Rutherford-Ernest: Arbiter (Step 3)

Arbiter, 2026-10-09. Inputs read: claims.md, physicist.md, mathematician.md, contrarian.md, sites.md (Leads), facts.json status, and the deep reads where the reviewers differ (radio-activity-e1fda6 section 2-3; Radioactive-Substances-f227f1 sections 3.5-3.6). Tags: [read] project documents and deep reads; [web] URL; [calc] my arithmetic (Python/hand); [memory] with confidence.

Source gap (as first written; see the QE revision at the end, which closes limits 1 and 2 below). None for the main works (all seven documents have text; equations were read from page images by the deep readers). Limits that carry into every verdict: (1) 1930 book Ch. IX-XI (artificial disintegration of light elements, attempts to change decay rates, satellite theory) and 1913 book Ch. I-X, XIV-XVII and 1904 book Ch. I-IV, VIII-XI are digest-only; (2) the 1913 deep read of the alpha-ray chapter (PDF pp.131-192) was lost in a file overwrite and has not been re-run, so the project holds no deep-read primary text for "alpha = helium" (the claim rests on the 1904-05 popular articles, the 1911 and 1930 statements, and [memory, high] for Rutherford and Royds 1909; a web summary now supports it, below); (3) the 1919 Nature letter (Rutherford and Compton, gravity) is abstract-only. Conclusions that rest on [memory] alone are marked.

Interested parties. Rutherford wrote every primary document and originated the disintegration theory and the nuclear atom; the experiments that tested these were mostly done by his own students (Soddy, Barnes, Geiger, Marsden, Gates, Bronson, Chadwick, Royds). Chadwick is co-author of the 1930 book that reports his own experiment. Le Bon and Moray-line sources have priority stakes. The New Energy Times article on Blackett has a stake in disputing institutional credit. Independent confirmation exists and is named where it does.

---

Overall judgement in brief

This is not an alternative-science project; it is the founding record of nuclear physics. Six of seven claims are Established in their core. What is instructive is (a) the seams: places where the author's own numbers were wrong or later superseded (alpha charge/energy, N = A/2, the whole-earth radium budget, the sun), (b) a short list of anomalies that were reported, then explained, or never followed up (Curie-Danne temperature effect, Dewar's liquid-hydrogen heat, Bumstead's X-ray heating, Russell-Rossi ionium lines, the Rutherford-Compton gravity test), and (c) the "detonator" passage, which is of particular interest to readers of the Moray and Le Bon projects.

---

Corrected claim list

QE revision (2026-10-09, later in this file) changes items 5, 10, 11 and 12 below and adds the points in its sections 1 to 11. Where an item below disagrees with the QE revision, the QE revision stands.

Corrections to claims.md, in the order a reader will meet them. Each is agreed by at least two reviewers unless stated.

1. Claim 4, "equals the atomic number (79) by Chadwick's 1920 measurements": wrong. Chadwick 1920 measured copper, silver and platinum: 29.3e, 46.3e, 77.4e for Z = 29, 47, 78 (recomputed ratios 1.010, 0.985, 0.992) [read: 1930 book pp.231-232, deep/Radiations]. Gold, within 2 percent of 79, is Rutherford and Chadwick, Phil. Mag. 50, 889 (1925) [read: 1930 book p.234]. Physicist, mathematician and contrarian all found this. The matching text in facts.json for the 1930 book must change.
2. Claim 4, "Coulomb's law holds to about 3e-12 cm": this is an upper bound on the gold nucleus radius, the closest approach at 135 degrees for 7.5 MeV alphas, not a measured scale of the force law. I recomputed: b = 2(79)(1.44 MeV fm)/7.5 MeV = 30.3 fm; closest approach d = (b/2)(1 + cosec 67.5 degrees) = 1.041 b = 31.6 fm = 3.2e-12 cm; copper 1.16e-12, silver 1.9e-12 (book: 3.2, 1.2, 2e-12) [calc]. Modern gold radius is about 7e-13 cm [memory, high].
3. Claim 4, "central charge about 100e in 1911 ... a central charge later called the nucleus": the 1911 paper does not use the word "nucleus", leaves the sign open ("it has not so far been found possible to obtain definite evidence", p.17), and allows satellites; "positive nucleus carrying most of the mass" is the 1913 book's statement (p.644-646) [read: deep/structureatom sections 8-9; 1913 deep read]. Physicist and contrarian agree.
4. Claim 3, "about a tenth of the speed of light": the 1903-05 popular figure only (20,000 miles/s = 3.2e9 cm/s = 0.107c [calc]). The 1911 paper uses 2.09e9 cm/s (0.070c); the 1930 book 1.922e9 cm/s (0.064c). Mathematician's correction accepted.
5. Claim 2, "3.7e9 cal/g": the figure is stated in the 1913 book (p.681) as "the total emission of energy for the complete transformation of one gram of radium". It is not derivable as 100 cal/h times a mean life (that gives about 2.0 to 2.5e9 depending on the mean life used, 2335 or 2900 years [calc]); the 1906 book's figure through radium C is 1.6e9. It is right in size: mass defect of Ra-226 to Pb-206 plus five He-4 is 35.3 MeV per atom = 3.60e9 cal/g; the five alpha energies alone sum to about 29.8 MeV = 3.0e9 cal/g [calc, atomic masses and alpha energies from [memory, high]]. Do not imply the product derivation. [QE revision section 2: the 1913 book derives 2.44e9 cal/g for the emanation chain exactly, and gives 3.7e9 for radium through F as a stated figure, not a step-by-step derivation.]
6. Claim 2, "about 100 cal/h per gram of radium" and "three-quarters from the emanation": the 1904 book's "42 cal/h" is per gram of radium bromide, not radium (the printed 65 cal/h per gram of bromide gives about 110 per gram of radium by the ratio 385/225 [calc]) [read: deep/radio-activity section 5]. Share from emanation and products: about 75 percent in Rutherford and Barnes; the modern alpha-energy share is 80 percent [calc: (5.49+6.00+7.69)/(4.78+5.49+6.00+7.69) = 0.80]. Use 65-80 percent with units.
7. Claim 2 (new item, missing from claims.md): in 1904 Rutherford's own three routes to the alpha power gave 40, 58 and 71 cal/h against about 100 observed [read: deep/radio-activity section 3; physicist and contrarian say "about 50"; the mathematician's three numbers are the reading I use]. He said the alphas "do not account for the whole" and invoked extra "rearrangement" energy. See the recomputation below: the shortfall was bookkeeping, not a missing energy source.
8. Claim 5, "can account for the earth's internal heat": true as a total heat budget only if the radioactivity is concentrated near the surface. The 4.6e-14 g Ra per g uniform through the whole earth is the threshold; crustal rock holds about 2e-12, so uniform crust composition through the whole earth would give about 200 times the conducted flux [read: 1913 book Sec. 260; recomputed below]. Strutt's thin-shell suggestion follows. Also: the 1905 article says "crust" where its own number is a whole-earth average (4.6e-14 [calc]).
9. Claim 5, "about a hundredfold": a scaling of Helmholtz's 40 million years to 5000 million, in the 1905 Harper's article only. The 1913 book says only "much longer" and gives the energy ratio 3.7e9 / 2.7e7 = 137 [read; calc]. A direct estimate gives 2.5e9 years (below).
10. Claim 6: the 1919 paper says the long-range particles are "probably charged atoms of hydrogen or atoms of mass 2" and that the evidence "is not sufficient to distinguish definitely" H from masses 2, 3 or 4 [read: 1919 Science, saved copy]; the residual nucleus was identified (oxygen-17) by Blackett's cloud-chamber photographs of 1925, published after about 23,000 photographs with 8 forks [web: https://en.wikipedia.org/wiki/Patrick_Blackett, via search 2026-10-09; the credit dispute at https://newenergytimes.com/v2/sr/Rutherford-Blackett/Rutherford-Blackett.shtml is an advocacy source with a stake and I did not read it]. Say "nitrogen disintegrated, emitting a hydrogen-like particle; product identified 1925". [QE revision section 5: the "six elements" in the Cambridge-Vienna contrast is wrong; Cambridge found 13.]
11. Claim 7, "no known physical or chemical agent changes" the rate: in the 1904 Bakerian Lecture (pp.198-199, 210) Rutherford accepted and had himself replicated the Curie-Danne finding that heating altered radium C's decay constant (about 1.4 times normal after about 1100 C), "the first case where temperature has been shown to exert an appreciable influence" [read: saved Bakerian PDF; contrarian read the page images]. Bronson 1905 (in Rutherford's lab) and Bumstead 1906 explained it as radium B being more volatile than C. The detonator passage predates the 1906 book. [QE revision section 7 corrects my dating: both the detonator sentence and the gravitation sentence are already in the project's own 1904 first edition, sec. 206, book p.339, read in the deep read; the 1905 second edition (sec. 270) and the St Louis address (September 1904) repeat them.] Claim 7 mixes two claims and is split below.
12. Claim 7, the 1930-list of agents (3000 C, 20,000 atmospheres, ...): not verified by any reviewer (digest-only, Ch. XI). The 1913 book (pp.647-648) supports the general statement. [QE revision section 7: the list is now confirmed word for word from the page image of the 1930 book p.319 (PDF p.352); the book gives no reference for any item.]
13. facts.json: Harper's "Radium the cause of the earth's heat" is February 1905, not March [web: https://harpers.org/archive/1905/02/radium-the-cause-of-the-earths-heat, per contrarian; I did not re-fetch]. The other article is January 1904.
14. Errors in the sources that travel with the verdicts (kept, not hidden): 1911 b = 3.4e-12 cm is about 6 percent above the paper's own constants (3.19e-12); p1 at p2 = 0.1 should be 7.4e-4 not 4e-4 [calc: exp(-7.21)]; three printed z/A^1.5 quotients (tin, copper, iron) and the Crowther silver row do not follow from the printed numbers (source of the error, journal or the web transcription, unknown); 1930 book footnote year 1921 should be 1911; 1913 radium D maximum "81 years" should be about 115; 1913 uranium atoms per gram 5.2e21 against 2.6e21 implied by its own Sec. 63; the 1913 "one-fourth" of solar output for a sun of uranium (below); OCR slips (3300 cal not 8800, 1/132 not 1/182, 3.77 days, 35 minutes, 13.6e10 not 136e10).
15. Physicist and contrarian both say "taking the alpha charge as e instead of 2e" explains the low alpha energy. That is incomplete; see recomputation R1.
16. Physicist and contrarian cite "Compton, Phil. Mag. 1920, Radioactivity and the gravitational field". I could not find such a paper by web search [web: search 2026-10-09 returned only the Nature letter, Nature 104, 412, dated 25 Dec 1919 in one index and 1920 in another]. Treat that title as an unverified citation; the Nature letter is the only confirmed record.

