Rankings Primer Arbiter Physicist Mathematician Contrarian QE Sites log Claims Sources Research notes

Larken (E. L. Larkin, Radiant Energy and Its Analysis, 1903): Science Contrarian review (STEP 2)

Reviewer role: Science Contrarian (judges by experimental and observational evidence; builds the strongest honest case; applies the same rigour to the inventor as to the critics). Tags: [read] = the book or a document in the folder, [web] = URL given, [memory: confidence] = recalled. Computations marked "calc" were run in the project venv.

Source note: one scanned book, text layer good. Every claim below is Larkin's own and rests on his single book plus the Investigator's later finds. The word "radiant energy" in the title means ordinary electromagnetic radiation in the 1890s sense (Melloni, Langley); there is no free-energy device here [read: pp.17-19, 177-178].

Corrections to Step 1 (and to the Investigator)

Date of the earth-current table: 16 July 1892, not 1882 (Investigator's correction confirmed). I checked it myself. The book's heading reads "JULY 16, 1882" [read: page image p.134, rendered by me]. But (a) 16 July 1882 was a Sunday and 16 July 1892 a Saturday (calc, Python datetime); (b) Larkin's own Science letter of 19 July 1892 describes "a magnetic storm ... on Saturday, July 16, 1892" with a telegraph disturbance from 10.30 a.m. central time [read: research/files/larkin-jstor-1766041.pdf; web: https://archive.org/details/jstor-1766041]; (c) trade-press snippets give the same 133 mA / 492 V for 16 July 1892 [web: Google Books snippets, research/2026-10-08-google-books.md]. claims.md claim 2 and facts.json (summary, key_claims "16 July 1882 ... 144 mA ... 280 mA at 644 V") should say 1892, "the book prints 1882". The deep read's "16 July 1882" (sections 4, 7, 8) likewise. The 17 Nov 1882 storm and Larkin's 15-19 Nov 1882 sunspot drawings are a different event and stay 1882 (17 Nov 1882 was a Friday, calc; it was a real great storm [memory: high]).
Do not say the table was "copied" from a trade source. research/2026-10-08-google-books.md says the table "was copied from an electrical-trade source". The evidence shows only that the figures agree with figures published in 1892-93. Direction of copying is unknown: both may come from telegraph-company records, or Larkin may have drawn on the journal. The book does not say Larkin measured them ("Made at Terminals, New York and Boston", no observer named) [read: p.134]. facts.json should call them "unsourced figures that match the 1892-93 trade press".
The 144 mA is the book's misprint, now with outside support. Table prints 144 mA / 492 V at 12:40 1/4; text says 133 mA. The 1893 journal gives 133 mA and 492 V together [web, snippet]. 492 / 0.133 = 3,700 ohm, not the printed 2,700 ohm, so the mismatch lies in the printed source, as the Investigator says. I would add: the volts were probably computed as I x R, so the "pressures" are not independent measurements [calc from the page image: nine of ten Boston rows give V/I = 3,700 +/- 1%].
Larkin's own timing in 1892 is better than the book's. The book says only "at the height of the upheaval on the sun" [read: p.134]. His 1892 letter gives clock times: telegraph disturbance from 10.30 a.m. central, and a sunspot "tongue" brightest 9.30-10.30 a.m. "at which time the electric wave disturbed the telegraphs" [read: Science 20, 65-66]. The table's 12:37-12:46 p.m. is probably Eastern time (about 11:40 central), inside his window. Step 1 did not have this; reviewers should weigh it (see claim 2).
claims.md claim 5 says the year-long test "is said to have ruled out atmospheric dust". Larkin's text says the floor was seen "on the same nights, when black space was seen between stars in adjacent ... regions" [read: p.300]. That is a within-night contrast control against uniform dust haze only. Say that, not "ruled out".
Digest p.134 line mixes two rows (the deep read already corrected this) [read].

