the finite-age numbers", and the "d = 80,000 ly (his scale)" input in the weighing above are withdrawn.) Bjerknes analogy: Historically superseded but instructive.

Claim 7: matter-radiation cycle; entropy not one-sided

Agree: no equations; the aura's "non-selective absorption" is unfalsifiable as stated; matter-to-radiation in stars is Established; radiation-to-matter happens as pair creation above 1.02 MeV but not as the book needs; the premise (infinite age, Lodge's dilemma) was replaced by a finite-age expanding universe (Hubble 1929, microwave background 1965 [memory, high]); Steady-State continuous creation (1948) was refuted by the microwave background [memory, high], though Very's proposal is not Steady State; interstellar dust is a real light-absorbing medium (Trumpler 1930 [memory, medium; unchecked]) but is not the aura; the "infinite source of energy which is spiritual" (PDF p.44) [read] is a metaphysical premise.

Disagree: labels only. Phys: Never properly tested (cosmological form Tested, not supported); Math: Not testable as stated; Con: Historically superseded but instructive, re-origination Never properly tested.

Weighing: the mass-accounting check (6.7e-4 solar masses a year converted to light by the Galaxy) shows energy bookkeeping does not forbid the idea, but with no rate or process, it cannot be tested, so Math's label is more exact than Phys's "Never properly tested" (that label presumes a testable claim). Con's label fits the Lodge dilemma, the premise. Part of Very's picture, an interstellar medium that absorbs light, has a real counterpart (dust and gas).

What would settle it: a stated rate of matter appearance per volume (or per absorbed erg) and a place it should be observed; or a bound on a diffuse, non-selective extinction compatible with supernova distances [memory, medium]. Very's 1909 mass equation dm/m0 = v/(V - v) (Science 30:495) [web: https://archive.org/stream/jstor-1635576/1635576_djvu.txt] is dimensionless and rises linearly at small speed (0.11 at v = 0.1 V), where relativity gives 0.005; it concerns mass added by radiation and the OCR is partly garbled (Math), so it is recorded but not weighed. His 1913 Popular Science Monthly paper "What becomes of the Light of the Stars?" is the fuller treatment and was not read.

Final verdict (split). Matter converting to radiation in stars: Established. Re-origination of matter from the aura and the claim about entropy: Not testable as stated. The infinite-age premise: Historically superseded but instructive. A continuous-creation version of the idea: related to Steady State, Tested, not supported (microwave background), but that is a different theory.

Summary of final verdicts

What would settle it (single items)

1. Claim 1/4: a written rule from which a fixed-size particle stream gives fringes proportional to wavelength, and the frame in which speed is c.
2. Claim 2: a derivation of E = h nu from the overturn picture with the stated angular momentum; a measured link between photon emission and sign change.
3. Claim 3: the second scan and Oct 1918 New-Church Review printing, to learn whether "1.660e22" and "10^33" are misprints or Very's; a single-radius, unit-consistent re-derivation of K.
4. Claim 5: nothing further is needed beyond the nuclear-size arithmetic; done.
5. Claim 6: a force law with G derived; a test of a gravity cut-off at the galactic edge (already contradicted by group dynamics).
6. Claim 7: a rate of matter creation from absorbed light.

Open questions

• Is the sign question for the electron's "collapse" (stress outward or inward) resolved in Fessenden 1900 or Very 1909? The mathematician flags it; nobody settled it.
• Does the Oct 1918 New-Church Review printing differ from the 1919 pamphlet in the numbers of claim 3? ("includes much additional material", per the bookseller note in research/.)
• What is the full content of Very's 1913 New-Church Review "A Scientific Estimate of Swedenborg's 'Principia'" (attribution by an AbeBooks listing only), 1913 "What becomes of the Light of the Stars?", "A Possible Limit to Gravitation" and the 1920 Nipher note? Each bears on claim 6 or 7 and none was read.
• Did anyone engage the pamphlet in print? See next section.
• Fessenden's claims of how his 1890-91 theory was received (withheld about a year, then printed with an editorial calling it "probably wrong") are documented in his own words in Science 12:325 (1900) [web: https://archive.org/stream/jstor-1629562/1629562_djvu.txt]. One source, an interested party's account, concerning Fessenden, not Very. Kept here.

