Mendeleev-D: Physicist review (Step 2)

Reviewer: PhD physicist. Scope: the six claims in claims.md, Mendeleev's essay "An Attempt Towards a Chemical Conception of the Ether" (written October 1902, English translation 1904), the Investigator's leads (sites.md, research/ notes, research/files/), and Stewart (2019, secondary). Tags: [read] from a document, [web] with URL, [memory] with confidence. Source folders were not modified. No source gap: the essay has a full text layer; tables on PDF pp.32 and 34 were read by Step 1 from page images. I re-read the key passages myself in _text/ (pp.50-52, 56-59 of the PDF text, and the Stewart text) and re-did the arithmetic.

0. The mainstream view in one place

In 1902 an "ether" was a hypothetical medium that carries light. Light shows polarization (Malus, Young, Fresnel), so its waves are transverse, and a medium that carries transverse waves needs shear rigidity. A gas has none: it carries only longitudinal (sound) waves, at a speed of the order of its molecules' thermal speed [memory, high]. That was the old puzzle of the elastic-solid ether. Michelson-Morley (1887) found no motion through the medium. Maxwell's equations (1865) give c = 1/sqrt(mu0 eps0) from electric and magnetic constants alone. Special relativity (1905) dropped the medium. Modern experiments test Lorentz invariance far beyond 1887: a rotating-cavity test sets a bound on anisotropy of the speed of light of about 1e-17 [web: https://arxiv.org/pdf/1002.1284, Herrmann et al., Phys. Rev. D 80, 105011 (2009)]. Mainstream physics does still speak of "space with physical qualities" (Einstein, Leiden 1920; Dirac 1951; the quantum vacuum and Higgs field), but none of these is a gas of atoms with a chemical element's mass and a temperature [memory, high; Lente (2019) as read by the Investigator cites Einstein 1922 and Dirac 1951]. Mendeleev wrote before relativity, the Moseley atomic numbers (1913) and the Bohr atom (1913), so his position was reasonable in its time and is superseded now.

Fairness note. Mendeleev himself calls the essay "a series of thoughts" and "a hypothesis", says it may be false and that its "one object is to bring this question to the fore" [read: p.39, 58-59]. He did not claim proof. He wanted the ether treated as a question for chemistry (does it have weight, does it belong in the periodic table). That motive is legitimate; the assessment below concerns the specific answers he gave.

1. Claim-by-claim review

Claim 1: The ether is a real chemical element, a monatomic gas with weight that permeates all bodies and cannot combine

a. Mathematics. The essay's claim here is mostly verbal. The one quantitative input is Kelvin's bound "a cubic metre of ether should weigh about and not less than 1e-16 g", against 90 g for hydrogen [read: p.10]. Mendeleev also says light is "a certain transverse vibration of ether" carried by a gas of "minute atoms in rapid motion", and that the atoms' speed is only "130 times" less than the speed of propagation of those vibrations [read: p.56-57]. That is where the internal difficulty sits. Using his own numbers (x about 9.5e-7, T = 193 K) the atoms' rms speed is 2.25e6 m/s, and 3e8 / 2.25e6 = 133, so his "130 times" is correct arithmetic. But in a monatomic gas the sound speed is sqrt(5kT/3m), about 1.7e6 m/s with those numbers. A wave in that gas travels at about the atoms' speed, not 130 times faster, and it is longitudinal, not transverse. To make a gas wave travel at c, the atoms would need v of about 3e8 m/s, which is his rejected lower limit x = 5.3e-11 [read: p.50-51]. He rejects it because it "would revive the emission theory of light". So the model needs two incompatible speeds. I count that as a self-inconsistency, not a small gap. [calc: this review]

b. Experiment. Ether-wind searches (Michelson-Morley 1887 and every later interferometer and resonator test, down to about 1e-17 [web, as above]) found no preferred medium frame. An ether gas at 2,250 km/s thermal speed also makes no prediction that distinguishes it from empty space, because Mendeleev says ordinary atoms "are not likely to be greatly influenced" by it [read: p.52]. A medium that does not interact with matter cannot be detected by its drag, but then it cannot be the thing light is made of either. Weight. Kelvin's bound, taken as a floor, is 1e-19 kg/m^3. The whole gravitating matter density near the Sun (the Oort limit, about 0.1 solar mass per cubic parsec) is about 6.8e-21 kg/m^3 [memory, high; calc: 0.1 x 1.989e30 kg / (3.0857e16 m)^3]. The ether floor would be about 15 times that, which stellar-dynamics measurements do not allow. The measured solar-wind density at 1 AU, about 5 protons per cm^3, is 8e-21 kg/m^3 [memory, high]. So the "not less than" figure from Kelvin is itself ruled out if the ether has ordinary gravitating mass. I did not verify Kelvin's derivation (not in the project files); I report only what the essay states.