Weighing the Investigator's findings (kept whatever my verdict)

Larkin's 1892 Science letter (own work): the earliest form of claim 2, with causation explicitly left open: "Whether the solar turbulence causes terrestrial magnetic upheavals is a question that future physicists must decide." The 1903 book asserts what he left open in 1892. He also records telegrams sent "by means of nature's electricity, without batteries" [read]. That matches the best-known 1859 case (operators ran about two hours with batteries disconnected) [web: https://en.wikipedia.org/wiki/Solar_storm_of_1859].
1881 Science series: Larkin held that no known force could start a primeval nebula rotating [web: archive.org jstor-2900594]; the 1903 electrical-induction answer is his reply to his own question of 22 years earlier.
PASP 14 (1902) 26: independent editorial record that the stellar-floor programme ran through 1901 and that his longer report was "too long to reproduce" and not printed [web: https://archive.org/details/jstor-40668215]. "Orion is always bound in white ... other portions of the celestial vault are jet black."
Letter to Hallock (c.1906): "nothing exists but electricity ... I allude to 'refined' matter in my book 'Radiant Energy.'" and Within the Mind Maze (1911): "nothing exists but electrons" [web: iapsop.com/ssoc/1906__hallock___luminous_bodies.pdf]. Shows where he took the corpuscle idea; its reception was in the spiritualist press (parties with an interest in that reading).
Tesla witness, 1916 book (single published account, date and place not given).
Birkeland's terrella (1896) and Arrhenius's radiation-pressure aurora (1901) are rival contemporary theories; both better than Larkin's on the sun-aurora link (below).
Critic's characterization: "The Earthquake and Professor Larkin", Science 24 (1906) 178-180, signed D. S. J., ridicules his Open Court earthquake article (animal behaviour, sulphur-smelling fish, blue flames at Petaluma). It is an institutional critic's view of Larkin as a popular writer; it tests none of the book's claims. Earthquake-light reports are now a studied topic [memory: medium], so ridicule without test is not a refutation either.
Reception: reviewed in the LA Herald by a self-declared non-scientist; William H. Knight "critically examined the work before publication" [web: https://cdnc.ucr.edu/?a=d&d=LAH19030405.2.366.13]; sold into the spiritualist market; USC calls him "among the most-read scientists of his time" with about 9,000 bylines [web: libraries.usc.edu]. Biography dates for Lowe directorship: 1 Jan 1901 (own words), not 1890.
Absence of evidence: no scientific rebuttal of the book's astrophysics has been found, and none of the stellar-floor observations has a second observer. Competing explanations for the silence: the book was a popular work not aimed at specialists; Larkin's reputation as a popularizer (and the 1906 episode) made specialists ignore it; the stellar-floor claim was visual and unmeasurable by 1900s methods. Records that would decide: the full 1901 ASP report, Knox College / New Windsor logs, the Feb-Aug 1902 Examiner series, any Lowe plates.

Claim 1. All radiant energy is one electromagnetic wave motion, carried by Thomson's corpuscles that saturate space and replace the ether; radioactive bodies and enzymes only direct this energy from space