Absence-of-evidence analysis: where the record goes dark

What is missing: any contemporary review, reply or citation of the 1919 pamphlet; the funding and fate of the Westwood Astrophysical Observatory after the patron Percival Lowell died in 1916; the whereabouts of Very's papers after his death in 1927; the contents of his 1913 and 1918 New-Church Review papers and the 1920 Nipher paper. The Investigator's search was brief (about 4-5 minutes of the 10-minute cap; HathiTrust, Google Books, the Pitt Allegheny Observatory file, the Lowell Observatory archives, Boston newspapers, and the Westwood Historical Society were not reached) and it found no review. Con ruled that "dismissed without testing" cannot be asserted; I agree, because no dismissal is documented either. The accurate state is: not engaged in the record located so far. Silence in the sources searched is not contradiction.

Competing explanations, with evidence for and against:

My reading: (a), (b) and (c) together explain the silence without needing suppression, and the evidence for (f) concerning this pamphlet is nil; the Langley item and Fessenden's account are context on how outsiders were treated, not on this book. The decisive records are the ones in the "record that would decide" column; they are the targets for the next Investigator run, together with the second scan of the book and the two New-Church Review printings.

Evidence kept in the record next to the verdicts

• Very, Science 30:491 (1909): first statement of claims 3, 4 and 7, with matter as "vacuities in an exceedingly dense ether, kept from collapsing by the centrifugal pressure of a rapidly rotating shell"; the dm/m0 = v/(V - v) equation; support cited: Kaufmann's mass-velocity data, Lebedew and Nichols-Hull light pressure (real measurements) [web, Investigator].
• Fessenden, Science 12:740 (1900): the numbers Very borrows; his stretched-glass vs Kerr comparison (30 x 10^6 against 25 x 10^6 dynes), which he calls "probably accidental"; and the proposed test that gases should weigh slightly more than solids (tested by torsion balances, composition independence far smaller; [memory, high], Fessenden's claim, not Very's).
• Fessenden 1913 ("every continuous medium theory must involve quanta") and 1914 ("Gyroscopic Quanta"): a vortex route to quantisation parallel to Very's, in print before the Planck-based picture was settled.
• Very's 1910 exposure of the Norwood "meteorite" as a fraud after he first reported the fall, and his statement that he "could not verify" a quotation (PDF p.42): evidence on how he handled data, offered by the Investigator and Con as character context.
• The Swedenborgian and devotional statements (PDF p.44 "infinite source of energy which is spiritual"; PDF p.49 "This is the finger of God", Exodus viii, 19; PDF p.55 telepathy) are in the record as Very's stated reason for the aura. The p.55 passage is, in his words, "left undecided": an earthly atmosphere of free ether-particles at rest relative to the earth "is perhaps the physical basis of exterior thought ... and especially of that telepathic thought which is received by harmoniously attuned instruments" [read: p.55]. It is hypothetical, with no experiment or mechanism offered.
• Strongest honest case for Very (Con): he says in 1918 that discrete light quanta make a continuous luminiferous medium unnecessary; he asks what holds an extended electron together; he argued that Andromeda is of "truly galactic dimensions"; an interstellar light-absorbing medium (dust, gas) has a real counterpart. Strongest case against: the only quantitative core (claim 3) fails its own arithmetic and units; every other claim has no prediction; the "aura" absorbs any anomaly.

Web verification

Budget about 8 calls; 4 used (plus a page-image check of Fessenden 1900).

QE revision (Step 4b remediation, 2026-10-04)

Basis: qe.md (verdict MINOR) and the three new deep reads appended to deep/The-Luminiferous-Ether-eb1408.md (PDF pp.42-51, 11-35, 51-59). Tags as above. I did not use new web calls: every change below rests on the pamphlet text as read by the deep readers [read], on arithmetic I checked [computed], or on [memory, high] facts already in the record. No headline verdict changes. Two sub-verdicts and several citations and labels change; they are listed here and the original text above is marked where it is superseded.

1. Corrections to my own and the reviewers' statements

2. New facts from the deep reads, weighed

Frame for c (claim 4). The deep read of pp.51-59 plus a whole-text search finds no passage that says what c is measured relative to. Free ether-particles move "with the speed of light" or "with the velocity of light" [read: p.52, 55, 59]; the only named rest frame is for a hypothetical class of slower particles "at rest in respect to the earth" (p.55), which are by his own description "not like ordinary ether". So neither an emission theory (source frame) nor a rest-aura theory can be attributed to Very. This confirms the mathematician's dilemma as logic (Reading A fails by 0.37 fringe; Reading B is the emission theory, excluded by de Sitter 1913 and Brecher 1977) and confirms my earlier call that "Tested, not supported" is too strong for an undetermined mechanism. Claim 4 stays Never properly tested for the shell hypothesis as stated.