c. Tests. Tested as an optical medium (see b): not supported. No one has tested "an inert gas lighter than hydrogen filling space", because the experiments that tested ether did so for the light-carrying medium. Mendeleev's own proposed route (low-pressure gas behaviour, his 1874 work with Kirpichev) he calls "practically impossible" to push far enough [read: p.11].

d. What would have to change if true. Special relativity and Lorentz invariance would need a preferred frame with a measurable wind; the speed of light would depend on the medium's temperature (c about thermal speed); the periodic table would need elements below atomic number 1, so the nuclear-charge basis of the table (Moseley 1913) would fail. Electromagnetic waves would have to be gas waves with transverse polarization, contradicting fluid mechanics.

Verdict (split): - Strong form (a gas of atoms with chemical-element properties carries light): Refuted by calculation (the gas wave speed is about 1/180 of c with the essay's own numbers and the wave is longitudinal), and Tested, not supported (Michelson-Morley and successors). - "The ether has weight, at least Kelvin's 1e-16 g/m^3": Refuted by calculation against local gravitating-matter density, medium confidence (Oort limit from [memory, high]; Kelvin's figure taken from the essay). - As a 1902 question ("what is the ether made of?"): Historically superseded but instructive.

Claim 2: The periodic system needs a "zero group" of inert gases and a "zero series" holding ether x, plus y (maybe coronium) between x and hydrogen

a. Mathematics. The zero-group part is a sound ordering argument: atomic weights He 4.0, Ne 19.9, Ar 38, Kr 81.8, Xe 128 fall between halogens and alkali metals [read: p.31, page images]. The x and y part rests on extrapolating ratios of neighbouring-period atomic weights: Cl:F 1.86, S:O 2.00, P:N 2.21, Si:C 2.37, Al:B 2.45, Mg:Be 2.67, Na:Li 3.28, Ne:He 4.98, Li:H 6.97 [read: p.35-36]. He concludes He:y is at least 10, so y is at most 0.4, and from Xe:Kr 1.56, Kr:Ar 2.15, Ar:He 9.5 he gets He:x = 23.6, x at most 0.17 [read: p.42]. The ratios do rise, but the extrapolation to "at least 10" is a curve-fit, not a law: Lente's reconstruction (a near-geometric series for y, a second-order parabola for x, from the German text [web: https://link.springer.com/article/10.1007/s40828-019-0092-5]) shows it can be reproduced, which says the fit is arithmetic and not physics. The quantity being fitted (the atomic weight ratio across a period) is not a smooth function in the first rows, since it jumps wherever atomic structure changes. The method has no support for extrapolating below hydrogen.

b. Experiment. Zero group: Established. Ramsay and Rayleigh found argon 1894, helium on Earth 1895, and Ne, Kr, Xe 1898; Errera (5 March 1900) and Ramsay (19 March 1900) put them in a group between halogens and alkali metals (Investigator, from Mendeleev's text; [read: p.21, 33]). The noble gases are now Group 18. Mendeleev's acknowledgement that he first thought argon was N3 and was wrong [read: p.27-28] is to his credit. Zero series (x, y): atomic number is the nuclear charge (Moseley 1913, Bohr 1913) [memory, high]; hydrogen has Z = 1, the lowest. No element has atomic weight below hydrogen's. There is a neutron (Z = 0, Chadwick 1932, heavier than hydrogen) and there are particles much lighter than hydrogen, the electron (mass 5.5e-4 u) and the neutrino (limit below 0.45 eV/c^2, about 4.8e-10 u [web: https://www.kit.edu/kit/english/pi_2025_029_astroparticle-physics-neutrinos-weigh-less-than-0-45-electronvolts.php]). Mendeleev's ether atom, x of about 1e-6 u, lies between those two numbers. That is a numerical coincidence only: electrons and neutrinos are not inert gas atoms, are not a medium for light, and neutrinos were not even conceived until 1930. I note it so the reader can weigh it; I do not count it as support. y as coronium: the green coronal line (printed by Mendeleev as 531.7 nm; the standard value is 530.3 nm [web: https://en.wikipedia.org/wiki/Coronium]) was identified by Grotrian and Edlén (1939-1942) as the forbidden line of Fe XIV, iron with 13 electrons stripped, in a corona at about a million kelvin [web: same URL]. So the coronium hypothesis was tested spectroscopically and not supported.

c. Tests. Zero group: yes, many. y: Edlén et al., see b. x: no test is possible beyond the failure of everything it was supposed to explain.

d. If true. Atomic number would not equal nuclear charge for the lightest elements; the Moseley law, the Bohr model, quantum chemistry's electron-shell filling (which gives the period lengths 2, 8, 8, 18...) would need an extra row below hydrogen. Nothing in quantum mechanics permits it.