---

Recomputations where reviewers differ

R1. Why Rutherford's 1904 alpha energy was low (5.9e-6 erg = 3.7 MeV against a modern 6 to 7.7 MeV). Energy E = (1/2) q v^2 / (q/m). His inputs: ion charge e = 1.13e-20 emu = 3.39e-10 esu (modern 4.80e-10), alpha charge taken as one e (true 2e), v = 2.5e9 cm/s, q/m = 6e3 emu/g [read: deep/radio-activity section 2c, 3]. Factor by factor against the true values for the same radium C' alpha (v = 1.92e9 cm/s, q/m = 4.82e3 emu/g): charge 0.352 (1/2.84), speed squared 1.69, q/m 0.80 (his q/m too high by 1.25). Product 0.476, so the true energy is 5.9e-6 / 0.476 = 1.24e-5 erg = 7.7 MeV [calc], which is radium C' (7.69 MeV). The shortfall comes from four errors that partly cancel (charge multiplicity, size of e, speed, q/m), not from the factor 2 alone; correcting only e to 2e (with his ionic charge) would give 7.4 MeV with a too-fast alpha, and with the correct 2e and e would give 10 MeV with his speed. Result: the shortfall in the heat budget was bookkeeping. His alpha count of method 3 (1.4e11 per second per gram) was within 5 percent of the modern 4 x 3.7e10 = 1.48e11 [calc]; modern alpha power for the chain to radium C' is 24.4 MeV x 3.7e10 x 1.602e-13 = 0.145 W = 125 cal/h, the whole observed heat (observed 110-134). No separate energy source was needed. Verdict on the "1904 factor-two shortfall" as an anomaly: Historically superseded but instructive (I do not adopt the contrarian's "Refuted by calculation"; it was never offered as a claim of new energy, and the explanation is arithmetic of inputs).

R2. Earth heat, 200 times versus 40 times. The earth-heat deep read gives two numbers: radium alone, 2e-12 against the 4.6e-14 threshold, a factor 43 [calc]; and "about 200 times" for the full uranium and thorium heat. Recompute: uranium 0.6e-5 g/g x 7.7e-5 cal/g/h = 4.6e-10; thorium 1.2e-5 x 2.7e-5 = 3.2e-10; total 7.8e-10 cal/g/h = 6.8e-6 cal/g/yr; times earth mass 5.97e27 g = 4.1e22 cal/yr against the surface outflow 39 cal/cm^2/yr x 5.1e18 cm^2 = 2.0e20 cal/yr; ratio 204 [calc]. Both numbers are right; they answer different questions (radium alone, or the whole crustal radioactivity). The book's 200 is reproduced. The 270-million-ton figure reproduces with conductivity 1.7 W/(m K): flux 0.0365 K/m x 1.7 = 0.062 W/m^2; times 5.1e14 m^2 = 3.2e13 W; divided by 0.1162 W/g (100 cal/h) gives 2.7e14 g [calc]. Physicist and mathematician agree; conductivity is my assumption (uncertainty about 1.5). Modern: about 47 TW surface heat flow, about 20 TW radiogenic by geoneutrinos [web: physicist's cited Gando et al., Nature Geoscience 4, 647 (2011), via search summary; not re-opened by me].

R3. The sun. (i) Power: radium needed for L = 3.83e26 W at 0.1162 W/g is 1.66e-6 of the solar mass; Rutherford's "2.5 parts per million" is 1.5 times larger (all three reviewers agree). Duration of radium (half-life 1600 years) would be thousands of years [calc]. (ii) Duration with 3.7e9 cal/g across the whole sun: 3.7e9 x 4.184 J x 1.99e33 g = 3.08e43 J; / 3.83e26 W = 8.1e16 s = 2.6e9 years [calc]. The "hundredfold" is therefore a scaling, not a result; the direct estimate is half of Rutherford's 5e9 years. (iii) The 1913 "one-fourth": a sun of pure uranium in equilibrium at 7.7e-5 cal/g/h gives 1.53e29 cal/h = 1.78e26 W = 0.46 of L [calc]. Unexplained factor 1.8; the conclusion (known radioelements cannot power the sun) holds either way. (iv) Fusion: 0.7 percent of mc^2 = 6.3e11 J/g = 1.5e11 cal/g, about 40 times the radium figure [calc], so Rutherford's guess was too small by that factor and wrong in mechanism.

R4. 1911 scattering numbers. Fraction scattered beyond 90 degrees from a gold foil of 4e-5 cm (n = 5.9e22 per cm^3): (pi/4) n t b^2 with b = 3.19e-12 (his N = 100) gives 1.9e-5 (deep reader; the contrarian's "1 in 53,000"); with modern Z = 79 and 7.69 MeV, b = 2.96e-12 cm and the fraction is 1.6e-5 (mathematician) [calc]. Geiger and Marsden's observed about 1 in 20,000 = 5e-5 is 2.6 to 3 times higher. The paper itself does not do this calculation, and the foil thickness behind "1 in 20,000" is not given in the reviews; same order of magnitude, not a precise test. Rutherford's 2.09e9 cm/s corresponds to 9.1 MeV, higher than any radium alpha (the fastest is 7.69 MeV, 1.93e9 cm/s) [calc]; that explains why b is 25 percent smaller with modern speed and Z (2.5e-12 cm, the contrarian's figure) than 3.4e-12. All reviewers' figures stand once the inputs are named.

R5. Decay-law example. Thorium X: half value 4 days, lambda = ln 2 / 96 h = 0.00722 per hour; recovered fraction after 1 hour 1 - e^(-0.00722) = 1/139 (book 1/140); after 1 day 0.159 (book 1/6); radium B to C maximum at ln(lambda2/lambda1)/(lambda2 - lambda1) = 35 min [read; physicist and mathematician computed the same]. Ra/U = (1600/4.468e9)(226/238) = 3.40e-7, matching Rutherford and Boltwood's 3.4e-7 [calc, half-lives [memory, high]].