Zero-drag statement (claim 4). PDF p.53, verbatim: the bound ether "whether quiescent or moving (and independently of its state of relative motion or rest) does exert a retarding influence on the passing radiant ether, but one which is determined by the density and electro-magnetic state, and not by the motion of the bound ether". On p.52 the moving attached ether "does not carry with it, control, or push aside" radiant ether. Read literally, this is zero first-order drag of light by a moving transparent body. Fizeau 1851 measured a partial drag, the Fresnel coefficient (1 - 1/n^2) [memory, high; not rechecked here], so that literal reading is Tested, not supported. The book never mentions Fizeau, Fresnel, Airy or Sagnac (checked in the deep read and by grep); whether Very would accept the literal reading is not stated. I record both: the text, and the test it meets.

Eichenwald 1903 (claim 4, kept evidence). Very's one experimental reason for an undragged aura: Eichenwald repeated Rowland's experiment with a charged, metal-coated glass disc rotated; "the magnetic effect was found, which proves that the aura is immobile in respect to the motion of matter through it" [read: p.23]. Very gives no field values or journal reference; neither deep reader nor I checked the original [memory, low on details]. It supports the reading in which the aura is a rest medium while bound ether travels with matter, which is the mathematician's Reading A (and so exposes the 0.37-fringe problem), but Very never says light's speed is c relative to the aura.

Hirn (claim 4 and claim 1). Very reports Hirn as showing that "the large material mass assigned to an interplanetary medium by Siemens and others, is an impossibility", then goes further: an ether with the needed mass "does not fill all space in the form of a continuous medium", and ether atmospheres are "strictly limited to the immediate vicinity of massive bodies" [read: p.35]. The deep reader notes the second step goes beyond what he reports Hirn showing. Hirn's computation is a resisting-medium test of planetary motion, a genuine kind of observation, and Hirn's actual work is unread by anyone here. Its place in the structure: it is Very's reason why a light-bearing ether cannot be a space-filling massive medium. It was dropped from claims.md; I add it to claim 4/claim 1 as a supporting argument, with the limitation above.

Leahy inversion (claim 6). Leahy (1889) objected that a pulsation theory implies repulsion beyond a quarter wavelength; Very turns this into a prediction by setting the gravitational quarter-wave to about the radius of an aural cell, with "evidence" for intergalactic repulsion being the crowding and speed of the galaxies [read: p.51]. The deep reader adds two consequences Very does not draw: gravity should change sign again at 3 lambda/4 and beyond (Leahy's "certain distances" is plural), and a finite-wavelength wave implies a finite speed, whereas Very calls gravity's speed "enormous" (p.50). Very's "reflected back upon itself at the remote boundary of the cell" option is his way of keeping a single sign change. He also admits there is "no suggestion ... as to any means" by which the synchrony Bjerknes' spheres require (negative gravitation being unknown) could be brought about (p.49) [read]. Nothing here changes the claim 6 verdicts: mechanism Not testable as stated; cell cut-off and repulsion Tested, not supported.

Entropy and Shapley (claim 7, kept evidence). Very states the second law fairly, quotes Lodge's dilemma, and gives his cycle (matter becomes radiation; the aura absorbs it and re-originates matter, so the building-up balances the tearing-down) [read: p.42-43]. The footnote (p.43) accepts Shapley's 1915 M13 result that there is no selective (colour-dependent) absorption in space, claiming he had reached it independently, but denies that this shows grey absorption is negligible, citing his own white-nebula study (AN 4536). The deep reader finds the logic sound: a colour test cannot detect grey absorption. Background [memory, high]: Trumpler (1930) found interstellar extinction, mostly selective reddening in the galactic plane, which is dust, and M13 lies off the plane. So Very's logical point stands; his physical attribution to "the aura" does not. The label for claim 7 (Not testable as stated; premise Historically superseded but instructive) is unchanged; "interstellar dust is a real counterpart" is strengthened by this passage.

Firefly spectro-bolometry (kept evidence). The one measurement in the pamphlet made by Very himself: Langley and Very, "The Cheapest Form of Light", Am. J. Sci. XL, p.97, Aug. 1890, cited at PDF p.21, found a broad short-wavelength band that "if it were thermally produced would correspond to a temperature of several thousand degrees" emitted at body temperature "without the corresponding infra-red rays" [read: p.21]. Cold light from a living source is mainstream [memory, high], and the observation is real. Very uses it by analogy (luminescence from satellite electrons, not thermal nuclear vibration), not as a test of the electron model. It also bears on the Langley credit-omission item above: the same 1890 work.