a. Mathematics: none is offered. Qualitative; internally he says corpuscles are negative and bent by magnets, so "the entire wave motion of the universe is electro-magnetic" [read: p.19]; but a transverse wave carried by corpuscles "vibrating like a pendulum" across the ray is a mechanical picture, not Maxwell's. b. Experiments: (i) Heat, light and "actinic" rays being one radiation: Melloni 1843, Langley, Hertz 1888 [read: pp.177-178]. Strongly consistent. (ii) Thomson's corpuscle (electron) is real. (iii) A corpuscle sea as the wave medium: his own primordial density gives a number. Spreading 10^8 suns as corpuscles (mass 1/1000 of H) through his 30,000 ly sphere gives about 10 per cm^3, spacing about 0.5 cm (calc), not "inches or even feet" [read: p.280]; the same mass as hydrogen gives spacing about 5 cm, not "miles" [read: p.280] (calc). He gives no present-day density. (iv) Radioactive bodies "simply more transparent to corpuscles from space" [read: p.18]: Rutherford and Soddy's 1902-03 transformation theory and decay half-lives, and Curie and Laborde's 1903 radium heat measurement, treat the energy as stored in the atom [memory: high]. Larkin hedges nowhere on this, but in 1903 the energy source of radium was a live puzzle, and an external-source idea was not absurd to contemporaries (Kelvin is reported to have entertained one [memory: low]). c. Tested: radioactivity, yes (half-lives constant whatever the surroundings; the energy per decay traced to nuclear mass differences [memory: high]); no one tested a corpuscle medium as such. No replication needed for the unity of radiation. d. If true: relativity's "no medium needed" would fail, and a corpuscle sea would need drag, scattering and dispersion limits. Free electrons do disperse radio waves (pulsar dispersion), but at about 0.03 per cm^3, not 10 [memory: medium]. Verdict (split): unity of heat, light and actinic rays as one electromagnetic radiation: Established. Corpuscles = electrons exist: Established. Corpuscle sea replacing the ether: Historically superseded but instructive (the ether was dropped, not replaced by a corpuscle gas). "Radioactive bodies only direct energy from space": Tested, not supported. Inventor's rigour: the density statements are asserted, not computed, and contradict his own numbers by about 10^3 to 10^10.

Claim 2. Solar jets and sunspot bridges disturb the earth's field by electromagnetic waves at light speed, causing magnetic storms, earth currents and aurora; the earth's field is "tuned" to the sun's; magnetism is a vortex in corpuscles fixed in steel; Crookes and Geissler tubes reproduce the aurora

a. Mathematics: none. The only numbers are the current tables. Boston peak 133 (text) vs 144 (table) mA; volts column = I x 3,700 ohm against stated 2,700 [read]. Providence: 644 V / 0.280 A = 2,300 ohm, other rows 2,200-2,800, so the ranges drift. "Dipping needles are inclined by an angle equal to the magnetic latitude" is only roughly true (tan dip = 2 tan latitude) [memory: high]. b. Experiments, strongest case for Larkin first: - Earth currents of hundreds of volts and over 100 mA on long telegraph lines in great storms are real and independently recorded for this very storm [web: Google Books snippets]. His 1892 report of a quadruplex-wire disturbance and "counter electro-motive force" is what geomagnetically induced currents do (changing field induces a voltage in a long wire) [memory: high]. His p.132 phrase "the currents induced in the earth ... by the sun's disturbed magnetic field" is, with the causal chain corrected, the modern picture of induction. - A light-speed solar effect on the earth's field is real in one form: a flare's X-ray and ultraviolet burst, arriving in 8 minutes, enhances ionospheric currents and gives a small "crochet" in magnetometers. At Kew the 1859 flare was matched by one [memory: high; the Kew crochet is not in the page I fetched]. So his same-hour timing observation (tongue brightest as the telegraphs were disturbed, 1892) is the right kind of evidence for that part. - Lab aurora: Geissler and Crookes tubes do imitate auroral light, because aurora is gas discharge excited by energetic electrons [memory: high]; a magnet changing the colours/forms matches Birkeland's terrella work (1896+) [web: Wikipedia Terrella, via Investigator]. Against: - Timing of the main storm. The accepted 1859 event had flare and storm onset about 17.6 hours apart [web: https://en.wikipedia.org/wiki/Solar_storm_of_1859], a coronal mass ejection (plasma cloud), not a light-speed wave. Larkin cites 1859 ("that evening") yet claims 8 min 19 s [read: pp.125, 128-129]. His own Nov 1882 jets were drawn on 19 Nov, two days after the 17 Nov storm [read: pp.126, 128]; he writes that great spots occur without storms and storms without spots [read: p.128]. - His own demonstration cuts against his mechanism. The tubes glow when current is passed through the gas: a particle-discharge effect, not light from a passing EM wave. Birkeland read the same laboratory result as evidence for particles from the sun. - "No new mode is seen in the auroral light" [read: p.139] was false in 1903: the green oxygen line (557.7 nm) was not reproduced in a laboratory until McLennan and Shrum in 1925, and even then disputed (Vegard 1926) [web: nature.com/articles/115047a0; JSTOR Daily]. - "Tuned" earth and sun fields [read: p.137]: no quantity, no test. Magnetism as "a vortex in corpuscles", steel fixes it, iron does not [read: pp.130-132]: the steel/iron difference is real (hard vs soft magnets) and is explained by domain pinning (Weiss 1907) [memory: high]; the vortex picture adds nothing testable. c. Tested: solar-geomagnetic correlation (Sabine 1852, Carrington 1859) and the timing question were tested by many observers over decades; the chief replacement (Chapman and Ferraro 1931, plasma clouds) is a particle theory [memory: high]. Never a direct test of "tuning". d. If true: electromagnetic waves from the sun would reach the earth at storm commencement with the storm following in minutes; the magnetosphere would act as a tuned receiver; both fail the timing record. Verdict (split): correlation of solar activity with magnetic storms, aurora and telegraph earth currents, and the measured currents: Established. Main storm caused by light-speed EM waves from jets: Tested, not supported (the part that is light-speed, a small crochet, is real). "Earth's field tuned to the sun's": Not testable as stated. Magnetism as vortex in corpuscles: Historically superseded but instructive. Tube-imitation of aurora: Established as a qualitative demonstration, but the "all colours, no new mode" claim is Tested, not supported (green line). How rejected: no one rebutted it by name. It was overtaken by the particle theories of Birkeland, Arrhenius and Chapman; I found no data suppressed or reanalysed.