---

Claim-by-claim arbitration

Claim 1. Spontaneous disintegration, exponential law, rate unaltered by any known agent

Agree: the law, the chains and the mathematics (Bateman solutions checked by SymPy; two book misprints, neither propagated) are Established; the 1904 Curie-Danne temperature effect was an artifact. Disagree: how to grade "no agent changes the rate". Physicist: Established to about 0.1 percent for alpha and beta, small explained exceptions for electron capture. Mathematician: Established for alpha; refined for electron capture and bound-state decay. Contrarian: strong form "Tested, not supported at the sub-percent level". Stronger argument: the physicist and mathematician. Rutherford's wording was "any known agency" and "not found possible", a statement about evidence. The electron-capture dependence is a real measured effect (Be-7 half-life about 0.8 percent shorter in C60 than in metal; larger pressure effects of 0.5 to 1 percent) [web: https://physicsworld.com/a/radioactivity-speeds-up/ per physicist, search summary only]. It does not refute the claim for the alpha and beta emitters he studied. The extreme case, bare Re-187 ions decaying with a 32.9 +/- 2.0 year half-life against 42 billion years neutral [web: Bosch et al., Phys. Rev. Lett. 77, 5190 (1996), via search 2026-10-09, https://www.osti.gov/biblio/404092; abstracts only], shows how large the effect becomes only when the electron environment is removed entirely. Contested solar-modulation claims (Jenkins, Fischbach and others; disputed by PTB and others) are unresolved; no reviewer read the papers; I give no verdict on them. The 1904 episode counts twice: it is evidence that the invariance claim was tested, and an example of a replicated anomaly (Rutherford replicated Curie and Danne himself) dissolved by a mundane mixed-species artifact (Bronson). What was not found: whether Curie and Danne accepted the explanation. Settle: for modern small effects, a controlled comparison of an alpha emitter and an electron-capture emitter in different environments, run by several laboratories under a pre-agreed protocol with annual-cycle controls (existing literature: Emery 1972; the PTB and Purdue series, none read here). Final verdict: Established (exponential law; chains; invariance under temperature, pressure, chemistry and ordinary radiation for alpha and beta decay). Split: "no agent at all" - Established for alpha decay at the tested precision; refined at the 0.1 to 1 percent level for electron capture, and large for fully ionised ions (not known to Rutherford). The 1904-05 temperature effect on radium C: Tested, not supported (explained). Modern cosmic-modulation claims: Never properly tested by an agreed neutral protocol (contested; not evaluated here).

Claim 2. The energy is the atom's own, about 100 cal/h per gram, store about 3.7e9 cal/g

Agree: Established; size right. The external-energy theories (Crookes 1899, Kelvin 1903, Curies 1902) are reported in the 1913 book and the St Louis address and were answered by experiments: heat inside an ice calorimeter, unchanged under several inches of lead (Runge and Precht, about 5 percent), heat that grows and decays with the emanation, and mass loss in the products. The author is an interested party but the arguments are experimental. Disagree: the contrarian notes that lead shielding cannot exclude every conceivable medium that passes through lead; the physicist treats the question as closed by the mass-difference accounting. Stronger: the physicist, because the later mass defect gives the energy to better than a few percent without any external term (35.3 MeV per atom, 3.60e9 cal/g against 3.7e9 [calc]). I keep the contrarian's point in the record as a limit on what the 1904 experiments alone could show. Dewar's liquid-hydrogen heat, "distinctly greater" (British Association 1903): reported once; Rutherford said it "requires confirmation". The contrarian's check of the printed gas volume (73 cc/min) gives about 42 cal/h, 60 per gram bromide, consistent with 65 and no excess [calc, per contrarian and physicist, who agree]. No confirmation or withdrawal found. Record kept; this is "Never properly tested" in the sense that nobody repeated it, though the printed numbers show no anomaly. Settle: nothing needed for the core. For Dewar, a repeat with an independent calorimeter, or the Dewar 1903 paper's own numbers. Final verdict: Established (internal energy; about 100 to 134 cal/h; 3.6e9 cal/g). The electron-orbit "latent energy" and radiation-loss picture of the store: Historically superseded but instructive. The 1904 factor-of-two gap: Historically superseded but instructive (R1). Dewar excess: Never properly tested (the printed numbers show none).

Claim 3. Alpha = helium, beta = electron, gamma = radiation; helium and lead end products; uranium to ionium to radium

Agree: Established. Early numerical errors (alpha mass "twice hydrogen", speed, radium life "not much more than a thousand years", thorium rate) were later corrected, mostly by Rutherford's own group. Evidence for identity: Ramsay and Soddy 1903 (helium spectrum from the emanation, which Rutherford said "must naturally be accepted with reserve"); Rutherford and Royds: helium spectrum accumulated from alphas in a sealed tube in 1908-09, with neon absent to rule out air leakage [web: https://web.lemoyne.edu/Giunta/royds.html; search summary 2026-10-09, Phil. Mag. 17, 281]. This upgrades the [memory, high] claim to [web]. Dissenting data kept in the record: Russell and Rossi found only thorium lines in a thorium-ionium preparation where about 1 percent ionium might have been expected; Gleditsch's high Cornwall pitchblende ratio "difficult to explain"; autunite ratios 0.21 to 0.68 of normal. Resolution of the first, which is not in the project files: ionium is thorium-230, an isotope of thorium, so it has the same chemistry and (almost) the same spectrum; Soddy's isotope idea (1913) grew from this kind of result [memory, medium]. This is a point worth a reader's study. The 1913 book itself says "no definite proof of the formation of radium from pure uranium has yet been obtained" [read]; the intermediates (uranium X2, uranium-234) were found later [memory, high]. Final verdict: Established. Historically superseded but instructive: 1904 alpha mass (about 2 H), charge e, speed 0.107c, and the dating of "thorium a million years".

Claim 4. The nuclear atom

Agree (all three): Established; the best-checked derivation in the project (SymPy checks of equations 1-5, recoil quadratic, 3 pi b / 8R); the corrected claim items 1 to 3 above. Disagree: only emphasis. The contrarian and physicist both note that the confirming experiments (Geiger-Marsden, Chadwick, Rutherford-Chadwick) came from Rutherford's own group; independent confirmation is Moseley (1913-14), Bohr's atom, Blackett's cloud-chamber deflections, Rose (1926) and all later scattering [read: 1930 book; memory, high]. Three residual points: the unexplained 35 to 50 percent excess of N/cosec^4 at 15 to 30 degrees in the 1930 book's first set (beam width and plural scattering likely, not stated by the authors); Crowther's thin-foil beta data, which the 1911 paper admits favoured plural scattering, answered by Marsden and "further experiments are desirable"; a Schonland table with a factor-10 inconsistency. The hidden assumptions (single large-angle encounter, heavy target, point charge below 1e-12 cm, classical mechanics, sign unknown) are all stated by Rutherford; the quantum Coulomb cross section is identical [memory, high]. Final verdict: Established (cosec^4(phi/2) law, Coulomb force to the closest distances reached (upper bound 3.2e-12 cm on gold), small massive charged centre). Charge about 100e for gold and N = A/2: Historically superseded but instructive (Z = 79). Sign of the charge: Not established in 1911; established by 1913-14. Composition of the nucleus as helium units and 1930's protons and electrons: Historically superseded but instructive (neutron, 1932).

Claim 5. Earth heat and Kelvin; the sun

Agree: Kelvin's conduction-only limit loses its basis: Established. The 1905 statement that heat gives no definite age is in the source ("no definite limit can be set"). Disagree: the contrarian grades "radioactive heat accounts for the whole" as Refuted by calculation (uniform earth: 200 times too much) and radium as the sun's power as Refuted by calculation; the physicist and mathematician grade the same items Historically superseded but instructive. Stronger: the mixed view. The two specific forms are indeed refuted by calculation (R2, R3), by Rutherford's own 1913 numbers; but they were offered as conditional ("if ... distributed"), the 1913 book itself carries the refutation and the shell correction, and modern geoneutrinos give about half the heat as radiogenic. So the right label for the history is "superseded but instructive", and the label for the arithmetic is "refuted". I give both as parts. Missing from the 1905 article (contrarian [memory, medium]): mantle convection as the second reason Kelvin's limit fails (England 1895, Perry). The sun speculation was right that a nuclear energy store exists in ordinary matter, wrong about the mechanism and the size (R3). Settle: for the earth, geoneutrino and heat-flow data (already accumulating); for the sun, nothing - fusion (Bethe 1938-39, [memory, high]) settled it. Final verdict: Established (radioactive heating removes Kelvin's limit; helium and lead ages are what date the earth). Whole-earth uniform radium budget: Refuted by calculation (about 200 times too much when rock values are used; the 1913 book says so) and Historically superseded but instructive. Radium (or uranium) as the sun's power: Refuted by calculation (thousands of years; one-fourth to one-half of L even for pure uranium). "Sun's life lengthened about a hundredfold by break-up of ordinary atoms": Historically superseded but instructive.

