RankingsPrimerArbiterPhysicistMathematicianContrarianQESites logClaimsSourcesResearch notes

Mills-Randal: central claims

Project source: one document, R. L. Mills, The Grand Unified Theory of Classical Physics (GUT-CP), PDF GUT-CP-2020-Ed-Web.pdf (1,876 pages; printed title page "December 10, 2018 Edition, Volume 1 of 3", yet the PDF holds the chapters of all three volumes and cites a 2022 paper). Author has a commercial stake (BlackLight Power / Brilliant Light Power, hydrino technology); everything below is what the book claims, not a verdict. Tags: [read] = from the document or the deep reads.

Source gap and coverage note

  • No work is unreadable: the text layer is good for prose (50 pages OCR'd), but most displayed equations and many table bodies are garbled or image-only in the text. Every numerical check by a deep reader therefore used page images. Later steps should not trust digest equation text.
  • Deep-read coverage (about 75 percent of pages after the QE round of 2026-10-09 [calc, approximate]): Introduction and front matter (pdf 1-86, evidence statements), Ch.1 (87-158), Ch.2 (159-200), Ch.3-4 (201-247), Ch.5-6 (248-286), Ch.7-8 (288-332), Ch.9 (333-350), Ch.10 (351-444), Ch.11 (445-504), Ch.12-13 (505-640), Ch.14 (641-738), Ch.15 (739-1151, including the functional-group chapters 801-1082, numerical table pages not rendered), Ch.16 hydrino-molecule section (1214-1252), boron/organoaluminium (1344-1400), Ch.24-30 (1521-1580), Ch.31 and first half of Ch.32 (1581-1620), Ch.32 cosmology (1621-1665), Ch.34 (1666-1679), Ch.35 (1680-1703), Ch.36-39 (1704-1746), Ch.40-42 (1747-1804), Appendices I-VI (1805-1874).
  • Digest only (not deep-read): Ch.16 outside pp.1214-1252 (1152-1213: aspirin, RDX, NaH, ice, noble-gas van der Waals), Ch.17-21 and Ch.23 (carbon, alkali halides, metals, silicon, semiconductors, transition metals; pdf 1253-1343 and the part of 1401-1520 outside Ch.22), most of Ch.33 (unification). The numerical table pages of Ch.15 (pdf 804-1085) were not rendered, so molecule counts per class are tallied from prose and symbol tables only.
  • Not in the project folder: the November 1995 edition (the claimed prediction of acceleration and top-quark mass), Hagen and Mills (EPR of molecular hydrino, Research Square 2020 / 2022), Mills's journal papers, all company reports (refs 23-42, 50-54, 56), and the original experimental papers Mills cites. Conclusions about experimental support for hydrinos therefore rest on the author's description of them plus [memory] for the mainstream view until the Investigator finds them.