Verdict (split): - Zero group of inert gases: Established. - Zero series with ether x: Historically superseded but instructive (the periodic table is now organized by nuclear charge, and no Z below 1 exists). - y = coronium: Tested, not supported (Grotrian, Edlén 1939-1942).

Claim 3: Ether atomic weight about 1e-6 of hydrogen's, speed about 2,250 km/s, from kinetic theory at -80 degC and escape velocities

a. Mathematics. I rechecked the arithmetic and it is internally consistent. - rms speed of H2 at 0 degC: sqrt(3RT/M) = 1838 m/s (essay 1843, from a slightly different H2 density) [calc]. - Formula v = 2191/sqrt(x) at -80 degC, so x = 4.8e6/v^2. Check: 2191 km/s at x = 1e-6 and 2248 km/s at x = 9.5e-7. The printed a = 0.000367 is a misprint for 0.00367 (also confirmed by the Russian text per the Investigator). - Earth escape speed 11.2 km/s gives x < 0.038; Sun 608 km/s gives x at most 1.3e-5; a star of 50 solar masses (radius scaled at solar density) gives 608 x 50^(1/3) = 2,241 km/s, so x of about 9.6e-7; a body that holds c needs (3e8/6.08e5)^3 = 1.2e8 solar masses, x of about 5.3e-11 [calc; all match the essay, p.45-50]. - Step 1 deep read correctly reports the range "0.00000096 to 0.000000000053". One point worth saying plainly: the adopted x of about 1e-6 is the upper edge of his own allowed range (the 50-solar-mass limit), not its middle; the lower edge was dropped by a judgement about the emission theory of light, not by calculation [read: p.50-51].

The arithmetic is not the problem. The premises are. 1. Temperature. Everything scales with the assumed -80 degC, "as a rule -100 and -60" [read: p.44]; the essay gives no measurement of a temperature of interstellar space, and says the temperature is not absolute zero because the space is filled with ether [read: p.44; the Russian text says he doubts the absolute zero, per the Investigator and ru.wikisource]. For fixed speed, x is proportional to T. At the cosmic microwave background temperature, 2.725 K (measured 1965 [memory, high]), x would be about 1.3e-8, not 1e-6 [calc]. Absolute zero and the Kelvin scale are established (thermodynamics, gas thermometry) [memory, high]. So the headline number 1e-6 is an artefact of one assumed temperature. 2. Escape-speed logic. It assumes a body with the Sun's mean density; massive main-sequence stars have lower density and white dwarfs and neutron stars much higher [memory, high]. White-dwarf escape speeds are thousands of km/s, which would trap an ether of this speed in the Sirius companion; the cap argument therefore fails. It also treats "mean speed above escape speed" as the loss condition, whereas for a Maxwell distribution Jeans escape needs a much smaller ratio [memory, high]; the mean-speed cut-off is crude. 3. Thermal equilibrium. A gas that is at -80 degC "its own temperature" and yet does not interact with matter is not in thermal contact with anything, so the temperature has no operational meaning.

b. Experiment. Mean kinetic energy per ether atom at these numbers is 4e-21 J (about 0.026 eV), which is just 1.5 kT [calc]. The de Broglie wavelength at 2.25e6 m/s is about 190 nm [calc], so the gas is classical, and no experiment reaches it. No measurement of an ether density or speed has ever given a nonzero result.

c. Tests. None as stated. The speed of light is not set by any thermal speed; wave speeds in a gas of x = 1e-6 at 193 K would be about 1.7e6 m/s [calc], a factor of about 180 below c.

d. If true. A medium with a definite thermal frame and atoms at 0.75 per cent of c would give an ether-wind effect and dispersion; relativity would be false.

Verdict: Refuted by calculation as a derivation of ether properties (the numbers follow from one assumed temperature and a stellar-density assumption, and the same model cannot reach c). The arithmetic itself is consistent.