Claim 6. Artificial disintegration; non-Coulomb scattering at close approach

Agree: Established. The 1919 paper was read in the saved copy by the physicist and contrarian (the mathematician could not read it: the text layer is a cover page, 26 lines; the contrarian read it from the OCR page images), and it is hedged exactly as above. Replication: Cambridge group (six elements: B, N, F, Na, Al, P) and Blackett 1925; Cockcroft and Walton 1932 [memory, high]. Disagree: the Vienna group (Kirsch and Pettersson) reported that almost all elements disintegrate; the Cambridge group did not find that (mathematician, from the 1930 digest). Kept in the record as an unresolved contemporary disagreement; the later literature sided with Cambridge on which elements, but no reviewer verified the resolution [memory, low]. The New Energy Times view that Blackett deserves the credit: a stake-holding source; the reasonable reading is Rutherford found the reaction and Blackett identified the product. Rutherford's 1933 remark that releasing atomic energy by splitting atoms was "moonshine" is on record in one newspaper cable (Levin Daily Chronicle, 13 Sept 1933) [web: https://paperspast.natlib.govt.nz/newspapers/LDC19330913.2.68, via the Investigator; neither reviewer opened it]; fission (1938) followed. Both are statements, not measurements. Digest-only: the departures from Coulomb scattering (1930 Ch. IX-X) are accepted from the digest and [memory, medium]. Final verdict: Established (alpha bombardment disintegrates nitrogen, emitting a hydrogen-like particle; product identified 1925). Departure from Coulomb at close approach for light nuclei: Established (digest-level in this review). 1920s models of the anomaly (plate-shaped regions of abnormal force, polarisation): Historically superseded but instructive.

Claim 7. Ordinary matter transforming; an external trigger ("detonator"); no rate change

Split into five parts (the claim mixed them): (a) Detonator position as Rutherford wrote it: "the energy liberated by the explosion ... is derived mainly from the atom itself and not from the detonator" and the trigger is conjecture. Agree among all reviewers that the position is stated openly and consistently across the St Louis address, Radio-activity 1905 sec. 270 and the 1906 book (p.280). Everyone also agrees that Le Bon's quotation ("in doubt whether the energy is borrowed from the atom or from external sources") is a real sentence from the French text (p.55) but is cut from the surrounding argument against external sources; the English version reads "whether the view is taken that the energy is derived from the atom itself or from external sources" [read: sites.md; research/archive-org]. The French wording sounds more open (mathematician). Le Bon is an interested party (priority since 1897). Anyone quoting this should quote all three passages. Disagreement: the physicist grades the detonator "Plausible, untested"; the contrarian says the detonator hypothesis is "Established in a different form" (induced reactions and fission are external triggers whose energy is the nucleus's). Both are right about different things: no one has triggered spontaneous decay of an alpha emitter, whereas projectile-induced reactions and neutron-induced fission are triggered nuclear transformations. The latter realize the hope of "disintegration by laboratory methods"; they are new events (a collision), not a change in the spontaneous rate. (b) Rate invariance under heat, pressure, field, chemistry, radiation (alpha and beta): Established (R5 context), Curie-Danne artifact explained, electron-capture refinement. (c) Ordinary matter slowly transforming (1906 sub-threshold alphas; Le Bon universal dissociation; Poincare): Tested, not supported. Rutherford's 1913 book assigns the residual ionisation to impurities; K-40 and Rb-87 are real exceptions (Campbell and Wood 1906-07); the Geiger-Nuttall rule makes slow alphas astronomically long-lived; the 1913 bound would need a half-life of about 5e17 years [calc: 1/(3600 x 2.7e4) = 1.03e-8 of uranium's rate]. Proton-decay limits beyond 1e34 years [memory, medium]. Historically superseded but instructive. (d) Bumstead's X-ray heating (lead heated "approximately twice as much" as zinc for equal absorbed fraction; the only hypothesis he found not in conflict with the facts was that X-rays break up atoms of some elements, though he examined secondary rays at length [read: Bumstead 1906 conclusions, per contrarian]; Rutherford 1906 called it suggestive, needing more metals and sources). Disagree: the physicist suggests the standard reading is secondary radiation and fluorescence [memory, medium]; the contrarian argues that fluorescence should make lead heat less, not twice as much, and that no replication or refutation was found. My weighing: the contrarian is right that the physicist's explanation has not been shown to work in the right direction, and my own web search found nothing linking the later secondary-ray work to Bumstead's data [web: https://en.wikipedia.org/wiki/Henry_A._Bumstead; https://ajsonline.org/article/127059-the-heating-effects-produced-by-roentgen-rays-in-different-metals-and-their-relation-to-the-question-of-changes-in-the-atom; both record only the lead-zinc factor of two]. A plausible mundane route (inference, mine, unverified) is calibration: the X-ray intensity was measured by ionisation in a gas, and the ratio of heat to ionisation depends on ray hardness and on which element absorbs; with the instruments of 1906 that could produce a factor 2. I do not claim this; I list it as a competing explanation. Final: Never properly tested (one laboratory, one paper, no replication or refutation found; Rutherford's 1913 book later says no evidence that any substance can be made radioactive by bombardment). (e) Gravity: Rutherford named it in 1905 as the one force "not under our control" that might influence stability; Schuster suggested a test; Rutherford and Compton developed a method for emanation decay over about 100 days at different places on the earth, and "the outbreak of the war interfered with this plan" [web: https://www.nature.com/articles/104412c0, abstract; I could not open the letter body: the site redirects to a login page]. Never properly tested (planned, interrupted, result unseen). The first modern decay-rate tests (Be-7 pressure, Re-187) are not gravity tests.

Final verdict: split as above. Detonator conjecture for spontaneous decay: Plausible, untested (and realised in other forms as induced reactions). Rate invariance: Established. Ordinary matter transforming: Tested, not supported. Bumstead X-ray heating: Never properly tested. Gravity: Never properly tested.

---

What would settle it

Reader-testable calculations (for the primer)

Each reproduces in a few lines of Python: the closest-approach table (b = 2Z(1.44 MeV fm)/K); the 90-degree fraction R4; the thorium X recovery curve R5; the heat budget R1; the sun's duration R3; the earth-shell calculation R2 (39 cal/cm^2/yr, 20 km, 300 C).

Open questions

1. Did the Rutherford-Compton gravity measurements ever run, and was the result published or only noted in notebooks?
2. What exactly produced Bumstead's lead/zinc factor of two, and why was it not followed up?
3. How did Curie and Danne respond to Bronson, and was their original claim withdrawn?
4. Was the Dewar liquid-hydrogen "excess" ever repeated?
5. Why did the first angle data in the 1930 book show a 35-50 percent excess of N/cosec^4 at 15 to 30 degrees, and who explained it?
6. What was the final resolution of the Vienna (Kirsch and Pettersson) claim that nearly all elements disintegrate?
7. Whether modern decay-rate variation claims (solar-correlated) have any real effect: contested, unread here.
8. What do Rutherford's private letters say about external-energy and trigger ideas (Cambridge MS Add.7653; Rutherford-Boltwood letters, Badash 1969; McGill letters)? The deciding records for his private view are unread.
9. For the Le Bon and Moray projects: does Rutherford's St Louis French text (p.55-57) support Le Bon's reading? Answer in this record: the sentence is real; the surrounding argument is against external sources (see Claim 7a).