Central claims

  1. Classical electron, closed-form atoms. The bound electron is a zero-thickness spherical shell of charge and current that obeys Maxwell's equations and Newton's laws and does not radiate; from it Mills derives spin, the Bohr magneton, the g factor (g/2 = 1.0011596521, "eleven-figure agreement", which used 1973-era constants: against modern alpha the formula is 2.87e-10 off, about 2,200 times the experimental uncertainty, and the three-term formula has two coefficients that nearly cancel), Lamb shift, fine and hyperfine structure, and the ionization energies of all 1- to 20-electron atoms, in closed form with integers and fundamental constants only. [read: Ch.1, 2, 7, 10 deep reads]
  2. Hydrinos. Hydrogen has stable states below n = 1 (n = 1/p, radius a_H/p, energy about p^2 x 13.6 eV, p up to 137 in the book; the 2025 Mills-Hagen paper gives p <= 126), reached by a nonradiative resonant energy transfer of m x 27.2 eV to a catalyst (He+, Ar+, K, Li, H2O, or hydrogen itself), releasing [(p+m)^2 - p^2 - 2m] x 13.6 eV; this is Mills's "BlackLight process" and the basis of a new energy source and new hydrogen chemistry. Offered evidence: EUV continua, hydride-ion binding energy (0.75418 eV) and NMR shifts, molecular hydrino H2(1/p), and an EPR signal (g = 2.00445) in a SunCell product measured by W. R. Hagen (Delft) and "replicated by Bruker". The Bruker statement appears twice (the book; Hagen-Mills 2022, which names a Billerica MA laboratory and cites Mills's own preprint), both Mills-co-authored, and no Bruker-authored record has been found; the 2022 paper's own agreement is "0.004 +/- 0.550 gauss over the 11 lines" (mean absolute 0.43 G [calc]), "semi-quantitative", not the book's 0.097 G, and it concedes an approximate conventional two-radical fit and an underlying broad signal of unknown origin. The Introduction states that hydrino states are confirmed "by the data presented in over 100 published journal articles and over 50 independent test reports and articles", with the company website as the only reference (p.37); the evidence refs [22-42] are by Mills and company-linked co-authors, apart from one hosted report (ref 29). Further evidence classes in the book: fast-H / broadened Balmer-alpha lines (refs 36-42) and hydrino-catalysed fusion (HCF), which proposes that deuterium or tritium hydrinos of high p fuse (neutrons, tritium, 3He) and cites detonations of hydrated silver shot with "power densities comparable to those of nuclear weapons [50-54]" (no HCF rate or yield is given). [read: Ch.5-6 pp.265-286, Ch.16, Introduction pp.37, 83-86, Ch.35 p.1689]
  3. Hydrinos as dark matter and cosmic emission. Hydrinos are the dark matter; hydrino disproportionation explains the 3.48 keV line, the 10.1 nm EUV cosmic background, 511 keV emission and gamma-ray bursts, and heating of the warm-hot intergalactic medium. On the 3.48 keV line: the book compares only with Bulbul 2014 and Boyarsky 2014; later Hitomi (Perseus) and XRISM (ten-cluster stack, limit 3-4 times tighter than Hitomi, still about 5 times above the Bulbul rate) and the Dessert-Rodd-Safdi analysis (contested by Abazajian) leave the line unconfirmed rather than excluded at the original flux [web via arbiter; memory medium]. [read: Introduction, Ch.5 pp.265-286, Ch.32]
  4. One method solves molecules and matter. Molecules are prolate-spheroidal molecular orbitals found by the same force balance; bond energies, lengths, angles and vibrations of hundreds of organic, inorganic, boron, organometallic and solid-state systems follow in closed form, with mean error about 0.1-0.2 percent in total bond energy (431 summary rows: mean 0.17 percent) and, in the book's words, "closed-form equations containing fundamental constants only" (no continuous fitted parameter). That statement must be read with the case-specific choices and experimental inputs recorded below: measured atom energies and rounding (Ch.11), a0 versus aH by species, empirical Morse and other imported formulas, and in Ch.15 about 19 E_T values and about 85 freshly specified group choices assigned with verbal justification (the book never says whether any were tuned, and no held-out molecule was found). [read: Ch.11-15, 22 deep reads]
  5. Quantum mechanics is wrong or unnecessary. The Schroedinger charge density "must radiate"; wave-particle duality, Stern-Gerlach, electron diffraction, the Aspect experiment (Bell violations), Schroedinger-cat SQUID/ion experiments and superfluid-helium electron mobilities are all classical; the uncertainty principle and the Born probability interpretation are refuted. [read: Ch.40-42, Appendices I-III deep reads]
  6. Spacetime, gravity, cosmology and particles from the same classical wave theory. Gravity and the Schwarzschild metric come from the classical wave equation with particle production; the universe oscillates (period 9.83e11 yr, H0 = 78.5 km/s/Mpc), accelerates from t = 0 (claimed predicted in 1995), has the CMB peaks of Mills's formula; lepton, nucleon and boson masses, magnetic moments, the X17 particle and the top-quark mass follow from alpha and geometric factors. The wave equation supplies only a light-speed front; the radial factor of the metric is assumed and the result (Eq. 32.38) is the textbook Schwarzschild metric, so the perihelion (Table 32.1) and light-deflection (Table 32.2) agreement does not separate the book from general relativity; the book's new element is that light has zero gravitational mass. H0 = 4 pi c^3/(G m_U) follows from the chosen m_U (78.27 km/s/Mpc for 2e54 kg; 2.32e54 kg would give 67.4); T_CMB needs a different mass (2.56e54 kg), giving H0 = 61.1 [calc]. Sub-claims with their own observational tests: (a) "no transverse time-dependent gravity wave effect in the far field consistent with the absence of the observation of gravity waves" (p.1678; also p.1642), resting on 2009-2010 LIGO and Shawhan references in an edition dated 2018; (b) the Higgs boson "will not be observed in experiments performed at Fermilab, the Large Hadron Collider (LHC), or any future collider" (p.1678, ref: Science 2010), followed by the reading of the CERN 126 GeV boson as "a high-energy neutron resonance" predicted at 128.75 GeV (Eq. 37.47); (c) "absolute space is confirmed experimentally" by the absence of quasar time dilation and the supernova SN 1997ex dilation (p.1678); (d) a free electron has zero gravitational mass (Witteborn and Fairbank 1967: "less than 0.09 m_e"), and a negatively curved "pseudoelectron", formed by photons of at least 44.2 eV near a very dense body, feels a repulsive fifth force (Ch.35; the Investigator found matching patents, 1989, 1994 and 2007). [read: Ch.31-35, 36-39 deep reads]
  7. Collective phenomena are classical. Statistical distributions, superconductivity (Tc from a dipole-array "band-pass filter"), the integral and fractional quantum Hall effect, the Aharonov-Bohm effect, pair production and positronium follow from the same electron model. Deep read (QE round): every chapter takes hbar (or h, or the flux quantum h/2e) as a given and postulates its integers (the dimension f = 1, 2, 3 in Tc; n in the Hall ratio; three more in the fractional case); the "positronium g/2" on p.1578 is the free-electron g factor compared with Van Dyck et al. 1987 (-5.1e-11, 12.75 sigma at the printed uncertainty); positronium levels use 6.795 eV where the book's own formula gives 6.8028 eV. [read: Ch.24-30 deep read, page images]