Claim 3. The sun's heat comes from gravitational contraction from a 25-trillion-mile sphere at absolute zero (Helmholtz and See: 49,048,200 degrees F water-equivalent, about 21.8 million years); meteoric impact is limited to about 1%; catalysis may have been the first mode of energy

a. Mathematics: mostly sound. The Helmholtz energy (3/5)GM^2/R = 2.3 x 10^41 J, as a water-equivalent 4.9 x 10^7 F, matches See's 49.05 million F within 0.4% [read: deep read]. The 1/100 Earth mass falling for one year works out (0.94 year at modern luminosity). Errors: Stefan's law in ordinary degrees, Lane's 16x vs 4x, 18+4 = 22 vs 21.8, "reduce its temperature" vs Lane, 27,855 and 81,154,000 F off by 3% and 1.3%, and a 1,475 lb that should be about 2,467 [read]. b. Experiments: the calculation gives, with modern luminosity, about 19 million years for a uniform sun (calc) and about 9 million with the virial half-share [web: https://en.wikipedia.org/wiki/Kelvin%E2%80%93Helmholtz_mechanism, about 8.9 Myr]. Against the Earth's age: radiometric dating from Rutherford and Boltwood (1904-07, hundreds of millions of years) and Patterson (4.55 billion) [memory: high], and solar neutrinos (Davis, then SNO) as a direct detection of fusion [memory: high]. Kelvin's own contraction needed about 9 in/day diameter shrinkage (Larkin's figure, 8 in/day by calc), too small to measure in 1903 and still of the order 0.01 arcsec per century. c. Tested: indirectly, as above; the failure was found by geologists first (the age conflict with Darwin and Lyell was debated for 40 years), which Larkin knew, since he calls the contraction time "probably not far from the truth" [read: pp.179-180]. d. If true: the Earth would be about 20 million years old; the geological and radiometric record would be wrong. Strongest case: Larkin hedges: contraction "all the heat ... assumption", and "some other and mighty cause must be sought" (p.165, text of 1882); catalysis "offered for what it is worth, if anything" [read: pp.165, 175]. In 1903 contraction was the mainstream view, not a contrarian one; the real contrarians were the geologists (Chamberlin 1899 argued for unknown energy sources [memory: medium]). His "enzymes and radioactives only direct energy" is a close cousin of the wrong turn in claim 1. Verdict (split): contraction as the sun's main source: Historically superseded but instructive for its time; as a present-day statement of fact, Refuted by calculation (about 10^7 yr vs 4.6 x 10^9 yr). The 1% meteoric bound: Plausible, untested in his text but consistent with later knowledge. Catalysis as the first energy: Not testable as stated. Inventor's rigour: sound figures, unsound physics; he treats a water-equivalent budget as a cooling rate and ignores the virial theorem.