Thornton's cube rule (claim 5, checkable). Very states that H, B, Al, Cu, Ru and Os, "the densest atoms of their respective thermal periods", have atomic weights proportional to the cubes of 1 to 6 [read: p.32], and uses six as the maximum number of electron "stories" (twelve shells; uranium's six satellite rings 6+11+15+18+20+22 = 92). It is unclear whether the sentence is Thornton's or Very's. Check against modern atomic weights [memory, high]: weight/n^3 = 1.008, 1.35, 1.00, 0.99, 0.81, 0.88 for H, B, Al, Cu, Ru, Os (B is 35% above its cube, Ru 19% and Os 12% below). Good for Al and Cu, poor for the rest. (The deep reader's "20-25%" for Ru and Os is slightly high for Os; QE's figures agree with mine.) This does not move claim 5 (already Refuted by calculation on size and binding energy), but it is a numerical claim the reader can test.

Other numbers kept (checked by the deep readers). The 6+11+15+18+20+22 = 92 sum, 922 + 921 + 1 = 1,844 and 238 x 1,844 = 438,872 are correct [read: p.24, 28; computed]. Lodge's "million million" steel rods of 30 ft diameter is self-consistent: about 540 MPa, roughly steel's breaking stress [computed by the deep reader]. Hydrogen satellite frequency 3.29e15 per second matches the Lyman limit [memory, high]. Very's "sun's equatorial rotation 4 1/3 times the earth's and 1/6 of Jupiter's" (p.55) is consistent with modern speeds near 2, 0.465 and 12.6 km/s [memory, medium]; it is the basis of his rejection of Swedenborg's solar and planetary vortices.

New slip in the star-spacing figure (not a headline). PDF p.46, from the page image: nearest-star dot spacing "(92.9 x 206,265)/10^6, or about 192 km" for 1 mm per 10^6 km. The expression evaluates to 19.2, not 192, and 92.9 is the sun's distance in millions of miles, mixed with a km scale. Consistently, 1 parsec at that scale is about 31 km (41 km for alpha Centauri). His "almost 20 km" at 1,000 times the density follows from 192, so the slip lies in the expression, not the later number. His conclusion that star-star collisions are negligible survives on any of these values. Recorded as an arithmetic and unit slip by Very; the galaxy-spacing argument does not use it.

Material in the pamphlet that no earlier file mentioned (kept as context, no verdict). Lenard 1902 (photoelectron energy independent of intensity, the reason for a "structural" energy); Bragg's corpuscular X-rays, reinterpreted by Very (the electron is not converted into the quantum; the stripped envelope is); Millikan's Kepler-law derivation of the Rydberg constant to one part in a thousand, used to reject Marion Erwin's 1,732-electron hydrogen [read: p.28 fn]; Swan's 1897 resin stress figures offered as evidence of polar inflow and outflow at the electron's poles [read: p.26-27]; Hicks (1898), whose theory Very says "can not accord with actual facts" as stated; the Millikan quote that a fibrous field structure is disproved unless "an enormous number of ether strings" end on one electron, which Very turns into support (p.18-19); Maxwell's ether density, said to contain a small numerical error (p.34-35); d'Auria's 1897 estimate, dismissed because he confined space to about 2,000 light-years and misread Maxwell's density as relative to air rather than water (p.35); Newton's "aetherial spirits" letter (1846 printing), which Very says is allied to his own aura, not to his ether (pp.56-57); Swedenborg's posthumous "three natural atmospheres" (pp.55-56) and R. W. Brown's 1916 critique (p.58); Zoellner's "Transcendental Physics" as "a semblance of experimental evidence" for hyperspace (pp.49-50). Also, Very's own estimate for Nova Persei 1901, 1,149 solar masses (AN 3771), is far above any modern stellar or nova mass [memory, high]; the derivation is not in the pamphlet. His direction of correction for Curtis's nebulae (10,000 to 80,000 ly) was right even though the scale remained about 300 times short.

3. Changed items

4. Still open after this round

• The second scan (archive.org luminiferousethe00veryrich) and the Oct 1918 New-Church Review printing have not been read, so whether "1.660e22" and "10^33" are misprints remains unsettled.
• Not checked on the web, by choice: Fizeau 1851, Taylor 1909 (primer evidence table; reword to "later single-photon interference experiments" or check), Barkla 1911, Trumpler 1930. Each is supporting only.
• Original sources for Hirn, Eichenwald, Thornton, Lenard, Swan, Hicks and Leahy were not read by anyone; they are known as Very reports them.
• Very, "On a Possible Limit to Gravitation" (Publ. AAS III, p.335), "Are the White Nebulae Galaxies?" (AN 4536), "What becomes of the Light of the Stars?" (1913) and the 1920 Nipher note remain unread.
• Investigator targets are in qe.md section 7 (downloads to research/files/, the sites marked not reached).