Orders for further deep reads

  • Done in the QE round (2026-10-09): Ch.24-30, Ch.11, Ch.31 and first half of Ch.32 with Ch.34, Ch.35, Ch.9, Ch.15 pdf 801-1082, Introduction, Ch.3-4.
  • Still open (digest only), for a later update: Ch.17-21 and Ch.23 (solids, semiconductors, transition metals); Ch.16 pdf 1152-1213; Ch.33; the numerical table pages of Ch.15 (render pdf 804-1085 to count compared molecules per class and read each E_osc/E_Kvib); the Investigator's downloaded Jansson report, Booker/Nansteel reviews and Mills's reply to Rathke.
  • Web checks still to do by the reviewers/arbiter: GW150914 and GW170817, Higgs mass, Lewis and Brewer 2023, Fomalont-Sramek, Witteborn-Fairbank follow-up and the Schiff-Barnhill effect, Suarez gas-phase NMR scale and absolute H2 shielding, Ramsey 1950 quadrupole convention, Loch-Stengler-Werth H2+ g factor, PRL 78, 424 (KEK).

Observations by the deep readers that bear on the claims (facts, not verdicts)

  • Inputs vs derived: hbar is a postulated input ("inalienable hbar"); spin 1/2 follows from a chosen hemispherical great-circle construction (Mills 0.495 hbar numerically); the full Bohr magneton needs an extra resonant Larmor photon postulate; N = 3/2 is inserted by hand; the printed field energy is (4 pi)^2 too large in SI. [read: Ch.1]
  • g factor: Eq. 1.228 reproduces to 1.0011596521215 with the book's alpha; against modern alpha it is 2.9e-10 away (about 2,000 sigma); the "eleven-figure agreement" used 1973-era constants with larger uncertainty; no fitted decimals, but discrete modelling choices. [read: Ch.1]
  • Muon g factor: Eq. 26 uses beta^2 = 2/g_mu from the measured g_mu itself, so it returns the input and is not a test. [read: Appendix III]
  • Nonradiation: vanishing depends on lambda_n = r_n (wavelength set equal to radius); for l >= 1 the correct factor is not zero. Stability (self-force) depends on postulates beyond Maxwell's equations. [read: Appendices I-II]
  • Atoms: He ionization 24.58750 vs 24.58741 eV uses a helium-only formula with a borrowed bracket; multi-electron ions show errors up to 4.6 percent with a smooth trend in Z; per-atom integers chosen "to give an energy minimum". [read: Ch.7, 10]
  • Hydride ion: 0.75418 eV agrees with Lykke 0.754195 +/- 0.000019 eV (about 1 sigma) using a special reduced mass; mainstream ab initio H- calculations also agree (memory, medium-high). [read: Ch.5]
  • Hydrino evidence in the book is the author's and company's papers; no independent replication is mentioned on pp.265-286; power examples (100 and 144 kW/m^3) are conditional estimates with efficiency factors set to one. [read: Ch.5 pp.265-286]
  • Molecular and organic energies: arithmetic reproduces; accuracy 0.17 percent mean on 431 rows but 1-3 percent for amino acids, DNA bases, phosphorus compounds, noble-gas van der Waals; whether group energies were tuned cannot be seen in the pages read. [read: Ch.14-16]
  • Cosmology/particles: H0 7-16 percent above measurements; T_CMB 13 percent high; CMB peak positions work only with a 10 Gyr age; top-quark mass 187 vs 172.57 GeV; B-mode r = 0.16 vs r < 0.036; muon/electron ratio matches to 1.6e-8 after field corrections justified in words. [read: Ch.32, 36]