Claim 4. Cosmogony: a corpuscular cloud 30,000 ly across; gravity made nebulae; electrical induction set them rotating; gemmation and tidal friction (Darwin, See) built suns, planets and moons ("completely accounts for the evolution of the sidereal universe"); dead worlds (about 11 quadrillion earth-like bodies, one live sun in 320) and a cycle back to corpuscles

a. Mathematics: the book gives no equations for tidal evolution; the 3-5 h first day and 55-day final state are Darwin's results, not derived. My conservation-of-angular-momentum calculation gives an inner synchronous state at 4.9 h and an outer at 47 days [read: deep read]. Fall-to-centre arithmetic (25 mi/s, 2, 8, 32 billion suns, 11 quadrillion) is correct on his premises [read: deep read]. Gaps: speed above escape speed need not mean a bigger mass; the 1,000x scaling of Kelvin's limit should be 10^9 (cube) [read]. b. Experiments: - Earth-moon tidal recession (about 3.8 cm/yr by lunar laser ranging) and tidal circularization of close binaries are established [memory: high]; this is the sound kernel, and See's idea that tides shape double-star orbits survives in that form. - Electrical induction spin: his croquet-ball charging by induction is correctly described, including the sign [read: deep read]. For nebulae, electric force would equal gravity for a charge imbalance of 1 part in about 10^18 (calc: sqrt(4 pi eps0 G) = 8.6 x 10^-11 C/kg, one electron per 10^18 H atoms); in an ionized gas the Debye length is 0.2-7 m (calc), so net charge is shorted out. No nebular charge is measured. Honest credit: the underlying idea, a passing neighbour spins a lopsided cloud by a couple and the spin amplifies on collapse, is conserved "areas" and survives as the gravitational tidal-torque theory of galactic spin [memory: medium]; Larkin writes "the gravity and inductance of some other nebula" [read: p.290]. - Missing mass: his Groombridge 1830 argument (a 200 mi/s star implies lots of unseen mass) has the right modern conclusion. He counts only 10^8 plate stars; the Galaxy has about 10^11. A rotation speed of 220 km/s at 15,000 ly implies 5 x 10^10 solar masses (calc; speed [memory: medium]), the same order as his 3.2 x 10^10, and Groombridge 1830 is a bound halo star [memory: medium]. But the invalid step is "no parallax in Newcomb, so unmeasurable"; and the dark matter is not "dead suns" (microlensing surveys excluded compact dark halos as the main component [memory: medium]). - Fission of the moon and of suns: the moon-fission idea gave way to giant impact; the pear-shaped (apioid) path to binary fission was shown unstable by Lyapunov and Jeans [memory: medium]; stars are formed by fragmentation of clouds. c. Tested: tidal evolution of Earth-Moon, yes; induction spin and gemmation of suns, no experiment; the sphere and cycle, no. d. If true: star formation would be a tidal-fission process driven by large nebular charges; plasma physics would need exceptions. Verdict (split): Earth-Moon tidal evolution, tidal friction in binaries: Established. Tidal friction "completely accounts for" the universe, moon and sun gemmation: Historically superseded but instructive. Nebular spin by electrical induction: Refuted by calculation (charge neutrality, above); the gravitational-torque cousin: Established in outline. Universe 30,000 ly across with 10^8 suns, primordial uniform corpuscle cloud and the cycle back to corpuscles: Historically superseded but instructive. Large unseen mass: Established (in modern form); identified as dead suns: Tested, not supported.