Absence-of-evidence analysis

1. The gravity test (Rutherford and Compton, 1919). Silent after "the outbreak of the war interfered". Competing explanations: (i) never carried out (war, then Rutherford's move to Cambridge and Compton's to Washington University, both [memory, medium]); (ii) done, null, only noted in a notebook or not written up; (iii) published in a place nobody has opened (the Nature letter body; a Compton paper whose title two reviewers cite but I could not locate). Evidence: the letter itself says the plan was prevented by war (for (i)); the same letter in December 1919 describes a developed method, which makes a later null result plausible but unrecorded (ii); no source found for (ii) or (iii). Suppression or concealment: no evidence either way; a null result on a minor sub-question needs no suppression to disappear. The record that decides: the Nature letter body; Cambridge MS Add.7653 notebooks (NB 26-35); Compton papers at Washington University (a finding aid appeared in search, not opened).
2. Bumstead's X-ray heating. Reported, mentioned by Rutherford (1906), absent from the 1913 book except as "no evidence that any substance can be made radio-active by bombardment". Explanations: (i) quietly explained by secondary rays in the Cambridge circle without a paper (physicist's inference, unverified); (ii) the effect was a calibration or geometry artifact found privately; (iii) the field moved to crystal and spectral X-ray work (Barkla, Bragg) which made the question moot [memory, low]; (iv) neglected because it had no theory to attach to. Evidence for any: none in the record. The record that decides: Bumstead's full paper and any correspondence with J. J. Thomson (Bumstead was at the Cavendish in 1905, per Wikipedia).
3. Dewar's liquid-hydrogen heat. Reported once at the British Association 1903; Rutherford asked for confirmation; no later confirmation or retraction found by any reviewer. Likely (i) a measurement artifact (printed numbers show no excess); (ii) never repeated because the calorimetry improved; (iii) a transcription issue in a conference report. Decides: Dewar's 1903 paper.
4. Curie and Danne after Bronson. No reply found. The Bakerian Lecture of 1904 and Bronson of 1905 are both in Rutherford's circle; the independent party (the Curie laboratory) is the silent one. Explanations: they accepted the volatility explanation without comment; they replied in the Comptes rendus or the Curie laboratory notes; no reply was needed. Decides: Comptes rendus 1905-06 and the Curie laboratory records.
5. Russell and Rossi (ionium lines) and Gleditsch's high ratio: reported honestly in the 1913 book, but the project holds no later explanation. The isotope reading (ionium = Th-230) is a standard resolution but is [memory, medium] here.
6. Rutherford's private views on external energy and a trigger: the private record (Cambridge, McGill, Boltwood letters) is the place where a difference from the public position would show; it was not read. Nothing suggests suppression; the record is simply not item-level searchable online.
7. No institutional dissent to Rutherford's main claims survives in the project record (the 1930 book, the 1913 book and popular articles quote the external-energy proponents fairly). Silence of other sources is not read as refutation.

Corrections to earlier documents (for the orchestrator)

• claims.md claims 2, 3, 4, 5, 6, 7: apply corrected items 1 to 12 above (Chadwick/gold; Coulomb bound; "nucleus" word and sign; speed; 3.7e9 provenance; per gram bromide; earth radium whole-earth versus crust, 200 times; hundredfold as scaling; 1919 hedge and Blackett 1925; Bakerian temperature effect and earlier detonator sources; claim 7 split in five).
• claims.md "Reading limits": the 1913 alpha-ray chapter deep read is still lost; add that the 1919 Nature letter body and the "Compton 1920 Phil. Mag." title are unverified.
• facts.json: Harper's article date (February 1905); 1930 book text on Chadwick and gold; 1911 paper note on printed table errors; 1913 book slips (radium D 115 years, 5.2e21/2.6e21, one-fourth/0.46).
• physicist.md, contrarian.md: the alpha-energy explanation (R1) and the "Compton 1920 Phil. Mag." citation (item 16) should be treated as superseded by this file; their other verdicts stand as graded above. The physicist's "Established" for claim 5 should be read with the split.
• sites.md leads: add that Rutherford and Royds 1909 is now supported by a web summary; add Compton 1920 title as unverified.

Summary of final verdicts

---

QE revision (2026-10-09)

Inputs read for this revision: qe.md (verdict MINOR, in full); the new deep reads (1930 book PDF pp.204-211 and 574-581, 350-369, 269-349, 422-447; 1913 book PDF pp.131-192, 340-346 and 525-535, 595-611; 1904 book PDF pp.340-364); the saved Bumstead 1906 text (sections 8-9). Web: one check (neutrino, below). Tags as before. Where I depart from qe.md I say so. It is organised by section; the table below says which section changes which verdict.

Source gap update. Closed: 1930 Ch. IX-XI (now deep-read, with page images), the lost 1913 alpha-ray chapter (re-run), 1904 Ch. X. Still digest-only: 1930 Ch. I-VII, XII, XIV-XVII; 1913 Ch. I-X apart from the alpha chapter; 1904 Ch. I-IV, VIII-XI apart from Ch. X sec. 199-207. The 1919 Nature letter body is still unread.

Verdicts that change

No other verdict changes. Claims 1 to 6 remain Established in their core; the Claim 5 refutations by calculation stand.

1. Dewar's liquid-hydrogen heat (Claim 2): QE C2 accepted, with limits

The 1913 book, Sec. 233, reports the Curie and Dewar measurement (Proc. Roy. Inst. 1904): a guard-bath calorimeter, with the radium tube in a small Dewar flask inside a second flask of the same boiling liquid, so the gas boiled off measured the heat. The rate was "about the same in boiling carbon dioxide and oxygen, and also in liquid hydrogen". The one number: 0.7 g of radium bromide, ten days old, gave 73 c.c. of gas per minute [read: deep/Radioactive-Substances p.595]. The deep reader's check gives about 39 cal/h against about 47 cal/h expected for a ten-day-old preparation (the gas temperature is not stated, so about 20 percent low is within the unknowns), which shows no excess [calc]. The 1906 book and the 1913 book Sec. 236 repeat the finding [read: QE C2]. My earlier text called the 1903 report "reported once ... no confirmation or withdrawal found". That was wrong: the 1904 measurement, with Dewar himself as co-author, is the follow-up. Limits I keep in the record: it is one number, from the same laboratory that made the 1903 report, with no tolerance given; Rutherford, who reports it, is an interested party. It shows that the heat does not depend on temperature from about -78 C to -253 C; it does not by itself identify the source. Verdict: Tested, not supported. The Absence-of-evidence item 3 (Dewar) is closed: the record did not go silent; the follow-up is in Curie and Dewar 1904. What is still unseen: the 1903 British Association text itself.

2. Where the 3.7e9 cal/g comes from (Claim 2): QE C3 accepted in part

QE says the 1913 book derives 3.7e9 in Sec. 238. The deep read of PDF pp.595-611 shows less. The book derives 2.44e9 cal/g for the emanation chain exactly: 109 cal/h per curie, 0.6 mm^3 per curie, atomic weight 222, lambda 0.0075 per hour, giving 1.82e5 cal/h per c.c., 1.83e7 per gram, divided by lambda gives 2.44e9 [calc, reproduced]. For radium through F it states "about one quarter" for an average gram of alpha-emitter (6e8 cal/g) and then "about 3.7e9 calories" with no steps [read: p.607]. A reconstruction (inference, the deep reader's and mine): five alphas for radium through F against three for the emanation chain, energy ratio 29.3/19.2 = 1.53 (4.78 + 5.49 + 6.00 + 7.69 + 5.30 MeV against 5.49 + 6.00 + 7.69, [memory, high]), times 222/226, times 2.44e9, gives 3.6e9 [calc]. Correction to my own item 5: the five alpha energies sum to 29.3 MeV, not 29.8 (still 3.0e9 cal/g; the full mass defect, 35.3 MeV per atom, gives 3.60e9). So: the number is in the book and is right; it is a stated figure, not derived in the book; "6e8 times anything simple" does not reproduce it. The 1904 first edition gave 2e9 to 5e10 cal/g (sec. 205, from 876,000 cal/year divided by lambda for the minimum and maximum life estimates), "the minimum estimate is probably nearer the truth" [read: deep/radio-activity section 6]. That 2e9 is the product of 100 cal/h and a mean life, as I first thought; the later 3.7e9 is a different, fuller figure. The 1904 text prints 876,000 in sec. 205 and 864,000 in sec. 207, a 1.4 percent slip in the source.

Heat split (1913, Vienna standard): alpha 123.6, beta 4.3, gamma 6.5, total 134.4 cal/h per gram of radium; the emanation chain 109.3. The share from the emanation and products is 81 percent of 134.4 [calc]; the early 1904 split ("three quarters"; 23, 32 and 45 percent) is superseded by the table. Calculated and observed alpha-plus-recoil heating of one curie of emanation agreed, 94.5 and 94 cal/h, which closes the 1904 bookkeeping gap (R1) from the experimental side [read: p.602-603].

One anomaly in this chapter that nobody had listed: Poole (1910) measured the heat of 500 g of pitchblende and found 12.8e-5 cal/h per gram of uranium against the calculated 7.2e-5, about 1.8 times higher; his thorium mineral orangite (1912) was far above calculation and he concluded much of it was "heat liberated owing to some slow chemical change in the mineral". Rutherford: "unsuspected chemical changes in progress in the large mass of mineral" [read: p.609-610, deep/Radioactive-Substances B7]. Reported by one observer; the book does not resolve it. I record it as Never properly tested (no repeat in the record held); the ordinary explanation (chemical change, gas retention, impure sample) is Poole's and Rutherford's, not a measurement.