  • Aspect/Bell: the "classical" curve (42.23) is the QM prediction formula of the Aspect paper (deep reader, medium-confidence memory); the book cites no Bell tests after 2007. [read: Ch.42]

  • Ch.3-4: the free-electron charge profile said to be "confirmed experimentally" at KEK (p.205) cites PRL 78, 424 (1997), a TOPAZ/TRISTAN measurement of the momentum-dependent electromagnetic coupling, which its authors report as agreeing with electroweak (vacuum polarisation) predictions; the book gives no alpha(Q) value. hbar is imposed by the boundary condition that the photon angular momentum is hbar (pp.229-232); Compton and photoelectric derivations are the standard photon ones. A free electron absorbing a photon gains 1.6e5 times the photon momentum (Ch.4, "space drive"). [read: Ch.3-4 deep read]

  • Ch.9: arithmetic reproduced for all 61 distinct rows; mean relative error vs NIST about 1.0e-3 (three figures, not "five"); r^2 = 0.999994 is real but plain -13.6/n^2 gives 0.967 with no parameters; the formula is a quantum-defect form with fixed defects 0.144 and 0.289 (observed about 0.145 and 0.305 for singlet and triplet S); the sidebar "strongly supports hydrinos" is the author's analogy, with no hydrino-side test in the chapter. [read: Ch.9 deep read]
  • Ch.11: H2 4.478 eV depends on rounding 27.20 eV and on a0 versus aH; the -28.0 ppm "experimental" NMR value is a composite of -28.5 and +0.48 ppm from two papers; the quadrupole integral omits a surface-element factor (consistent value 0.373e-16 vs printed 0.476e-16 cm^2, both inside Ramsey 0.38 +/- 0.15); the H2(1/p) results need a two-centre integer field acting on the proton repulsion, a Doppler term scaled inconsistently (p^2 or p^3), and an unpaired-electron "doublet" imported from Ch.16 that contradicts the paired-electron force balance. [read: Ch.11 deep read]
  • Ch.15 functional groups: about 19 E_T values and about 85 per-group choices selected with verbal justification; poly-halo groups CCl4, CBr4, CH2I2, CHI3 each have their own E_T and two iodide groups serve one molecule each; the text neither says parameters were tuned nor gives a rule fixing them in advance. [read: Ch.15 deep read]
  • Ch.24-30: all inputs (hbar, integers, dimension f) supplied by hand; Tc formula gives Al 46 K (observed 1.18 K) and Cu 28 K; Eq. 27.8 dimensionally wrong; printed eta = 376.730519 ohm (correct 376.730313); positronium ionization 6.795 eV (formula 6.8028); the p.1578 "g/2" is the free-electron value. [read: Ch.24-30 deep read]
  • Ch.31-34: perihelion and deflection tables are the GR numbers (Mercury 42.98"/century, solar limb 1.751"); the observed columns are 1970s data; "not conservative" versus Fock's form is a coordinate relabelling; gravity waves and the Higgs are predicted absent in text dated 2018 (see claim 6); SN 1997ex dilation is 1 + z, not an ejecta-speed effect. [read: Ch.31-34 deep read]
  • Ch.35: the 9.39e17 kg/m "required M/R" is Earth's ratio, the inequality gives about 1.17e23 kg/m [calc]; Witteborn-Fairbank is an upper limit (< 0.09 m_e); the 2013-14 CERN Courier remark describes shielding fields as the reason electron drops give g near zero; no acceleration, curvature or negative gravitational mass is computed; the Eq. 35.23 energy is -3.56e-4 eV (not 3.56e4). [read: Ch.35 deep read]
  • Introduction: no instrument, sample, run, power figure or spectrum is described in pp.1-86, and no negative result or failed replication is mentioned; the "over 50 independent test reports" sentence cites the company website only; the evidence refs [22-42] print no affiliations. [read: Introduction deep read]
  • NASA TM-107167 (Niedra 1996), for fairness to claim 2: the authors write that explaining all the excess by recombination "requires a recombination efficiency of quite a high magnitude" (above 0.55, rising to about 1 at 5 A) which the data "marginally admit", and quote the maker's documentation of "50 W steady excess heat for a continuous period exceeding hundreds of days" (the company's claim, quoted by NASA). [read: research/files/NASA-TM-Niedra-1996, pp.1, 4]