Claim 4: Radioactivity, the Curie ZnS fluorescence demonstration and Dewar's low-temperature phosphorescence support the ether

a. Mathematics. Qualitative only. Uranium, thorium and radium (U 239, Th 232, Ra 224 as printed; modern values 238, 232, 226) are "suns" accumulating ether, like a gas dissolving in a liquid [read: p.53-54].

b. Experiment. The Curie demonstration [read: p.55-56]: two flasks joined by a tube with a stopcock, a radium solution in one and zinc sulphide suspended in water in the other; opening the cock makes the ZnS glow. Mendeleev reads it as "an emissive flow of something material". Something material does flow, but it is the radium emanation, a radioactive gas (radon, Z = 86, atomic mass 222) first observed by Dorn in 1900 and characterized by Rutherford and Soddy in 1902-03 [memory, high]; the glow is scintillation from alpha particles hitting ZnS (Crookes' spinthariscope, Royal Society, 15 May 1903 [web: https://lapada.org/art-and-antiques/a-cased-william-crookes-spinthariscope/]). That the ZnS in a separate flask lights up through a tube is expected for a gas that diffuses. [memory, medium for the link between this exact experiment and radon; Mendeleev's text does not name the emanation.] Mendeleev's essay (October 1902) does not mention emanation, the then-new transformation theory (Rutherford and Soddy 1902-03), or alpha particles. Dewar's observation (phosphorescence of paraffin brighter at about -190 degC) [read: p.57-58]: the ordinary explanation is thermal: lower temperature slows the escape of electrons from traps and so changes the brightness and persistence of phosphorescence [memory, medium; I did not verify this on the web]. The essay's ether-condensing-in-cold-paraffin picture adds no measured quantity. The solar-energy remark (ether piling up around massive suns to explain luminosity, added in the Russian text, per the Investigator) is replaced by nuclear fusion (Bethe 1939) [memory, high].

c. Tests. The phenomena are real and reproduced (Curie, Crookes, Rutherford). No one tested the ether interpretation separately.

d. If true. Radioactive decay would have to be an ether-exchange process and not a change of nucleus; the energy balance of alpha and beta decay (measured MeV per decay) would need an ether energy source.

Verdict (split): the phenomena themselves, Established. The ether interpretation: Historically superseded but instructive (1902 knowledge could not yet explain them; Rutherford-Soddy and Crookes did within a year) and Never properly tested as an ether claim.

Claim 5: No single primary substance; atoms are not divisible into electrons or transmutable; only the ether is a unity

a. Mathematics. Argument by definition. Atoms of chemistry are individuals with mass (Lavoisier, Dalton) and so division into electrons is "metachemical" [read: p.13-15, 25]. The reasoning is circular: it defines the atom as indivisible and then rejects division.

b. Experiment. Electrons were identified in 1897 (Thomson) as a component of atoms (cathode rays, photoelectric effect, ionization) [memory, high]. Atomic nuclei were found in 1911 (Rutherford) and disintegrate (Rutherford and Soddy 1902-03; artificial transmutation 1919) [memory, high]. Isotopes (Soddy 1913) show that atomic weight is not fixed per element. So atoms are divisible and transmutable. By contrast, Mendeleev's dismissals of Emmens (silver to gold) and Fittica (phosphorus to arsenic, 1900) [read: p.17] concern chemical-lab claims which, in my memory, did not stand up [memory, medium; I did not check]. The "no primary substance" part is not false in a simple way: modern physics has several elementary particles (quarks and leptons), not one, and not "hydrogen" either. Mendeleev's resistance to Prout's hypothesis (all elements built from hydrogen), which Lente calls a personal motive per the Investigator, is partly vindicated in a loose way: nuclei are built from two kinds of nucleon, not from H atoms. I flag that as a nuance, not a rescue.

c. Tests: electron, nucleus, transmutation are all heavily tested and replicated.

d. If true. Atomic physics, radioactivity, nuclear physics would not exist in their present form.

Verdict (split): atoms indivisible and untransmutable: Tested, not supported (and superseded). "No single primary substance": Plausible, untested in the sense Mendeleev meant (modern answer: several elementary particles). "Ether is the unity": same as Claim 1.