3. The beta-energy puzzle (new; Claim 2): a real energy-accounting anomaly

QE missed-material item 1 is confirmed from the page images. Alpha emitters release one energy; beta emitters do not. If every radium E nucleus released the maximum of its spectrum, heat per disintegration would be 1,050,000 electron-volts; if each electron left with its own energy, the mean, 390,000 +/- 60,000. Ellis and Wooster (1927, lead calorimeter, all betas absorbed, with the polonium that grows in separated by its known period) measured 350,000 +/- 40,000; Meitner and Orthmann (1930), 337,000 +/- 20,000. The book concludes "the nucleus can break up with emission of an amount of energy that varies within wide limits. No satisfactory interpretation has yet been given". The deep reader notes that the book does not mention a neutrino in these pages [read: deep/Radiations section 1, p.440-441]. A second line (ionisation of radium B+C betas, 9.65e14 ion pairs per second at 40 V each, about 1e6 V against a 2.7e6 V sum of end points) agrees [read; calc: 1.04e6 V]. The book also rules out the idea that the spread arises after emission (Aston's gamma limit about 10,000 V per disintegration against the 600,000 V that idea needs). Interested parties: Ellis is a co-author of the book that reports Ellis-Wooster; Meitner-Orthmann is a Berlin repeat by a different group.

What mainstream physics did: Pauli proposed the neutrino in 1930 to rescue energy conservation for the continuous spectrum, and it was first detected by Cowan and Reines in 1956 [web: https://en.wikipedia.org/wiki/Neutrino, fetched 2026-10-09; Fermi's 1934 theory and Bohr's openness to giving up strict conservation are [memory, high], not checked]. For the reader: the measurement shows the mean heat matches the mean electron energy; it does not say where the rest of the maximum energy goes. This is the clearest case in the project where an accepted law was kept by postulating something unseen and then confirmed 26 years later; a reader testing mainstream claims should read it beside the cases where a postulated unseen thing was never found. Verdict on the anomaly as a challenge to Claim 2 ("the energy is the atom's own and is accounted for"): Historically superseded but instructive. The core of Claim 2 stands, because for alpha emitters the books' accounting closes to a few percent.

4. The sun and the earth (Claim 5): QE C4 accepted, checked

The 1904 first edition shows where the 2.5 parts per million comes from. W. E. Wilson (Nature, 9 July 1903) used Curie and Laborde's 100 cal/h per gram of radium and Langley's solar emission, 8.28e6 cal/cm^2/h; the book's two forms agree (2.5 ppm, or 3.6 g per cubic metre at density 1.44) [read: deep/radio-activity section 8, with the deep reader's check]. The modern emission is 5.42e6 [calc from L = 3.83e26 W]. The ratio is 1.53, which is my "factor 1.5" in R3: it is Langley's constant, not an error by Rutherford. With Langley's value the required radium is 2.5e-6; with the modern value 1.66e-6, as I computed. The 1913 "one-fourth" for a sun of uranium: 0.46 of L with the modern constant, 0.46/1.53 = 0.30 with Langley's [calc]. So my "unexplained factor 1.8" shrinks to about 1.2 (0.30 against 0.25). The conclusion stands either way. The 1904 text also gives "50 to 500 times longer than Kelvin" for the sun's life, stated "if" and without derivation, as a conditional guess that ordinary elements carry radium's store and break up faster in the sun [read: p.362]; my own check gives 74 to 1,850 times Darwin's gravitational energy, so the printed range is not derived from those numbers. Darwin's gravitational store, 2.7e7 cal/g, checks [read; calc]. Verdicts unchanged: radium as the sun's power, Refuted by calculation; sun's life lengthened, Historically superseded but instructive.

Earth (Claim 5): the 1904 first edition already contains the heat calculation X = 3KT/R = 7e-15 cal per cm^3 per second, equivalent to 4.6e-14 g of radium per gram of earth, with K = 0.004 and gradient 0.00037 C/cm [read: p.362-363; checked 6.97e-15 per second]. That 4.6e-14 comes from the 1904 book, not from the 1905 Harper's article. His evidence was Elster and Geitel's clay: about 1e-13 g radium per g, "about the right order of magnitude" [read: p.364]. Only about twice the threshold in 1904; by 1913 crustal rock had been found to hold about 2e-12, which is why the uniform-earth picture failed and the thin-shell picture followed (item 8). This is a fair record of how a number changed, and I keep it for the primer. The source also shows the argument assumed uniform distribution and surface conduction as the whole loss [read: notes in section 9].

5. Claim 6

Elements. QE C5 confirmed: by the long-range-proton method Cambridge found B, N, F, Na, Al, P; by the right-angle method (1924) it added Ne, Mg, Si, S, Cl, Ar, K. "All the light elements up to and including potassium can be disintegrated ... with the exceptions of hydrogen, helium, lithium, beryllium, carbon, and oxygen" [read: 1930 p.321, PDF p.321]. Thirteen elements, at about 20 per million alphas (nitrogen), 8 per million (aluminium), about 1 per million (Cl, Ar, K) [read: deep/Radiations section 1]. My earlier "six elements" in the Cambridge-Vienna contrast is wrong and is withdrawn.

Departures from Coulomb scattering (read from page images, 1930 Ch. IX): hydrogen, 8.2 cm alphas at 30 degrees, observed 4.3e-5 against Coulomb 4.4e-7, a factor of about 100 (Chadwick and Bieler); Coulomb holds to about 4e-13 cm for central collisions. Helium: factor about 20 at 40 to 50 degrees, as low as one third at small angles (Rutherford and Chadwick 1927); central Coulomb to 3.5e-13 cm. Aluminium and magnesium: below Coulomb, to about 0.35 at 135 degrees, then above ("several times" at 155 degrees, unpublished); Coulomb holds to about 13e-13 cm. These are quoted from the book by its authors, who did the experiments (interested parties); independent confirmation for the light-nucleus departures is in the later literature, [memory, medium], not checked here. Models: oblate-spheroid alpha (8 x 4e-13 cm; a 1930 footnote withdraws the shape), Bieler's 1/r^4 attraction, Debye-Hardmeier polarisation (works for Al and Mg, fails for He). The book concludes "a hitherto unknown, though not altogether unsuspected, force" [read: deep/Radiations sections 1, 3]. Verdict: Established as an observation; models Historically superseded but instructive. The book's own defects: uranium closest approach "38.0" should be about 35e-13 cm; Blackett's quoted "19 percent" energy loss does not match the six ratios in the book (mean 0.843, a loss of 16 percent, [calc]).

Blackett's forks: 23,000 photographs, about 415,000 alpha tracks, 8 forks of two arms (proton and recoil); the alpha is captured; residual nucleus mass 17, atomic number 8. Six measured energy ratios 0.83, 1.02, 0.72, 0.68, 0.88, 0.93: energy was lost in five of six, and (as QE noted) the change ranges from -0.33 to +0.02 of the alpha energy; the book later says the nitrogen reaction absorbs energy and the aluminium reaction releases it (protons up to 1.37 times the alpha energy). The book's own account of other fork reports that broke conservation (Auger and Perrin, Akiyama, Wilson's beta forks) is in the digest and not re-read here, so I keep it as listed in QE section 3 item 8 and do not weigh it.

Vienna and Cambridge. As the book states it: "nearly every element can be disintegrated ... and the chance of disintegration is much greater". The book declines to explain and "to avoid fruitless discussion" reports only Cavendish experiments; no Vienna data are in the book. So the Cambridge text is one party describing the other. The later history (a Chadwick visit; Vienna's counters judged unreliable) is [memory, medium] and unverified here. Verdict: unresolved in the record I hold; the weight of later literature is with Cambridge [memory, low]. Settle: Kirsch and Pettersson, Atomzertruemmerung (1926), and Chadwick, Phil. Mag. 2, 1056 (1926).

6. Transmutation by ordinary means (new; QE missed-material item 2)

The 1930 book Sec. 74 lists six claims. For each I give the claimant's description, then the book's account of how it was answered (the book's authors are not neutral: they are the group whose theory says it cannot happen).