Corrections to earlier documents

The deep readers found errors in digests (earlier AI-written, secondary). The orchestrator should apply them to the digests if they are kept:

  • digest GUT-CP-2020-Ed-Web-b290db.28.md, p.301: says "Ionization Energy = Electric Energy + (1/Z) Magnetic Energy"; the page image reads "= -Electric Energy - (1/Z) Magnetic Energy". p.307-308: the denominator of Eqs. 7.81, 7.86, 7.87 is 12 m_e a0 (1 + sqrt(s(s+1))); Eq. 7.87 is -(p 29.9 + p^2 1.59e-3) ppm with factor (1 + p alpha^2). p.299: E(electric) in Eq. 7.45 carries a minus sign. [read: page images]
  • digest ...110.md, p.1625: "R_U(t) ~ 1.45e51 [1 - cos(...)]" is wrong: the image reads [1 + cos(...)], and R_U is in W/m^2. Box 32.3 "ratio 0.8627 for r_U/r_g" is wrong: 0.8627 = aleph_0/r_U = 2 pi/(1 + 2 pi). ...111.md, p.1636: the 1995 acceleration claim does have a citation ([101], November 1995 edition, ch. 22). p.1640: "1 Gyr = 3.1358e16 s" is the book's own error (correct 3.156e16 s).
  • digest ...121.md, p.1760: "Bohm [34]'s variant" repeats a misprint; Bohm is [33], and [34] is Aspect et al. 1981.
  • The Introduction digest says the printed edition is 2018 while the file name says 2020; the PDF also cites a 2022 paper (Hagen and Mills, ref 30). Treat the content as post-2018 text of the 2018-dated edition.
  • Deep read file deep/GUT-CP-2020-Ed-Web-b290db.md: the "Corrections" bullet of the pdf pp.1214-1252 section about book p.1193 (neon U(Ne) calculation printed inside the solid-H2 section) was cut off by a later writer; that deep read's remainder should be redone if needed. Deep reads for different page ranges were appended to one shared file by several agents; the pp.1083-1151 entry and its later sections read as intact.
  • Book errors found by deep readers, for the reviewers (not corrections to our files): the muonium experimental uncertainty printed "12 ppm" (should be 12 ppb); the proton magnetic moment printed 2.79268 (correct 2.792847); proton mass printed 1.672648e-27 kg; neutron half-life "12 minutes" (about 10.2); Sun power density 272 W/L (2.72e-4); Table 15 relative errors for guanine and eight other rows do not follow (exp - calc)/exp; "fluorobenzene" labelled C6H5Cl.

Corrections from the QE round (2026-10-09)

  • Earlier text of this file said claim 7 and Ch.11, Ch.15 pdf 801-1082, Ch.9, Ch.3-4, Ch.31-35 were "digest only" or "not assessed": that is now out of date (deep reads done; see the coverage note).
  • Earlier claim 7 text implied a derivation of the fine-structure constant and an Aharonov-Bohm effect "via flux linkage": the deep read shows alpha is identified (Eq. 29.9 is the identity alpha a0 = hbar/(m_e c)) and the Aharonov-Bohm treatment is a damped oscillation with an assumed period h/2e and a free damping factor.
  • Arbiter and contrarian text describing "positronium g/2 = 1.001159652137" as a positronium measurement comparison: the page image (p.1578) shows it is the free-electron/positron g factor re-quoted from Ch.1 Eq. 1.236, compared with Van Dyck, Schwinberg, Dehmelt 1987.
  • Claim 7 Tc "factor 25" for the dimension choice: at fixed n, Tc goes as 1/f^3, so f = 3 to 1 is a factor of exactly 27 (the printed 7 K to 180 K is 25.7 because 7 K is a rounded 6.64 K).
  • Any file giving the Eq. 35.23 energy as "3.56e4 eV" or the required M/R as 9.39e17 kg/m, or Witteborn-Fairbank as having "measured zero", should be corrected to the values in the Observations (-3.56e-4 eV; about 1.17e23 kg/m; "less than 0.09 m_e").
  • The "KEK confirmation" of the free-electron charge distribution is the book's wording; the cited paper measured the running of the electromagnetic coupling.
  • Applied from the arbiter (corrections 1-4, 7): UK appeal [2008] EWHC 2763 (Pat) was allowed and remitted, not dismissed (for physicist.md and sites.md, applied by the orchestrator); claim 2 (p <= 126, Bruker qualifiers), claim 4 (qualified "no fitted parameters"), claim 6 (H0 comparison) and the 3.48 keV note are applied in this file.