Claim 6: Mendeleev's earlier periodic predictions largely succeeded; the ether and coronium were not predictions (Stewart 2019, secondary)

a. Mathematics. The interpolation method is sound: gaps in a table of atomic weights and properties. Stewart's counting: eight successes, one multiple error (the lanthanide "dead zone", five elements of atomic weight 140-175), and two irrelevancies (ether, coronium) [read: digest; Stewart pp.3-9]. Eka-silicon predicted atomic weight 72 and density 5.5; germanium (Winkler 1886) is 72.6 and 5.35; eka-aluminium 68 and 5.9 versus gallium (Lecoq de Boisbaudran 1875) 69.7 and 5.9; eka-boron 44 versus scandium (Nilson 1879) 45 [memory, high].

b. Experiment. The three detailed predictions were confirmed (read: Stewart; [memory, high] for numbers). Of the briefly listed ones, Stewart counts Tc (100), Re (190 vs 186.2), Po (212 vs 209), Fr (220 vs 223), Pa (235 vs 231) as hits [read]. I agree the contrast with the ether is real: x and y have no table-based basis, which Mendeleev himself says in his own words ("I feared to injure the first impression of the periodic law", "could not possibly have been foreseen" [read: p.33]).

c. Tested by discovery of the elements by independent chemists; replicated in the sense that the elements are routinely produced.

d. If true: nothing changes; it is part of the textbook.

Verdict (split): the three detailed predictions: Established. The briefly listed ones: Established for Tc, Re, Po, Fr, Pa as Stewart counts them, with the lanthanide failures real. Ether and coronium as "predictions": Not testable as stated (not predictions from table properties), and see Claim 2 for their separate fate.

2. The Investigator's findings (kept in the record whatever the verdict)

Priority 1, the inventor's own work [web: research notes; sites.md]: - The Russian original (Wikisource, 31 footnotes) and the 1905 St Petersburg edition (Frolova; censor date 7 Aug 1905; scan saved in research/files/). It has material missing from the 1904 English: the name "newtonium" for x ("in honour of the immortal Newton"; I confirmed this in the Wikisource text, line 327 [read]); the ether as possibly a mixture of several limiting gases (I confirmed, line 68 [read]); his doubts about absolute zero (confirmed, lines 358-360 [read]); the "trinity" opening; the idea that ether piling up around massive suns could explain their excess energy; a possible light halogen between H and He. - Physicist's note: the ether-as-mixture idea does not repair the wave-speed problem in Claim 1 (a mixture of gases is still a gas, with no shear rigidity); the "doubts about absolute zero" are mainstream-contradicted (Claim 3). - German translations: Prometheus XV (1903), translated by S. Tschulok, in Nr 735, 737 and 738 (Nr 736 missing); Naturwiss. Rundschau 19 (1904). Per Lente the German alone says x = 0.17 came "from a second-order parabola". A third English printing: Principles of Chemistry, 3rd English ed. (1905), vol. 2, Appendix III, where the preface says that giving the ether "a chemical dress" serves "the cause of unity in natural philosophy". The essay was widely printed in Mendeleev's lifetime and he did not withdraw it in those printings. - Not yet obtained: Основы химии 7th ed. (1902) footnote 68 bis; 8th ed. (1906) pp. 432-433 and supplement 565.

Priority 2, support by others: - Errera (1900) and Ramsay (1900), credited by Mendeleev with the zero group. This supports the zero group, which I rate Established. I.O. Yarkovsky, "The density of the luminiferous ether" (Bryansk 1901), cited by Mendeleev as an independent density estimate; not located, so I can say nothing on its content. - No replication or experimental test of an element lighter than hydrogen was found by the Investigator, and I know of none.

Priority 3, related work: - Lente (2019, open access): reconstruction of the x and y extrapolations; he calls them failures. That is Lente's characterization, and I have stated his reconstruction in Claim 2. - Later "element zero" and sub-hydrogen proposals: nebulium (1898), Antropoff's neutronium (1926) and a new ZAAC paper (June 2026), protofluorine (Habr). Note that the neutron (Chadwick 1932) is a Z = 0 species that exists but is not lighter than hydrogen; whether Antropoff's idea anticipated it is a historical question I did not check. - Modern ether advocates: Atsyukovsky ("General etherodynamics"), who says the zero row "was later removed"; Rodionov, who claims a "genuine 1906 table" with a zero row. These are the advocates' own claims. The Investigator did not obtain their texts and neither did I.