The book's general argument is energy scale (nuclear binding "several million volts" against chemical energies) and that "the presence of an element has been mistaken for its creation". Add the earlier 1907-08 episode (1913 book Sec. 116): Cameron and Ramsay reported neon and argon from emanation acting on water, and lithium (with sodium and potassium) from copper solutions. Rutherford and Royds repeated it and found only helium, attributing the neon to a leak Cameron and Ramsay "admittedly" had; Curie and Gleditsch found lithium in glass and quartz vessels and none in platinum; Gleditsch and McCoy found no relation of copper, radium and lithium in minerals. Ramsay and Usher's carbon dioxide from thorium and zirconium is answered by Rutherford only with an argument (organic matter makes carbon dioxide under emanation), not a recorded repeat [read: deep/Radioactive-Substances Part B]. Ramsay's own replies are not in the pages read (he is an interested party and a Nobel chemist, not a fringe figure). Why it matters: Paneth-Peters is the best-known 1920s antecedent of the 1989 palladium claims [memory, high], and the specific control (helium absorbed by glass from the air) is the one any later helium claim must meet.The Contrarian's question is how each claim was rejected: here the rejections were by repetition and by the claimants' own withdrawals in four of six cases, not by dismissal.

7. Detonator and gravity (Claim 7a and 7e): QE C1 and C8

C1 confirmed. The 1904 first edition, sec. 206, book p.339, has both sentences in the project's own copy [read: deep/radio-activity section 7]: "It is conceivable, for example, that the application of some slight external force might cause instability and consequent disintegration, accompanied by the liberation of a large amount of energy, on the same principle that a detonator is necessary to start some explosives." and "unless possibly it is assumed that the force of gravitation which is not under our control may influence in some way the stability of the radio-atoms." Rutherford adds that the exponential law "would be expected equally on either hypothesis". The sec. 204 text goes further on gravity: "if gravitation is the result of forces having their origin in the atom, it is possible that, if the atom were disintegrated, the weight of the parts might not be equal to that of the original atom", and that Heydweiller and Dorn reported weight changes of radium, "not confirmed", and Forch saw none [read: p.354]. That is a second gravity-related question (weight of the radium, not decay rate) and I add it to the record. Correct "Rutherford named it in 1905" to 1904 (this is an error of mine and of the primer and sites.md lead; the 1905 second edition repeats it). The St Louis address dating stands.

Also in the 1904 book, the earliest test of a radiation trigger: J. J. Thomson (Nature, 30 April 1903) suggested radium's own radiation might drive its decay; Rutherford dissolved a few milligrams of radium bromide into barium chloride solution at least 1000 times greater in volume and saw "no appreciable change of the activity measured by the gamma rays ... over the period of one month", concluding the radiations have "little if any influence" [read: p.357-358]. His own caveat from the reader notes: one run, one experimenter, gamma readings only.

Radiation triggers, upper limits in the 1930 book (page images, Sec. 76): Rutherford 1911, radon decay constant constant to 0.05 percent over a thousandfold fall in concentration (about ten half-lives, 38 days [calc]), confirmed by Mme Curie 1924; Danysz and Wertenstein, alpha on uranium, at most 1 alpha in 5 million effective, on mesothorium 1 in 26,000; Herszfinkiel and Wertenstein 1928, alpha on thorium, 1 in 8 million (a test of Rutherford's own satellite theory's "triple transformation" prediction); Hevesy 1922, gamma and beta from 800 mg radium on uranium for six weeks, no change as great as 0.1 percent (not more than 15 in 1e10 absorbed gamma quanta), radium D under 0.2 percent in 51 days; Jorissen and Vollgraf 1915, cathode rays on uranium, none; Walter 1926, 200 kV X-rays on uranium, none [read: deep/Radiations section 9]. Citation issue: the book prints Danysz and Wertenstein as "C.R. 161, 784, 1925"; volume 161 is 1915 [memory, high], so 1925 is probably a misprint and Danysz probably died in 1914 [memory, medium]. Not checked on the web. The printed count of 6e14 alphas falling on the oxide is about the whole emission, so that limit may be optimistic by up to a factor 2 [calc by deep reader]; it does not change the conclusion. Interested parties: all by Rutherford's group or collaborators except Hevesy, Jorissen-Vollgraf and Walter.

Split verdicts (QE C8 adopted): - Trigger by radiation: Tested, not supported, at these limits. Note a limit of the tests: they probe only the radiations that were available and the decay of alpha and beta emitters. - Trigger by an unspecified external force: Not testable as stated (no force is named). - Gravity: Never properly tested. The 1930 book says: "The effects of temperatures up to 3000 C., of pressures of 20,000 atmospheres, of powerful electric, magnetic, and gravitational fields have all been looked for, and all in the final analysis without a positive result" [read: page image PDF p.352]. No reference is attached to any item; the first footnote on the page belongs to the next paragraph. Rutherford co-wrote the book and was the one who planned the 1919 test. So this is the authors' summary claim, not a cited measurement. It is pertinent evidence that some gravity looking was done and found nothing, from a party with a stake in the answer; I do not count it as a controlled test. It does not match the "interrupted by war" record, and I do not know which is right. - Triggered nuclear events (artificial disintegration, fission) are new collisions; they realise the hope of laboratory disintegration but do not change the spontaneous rate.

Absence-of-evidence item 1 (gravity test), added evidence: the 1930 sentence, and the 1904 weight-of-radium conjecture with its three unconfirmed or null reports. This strengthens explanation (ii) (done, null, not written up) a little and does not remove (i) (never carried out); no source decides between them.

8. Claim 7b, 7c, 7d

7b (temperature and pressure), 1913 book Sec. 200 [read: deep/Radioactive-Substances Part A]. Gates showed active deposit is driven off a wire above red heat; Curie and Danne found activity varied with temperature up to 1350 C; Bronson (1907) showed radium B is more volatile than C, the wire then decaying at 19 minutes; gamma activity of a quartz tube unchanged to 1300 C; all observers agree the gamma and emanation decay does not change; the beta anomalies (some saw a temporary rise, some a fall) are explained by Russell (1912) by redistribution of the products in the quartz altering beta absorption, with volatilisation at about 600 C revised to about 360 C in hydrogen; Rutherford and Petavel, cordite bomb at about 2500 degrees and 1000 atmospheres, no change at the moment of explosion (later changes over hours attributed to redistribution); Schuster, gamma unchanged at 2000 atmospheres, and Eve and Adams similar. The summary "-186 C to 1500 C" has no experiment attributed to it in the passage. One correction to my earlier item 11: the book describes Curie and Danne's effect as explained by volatilisation, and gives the temperature of Bronson's wire test; my "1.4 times normal after 1100 C" comes from the 1904 Bakerian Lecture (contrarian's page images) and I did not re-read it. Verdict unchanged: 1904 temperature effect on radium C, Tested, not supported (explained); invariance, Established.

Geological time (1930 book, haloes, PDF pp.204-211). Halo ring radii match alpha ranges in air for six of eight groups in fluorspar (Gudden), with unexplained residuals: the uranium I and II rings are 2.83 and 2.93 cm against Laurence's 2.73 and 3.28, "not easy to account for"; Joly found the inner ring radius grows with the age of the mica. Explanations offered, none adopted: Joly's idea "at one time" that the decay rate and alpha range changed through geological time; short-lived uranium isotopes now extinct (Wilkins); damage to the mica (Lotze, whom Joly disputed); chemical colouring (Poole). The authors say a systematic examination of varied materials is required. Joly and Rutherford (1913) estimated several hundred million years for some haloes. Unexplained "X" haloes (ranges 1.5 to 2.25 cm, Joly) and "Z" haloes (1.2 to 2.1 cm, Iimori and Yoshimura) match no known alpha range; "their exact origin is unknown". I add this to Claim 1 as evidence for constancy over geological time and as an open anomaly. Verdict: constancy of alpha energy, so of decay, over geological time: Plausible, with the residuals above Never properly tested; I do not claim the haloes refute or prove constancy. The caution for any reader: the Joly passage is the only place the project's sources allow that rates changed over time, attributed "at one time", not endorsed, and not refuted in the book.

7c (ordinary matter transforming). Hoffmann (1920, 1927) found weak alpha activity in platinum and brass "not due to known radioelements"; Ziegert (1927) traced the brass to its zinc and, by chemical separation, got a residue with three alpha groups at 0.55, 1.15 and 2.05 cm, and suggested a link to the X and Z haloes. The authors say the Geiger-Nuttall relation would make such particles very slow-decaying, inconsistent with the rate observed from the small residue: "It is indeed difficult to offer any suggestions as to the source of these alpha particles, and further information must be awaited" [read: deep/Radiations part B3]. Final statement of the book: no adequate proof of intrinsic activity in ordinary elements, leaving Ziegert's observations for further investigation. The v. Schweidler estimate puts the half-life of copper, if the residual ionisation were intrinsic, at about 3e15 years (the text layer's "3e5" is a scan error). Potassium and rubidium are real: weak beta emitters, no alphas, per-atom activity independent of origin or age; the book gives half-lives 1.5e12 (K) and 1e11 years (Rb), assigns potassium's activity to mass 41 (the radioactive isotope is potassium-40 [memory, high]), finds no decay products. The book's K-41 is a misassignment; it is instructive and I keep it. Verdict: ordinary matter transforming, Tested, not supported; Ziegert's short-range alphas, Never properly tested (one observer, one residue, no repeat in the record); I did not find a later resolution [no web check made]. Also to record, from QE R2: Gates' tests of Le Bon's claims (quinine sulphate 1904; heat volatilises rather than destroys the active deposit, 1903) and Le Bon's dispute of her ultraviolet explanation are in sites.md; they were not re-read here.