Claim 5. A luminous "stellar floor" lies beyond the Milky Way, seen from the Lowe Observatory with over 100 rents in Scorpius and Sagittarius and "starless fields" (tubular empty lines of sight); a year of tests ruled out atmospheric dust

a. Mathematics: the positions are given (page-image tables [read: p.301]); "tubular spaces" is a geometric inference. No photometry. b. Experiments: - Observations that fit real objects: the starless-field positions cluster in Ophiuchus/Scorpius/Sagittarius, the richest region of dark nebulae. Field 2 (17h 25m, S 26 35) lies about half a degree from the centre of the Pipe Nebula (17h 27m, -26 56) [web: https://en.wikipedia.org/wiki/Pipe_Nebula; offset calc, epoch of his coordinates not stated]. Barnard catalogued 349 such dark markings in 1919 [memory: high] and Herschel noted "holes in the heavens" [memory: high]. Orion's real extended nebulosity (Barnard's Loop) may be his "Orion bound in white" (Investigator lead) [memory: high]. - His same-night contrast test is a legitimate control against uniform dust haze, and he reports openings "not jet black within", "dark by contrast" [read: pp.300, 302], i.e. he states the weakness himself. Against: - A pavement of unresolved extragalactic stars beyond the Galaxy is far too faint to see: the extragalactic optical background is about 1-24 nW m^-2 sr^-1 [web: arxiv.org/pdf/1905.08825 and 2101.01500 snippets] against a dark sky of about 22 mag per square arcsecond, roughly 1,500 nW m^-2 sr^-1 (calc; the 22 mag figure is [memory: medium]). So 1% or less of the sky glow. - The visible diffuse glow has known sources: airglow (patchy, varying within a night; recognized after 1900), zodiacal light, diffuse galactic light, unresolved stars [memory: medium]. His dust test did not address airglow or faint-star glow. - Visual, one observer, no counts, no second observer, full report unprinted [read: p.300; web: PASP 1902]. c. Tested: not by anyone else. Never dismissed with a published argument; the 1906 Science critique is about the earthquake only. His 1910 Scientific American "abyss in Orion" repeats it (later quoted in religious literature, a reception history). d. If true: a solid or stellar boundary beyond the Galaxy would need explaining (Olbers's paradox); a long empty tube would be a statistical oddity. Verdict (split): dark patches in Scorpius/Sagittarius against a glowing background: Established (as dark nebulae and background glow). A distinct pavement of distant stars/nebulous envelope as the source of the glow: Refuted by calculation for the extragalactic version; as a visual observation of a "floor", Never properly tested (not dismissed; unmeasured and unreplicated). Tubular holes: Historically superseded but instructive.

What would settle it

Claim 2: Kew/Greenwich magnetograms and Greenwich sunspot records for 16 July 1892, and the telegraph records, against Larkin's 10.30 a.m. and 9.30-10.30 a.m. times; plus the 15-19 Nov 1882 magnetograms against his drawings.
Claim 5: a photometric sky survey (Sloan, Gaia and modern all-sky maps) of his seven "floor" and eight "starless" positions [web: arxiv.org/pdf/2101.01500], plus the Lowe plates or ASP report if they exist.
Claim 3 and 4: the calculations above; example worth running: U/L for the sun, and the charge-balance bound.

Open source gaps: Examiner 1902 series, ASP full 1901 report, Knox College logs, the Open Court July 1906 article (not found), HathiTrust full text of The Electrical Engineer vol. 14 p.79.