Priority 4, the historical record: "removed from the table" - Competing explanations for why the zero row left the table: (a) the hypothesis was dropped as relativity, Moseley's atomic numbers and the Bohr atom arrived (the view of Lente and ru.wikipedia); (b) the advocates' claim of deliberate removal or falsification. - Physicist's assessment on the evidence available. For (a): Moseley's 1913 result gives a physical reason for the first row to start at hydrogen; Mendeleev died in February 1907 [read: Stewart], before it. For (b): no document has been found that shows a decision to remove it, and the Investigator has not yet seen the 8th edition (1906) or the 9th (1927) tables, which would decide the question. I therefore leave (b) open as unsupported by any record seen so far, not as refuted. Silence about the row in the later editions is an anomaly to be analysed with those editions in hand. What is clear from the sources read: the essay and its table were published in several languages in Mendeleev's lifetime, which is hard to reconcile with a suppression story for the essay itself. - ru.wikipedia says even the 1905 tables lacked x and y; the Investigator found that contradicted by the 1904 English page images (pp.32, 34 per Step 1 [read: page images]) and Habr's photo of the 1905 table. I accept the Investigator's reading. - Publication chain (Oct 1902 to 1958 Kedrov edition) and the Polytechnic Museum holding are recorded in sites.md; they bear on provenance, not on physics.

3. Corrections to Step 1

1. facts.json (Stewart entry) and the "mainstream_conflict" line carry the phrase that the green line was identified as ionised iron "(by Gruenwald)". Stewart's sentence says it was "eventually identified as due to a highly ionised form of iron and named 'coronium' by Gruenwald" [read: _text/mendeleev-predictions-bdcaef.txt]; his footnote 12 is Gruenwald, "On Remarkable Relations Between the Spectrum of Watery Vapour and the Line Spectra of Hydrogen and Oxygen", Chemical News 56 (1887) [read]. That is not an iron identification. The identification of the green line as Fe XIV is by Grotrian and Edlen, 1939-1942, and the name "coronium" dates from 1869 observations by Young and Harkness [web: https://en.wikipedia.org/wiki/Coronium]. Later steps should not credit Gruenwald with the identification. Stewart's own sentence is ambiguous, so this is a reading problem in Stewart as much as in Step 1.
2. The wavelength of the line: Mendeleev prints 531.7 (nm), the standard value is 530.3 nm [web, same URL]. Small, but the essay's figure should not be repeated as the line's wavelength.
3. Deep-read name fixes (from the Investigator, which I checked for one): "Rostovsky" is Rogovsky (the Russian text links to Роговский [read: ru.wikisource text, line 319]). "Kirpitchnikoff" as Kirpichev I take from the Investigator and did not verify myself.
4. The deep read left a question mark on the p.57 sentence about "130 times". I read the page: it says the velocity of the ether atoms "is not very much less (130 times) than that of the propagation of their transverse vibrations" [read: p.57]. It matches 3e8 / 2.25e6 = 133. The sentence is a direct admission that his gas atoms move at 1/130 of the wave speed he assigns to light, which is the inconsistency discussed under Claim 1.
5. The deep read lists the range "0.00000096 to 0.000000000053" correctly but does not say that the adopted value sits at the upper limit set by a 50-solar-mass star. Add that to claim 3 of claims.md.
6. claims.md claim 3 should add that the result depends on the single assumed temperature -80 degC (x scales with T). It says so in part, but not that the value for x would differ by about 70 times at 2.7 K.
7. Items I could confirm: the formulas (I) and (II), the escape velocities, the 120 million solar-mass figure, the 1843 and 2191 constants, the 1e-16 g/m^3 Kelvin figure and the misprint in a. Step 1 was accurate on these.

4. Web checks done (budget 6; 4 used)

• Coronium and Fe XIV (Grotrian, Edlen), retagged [web]: https://en.wikipedia.org/wiki/Coronium.
• Neutrino mass limit (KATRIN, below 0.45 eV): https://www.kit.edu/kit/english/pi_2025_029_astroparticle-physics-neutrinos-weigh-less-than-0-45-electronvolts.php.
• Lorentz invariance bound of 1e-17: https://arxiv.org/pdf/1002.1284.
• Spinthariscope and alpha scintillation on ZnS: https://lapada.org/art-and-antiques/a-cased-william-crookes-spinthariscope/ (this does not itself describe the Curie flask experiment or radon diffusion; the radon explanation is [memory, medium]). Not checked on the web: the Oort limit (memory, high), Dorn and Rutherford-Soddy dates (memory, high), the thermal explanation of Dewar's phosphorescence (memory, medium), Emmens and Fittica (memory, medium). None of the verdicts depends on the medium-confidence items alone: Claims 1 and 3 rest on the essay's own numbers; Claim 4's verdict rests on the essay offering no ether-specific quantity.

5. Summary of verdicts