7d (Bumstead). QE C7 verified in the saved text. Bumstead measured the fraction of rays absorbed with the radiometer (which measures energy) and found zinc 88 percent and lead 91 percent of the energy that passes the aluminium window, against the electroscope's 78 and 79 percent for the same pieces [read: research/pages/archive-amjsci-1906-01.txt, printed p.18, sec. 8]. So my suggested calibration explanation (gas ionisation versus energy) does not explain the factor of two, as far as Bumstead could test it. Bumstead himself raised the secondary-ray explanation (sec. 9, "An examination of this possibility is therefore necessary") and, per the contrarian, argued it does not work; the mathematician's reading that secondary rays "explained" the effect is a misreading of sites.md. I withdraw my calibration suggestion from Absence item 2(ii). Remaining: Never properly tested (one laboratory, one paper, no replication or refutation found).

9. The alpha chapter and Claim 3 (QE C6)

The lost 1913 deep read is re-run (PDF pp.131-192). Source gap limit 2 is closed. The book reports: Rutherford and Geiger (1908) measured the charge of one alpha directly as 9.3e-10 e.s. units, twice the unit 4.65e-10 that the book adopts (itself derived from this measurement); Regener (1909, scintillation counting, polonium) 9.58e-10, unit 4.79e-10; modern 4.80e-10 [memory, high]. Rutherford notes Regener's value may be too high because recoil atoms carried charge with the uncovered polonium plate (Rutherford's critique of a competitor; the two agree to 3 percent). Alpha count: 3.4e10 per second from one gram of radium and 13.6e10 with its three alpha products, replacing the 1905 value of 6.2e10 that rested on an assumed charge of 3.4e-10; modern 3.7e10 per gram [memory, high], so the 1908 count is 8 percent low. Rutherford and Royds (1909): thin glass tube (wall under 1/100 mm) of emanation inside an exhausted tube; yellow helium line after two days, the whole helium spectrum after six; control: helium put in place of emanation gave no trace of helium in the outer vacuum after several days; and a lead sheet in open air over the tube held helium. The book: "There can thus be no doubt that the alpha particle becomes a helium atom when its charge is neutralised." Consistency: 13.6e10 atoms per second per gram is 154 mm^3 of helium a year (the book prints 158) [calc], agreeing with the measured production. The text layer was wrong on several exponents; the page images are right. Claim 3 verdict unchanged: Established. My earlier source gap statement, that the project holds no primary text for alpha = helium, is now false in both directions: the digest-level pages and this deep read both carry it. Also kept: alpha energies from the book's table (radium C alpha 1.31e-5 erg, 2.37e5 ions, 35 eV per ion pair) agree with modern values to about 5 percent [read; memory].

10. Uranium-radius paradox and the satellite theory (1930 Ch. XI)

Scattering experiments (Rutherford and Chadwick 1925; Chadwick, unpublished) found uranium obeys Coulomb scattering, 1/v^4, within 5 percent down to 3.9e-12 cm, so the radius is under about 4e-12 cm. Decay needs the alpha to start at r = 2Ze^2/E = 6.3e-12 cm (Z = 90, E = 4.07 MeV) [read; calc 6.37e-12]. The book calls this "difficult to reconcile on any theory based on the classical mechanics" and says wave mechanics (Gamow; Gurney and Condon) explains it. Rutherford's own 1927 satellite theory (neutral alpha satellites in quantum orbits around a core under 1e-12 cm, bound by polarisation) is criticised in the book itself on three counts: it gives no exponential law or Geiger-Nuttall relation, it needs unspecified assumptions, and it predicts disintegration or non-Coulomb scattering that was not seen when fast alphas hit uranium; the Herszfinkiel-Wertenstein null test (1 in 8 million) went against its triple-transformation prediction. This is a good example of a published theory refuted by the same group's data. Claim 4 verdict unchanged; the 1920s satellite model: Historically superseded but instructive (the book calls the Gamow theory "tentative"; the worked results it reports are Gamow's and Gamow-Houtermans').

11. Retention items (QE R1 to R5): restored here

• Rutherford at the British Association (Levin Daily Chronicle, 13 Sept 1933): hopes of releasing energy by splitting atoms proved "moonshine" [web via the Investigator]. Richardson at the British Association (Press, Christchurch, 13 Sept 1921): Rutherford's work made transmutation "an established fact. If atomic energy could be intensified and controlled, we would have at our disposal an illimitable supply of power" [web: https://paperspast.natlib.govt.nz/newspapers/CHP19210913.2.57; OCR noisy, per the Investigator]. Both are statements of expectation, not measurements, and from an interested party (Richardson) and the inventor.
• The contrarian's point that the Chadwick-Mercier gold data in the 1930 book imply an effective charge about 3 below 79 (deep/Radiations 6.5) stays in the record as stated by the contrarian; I did not re-derive it.
• Mantle convection (England 1895; Perry) as a second reason Kelvin's age failed stays in Claim 5.
• The 1924 Franklin Institute address and "Some Cosmical Aspects of Radioactivity" (1907) are leads in sites.md for Claims 6/7 and 5; the primer lists them as further reading.

12. Where I differ from qe.md

• QE C3: the 3.7e9 cal/g is a stated figure in the 1913 book, not a derived one (section 2 above). Only 2.44e9 is derived step by step.
• QE C4: accepted and extended: Langley's constant explains the 2.5 ppm exactly and reduces the "unexplained factor 1.8" for the 1913 "one-fourth" to about 1.2, not zero.
• QE C7: accepted and verified against the saved text.
• QE C2: accepted, with the limits in section 1.
• QE section 6, scale wording: accepted.
• QE retention item R3: not re-derived; kept as the contrarian's statement.

13. Updated "What would settle it" additions

Corrections to earlier documents (for the orchestrator, from this revision)

• arbiter.md earlier text, now corrected here: "named gravity in 1905" (1904); Dewar "never repeated" (Curie and Dewar 1904); "six elements" (13); "3.7e9 not derivable" (the 1913 book states it; only 2.44e9 is derived); "unexplained factor 1.8" (about 1.2 after Langley's constant); five alpha energies "29.8 MeV" (29.3); "the project holds no primary text for alpha = helium" (digest-level and now deep-read).
• claims.md: claim 2 (Dewar; 3.7e9; beta-energy puzzle), claim 5 (4.6e-14 is from the 1904 book, Langley), claim 6 (13 elements; Coulomb departures; transmutation claims by ordinary means as related claim), claim 7 (gravity and detonator 1904; 7a split; 7c Hoffmann and Ziegert; 7d Bumstead radiometer), Reading limits (what is now deep-read).
• facts.json: 1930 book entry (summary, key_claims, how_read: Ch. VIII, IX-XI, XIII 88-94, haloes, ordinary-matter residue now deep-read; add Sec. 74, 76, 93, 124); 1913 entry (alpha chapter re-read; Sec. 200, 233-240, Ramsay-Cameron); 1904 entry (sec. 199-207, gravity sentence in sec. 206 and weight conjecture in sec. 204).
• primer.md: Section 4 detonator and gravity (1904), sun row (Langley), Section 6 evidence rows for Dewar, radiation triggers, transmutation claims by ordinary means, beta-energy spread; Section 7 open items; further reading (1924 Franklin address, 1907 cosmical aspects); define cosec; add "moonshine" and Richardson.
• sites.md lead 1 (2nd edition adds gravity: it is in the 1904 first edition).
• physicist.md, mathematician.md, contrarian.md: superseded on the points above (Compton 1920 citation, Bumstead secondary-ray reading, alpha-energy explanation, "could not trace 3.7e9"); not rewritten.
• Stale phrases to search for in review/, work/rankings/entries/ and site/: "named gravity in 1905", "never repeated", "six elements", "not derivable", "unexplained factor 1.8", "Plausible, untested" for the detonator, "29.8".