Flat-Earth: Science Contrarian review (Step 2)

Role: Science Contrarian. Standard: the same PhD-level rigour as the physicist and mathematician. Judgement rests on experimental evidence, not on consensus or reputation. Mallove is a model for the standard, not a persona.

Tags: [read] = the project's documents or deep reads (page cited where useful); [web] = Internet source with URL; [memory] = recalled, with confidence; [calc] = my own arithmetic (Python, R = 3958.8 mi, k = refraction coefficient, apparent drop = (1 - k) x geometric drop).

Coverage note. I read all of facts.json's deep reads (Rowbotham 1865 and 1881, Hampden 1869 by summary, Carpenter, the Zetetic Review, Blount and Smith, Dubay, the TFES pages on Bedford, Sinking Ship, Airy, Foucault, Coriolis, RLG, MGP, Universal Acceleration, Electromagnetic Acceleration, WGS84, the NGA standard, Foucault 1851, Oldham 1901, Wallace 1905, Hirt 2010, the Koreshan survey, NASA TM X-2514 and CR-4568, Skiba's Chicago, FEIC and balloon videos), claims.md, sites.md and all ten Investigator notes. NASA RP-1207, CR-62898, the PTTI paper, Michelson 1925 (I and II), the DMA text, Voliva items, the Lead Stories fact-check and Hampden 1869 I took from facts.json summaries and the first pages of their deep reads, not line by line. Skiba's videos are auto-captions: numbers may be misheard. No Source gap for the key works.

0. The contrarian's frame for this project

The project is unusual: it asks me to take seriously a position that almost every working scientist regards as closed. The honest way to do that is to separate five different things the documents mix together.

1. Observations that are real and repeatable (the Old Bedford sightings, Chicago's skyline over Lake Michigan, lighthouse ranges, ring-laser beat frequencies, pendulum precession). These are evidence and are treated as such.
2. Arithmetic (the "8 inches x miles squared" rule and what it is applied to). This can be checked line by line.
3. Interpretation (what the sighting means: refraction, perspective, curvature, a plane).
4. Institutional claims (agencies conceal, test results were suppressed).
5. Theology.

The strongest honest case for the flat-earth claims lives almost entirely in group 1 plus one unsettled physical question: how large and how variable is refraction over a few metres of air above water, and was it ever measured at the time of a Bedford-type test? That question is real. Nothing in the folder or the Investigator's finds answers it for any of the historic Bedford tests (Rowbotham 1838-1870, Wallace 1870, Oldham 1900-01, Blount-Clifton 1904). I will say so for each.

The case against is also nearly all experimental, and it is stronger than the popular "obviously round" framing suggests, because (a) the one pre-agreed, both-sides-signed test (Wallace 1870) came out for curvature, (b) the quantitative flat-earth case contains repeated arithmetic mistakes that all push the same way, (c) a plane cannot supply a sunset, southern constellations or a sun-altitude table at all, and (d) several of the "NASA admits" and "WGS84 is flat" claims fail on the sources' own words.

1. Corrections to Step 1 (and to the Investigator)

Per the corrections rule, plainly:

1. TFES Bedford deep read (deep/Related-Research-2022-TFES-wiki-Bedford-Level-Experiment.md), "Rowbotham, 1838", check and "Points of note" item 4. It says the wiki's "eleven feet" is "not derived" and that the method gives "about 16.7 ft, not 11 ft". That is wrong. 8 in x 25 = 16.7 ft is the hidden height of the waterline at the far end after subtracting the 1-mile horizon of an 8-inch eye; the top of the 5-ft mast is 16.7 - 5 = 11.7 ft below the line of sight. The wiki's "some eleven feet" is right for a 5-ft mast [calc: 11.67 ft at k = 0; 9.1 ft at k = 0.13; matches Rowbotham 1881 Exp. 1, deep/1881 section 1]. The same file's check of the wiki's "more than 20 feet" for Clifton's sheet also needs fixing: for an eye 2 ft above water, sheet bottom at the water, 6 miles, hidden height is 12.1 ft at k = 0, 10.0 ft at k = 0.13 [calc], not "about 15 ft". The wiki's "more than 20 feet" is the drop below the tangent at 6 miles without any horizon allowance (24 ft), which is the wrong figure for this geometry. Net: the wiki overstates the hidden height by about a factor of two; the right figure is smaller and so makes the refraction explanation easier, not harder (see claim 1).
2. claims.md, Corrections, Wallace bullet ("he received £500 and repaid it"). Incomplete. The umpire's printed decision proposes to pay Wallace "£1000", i.e. both stakes [web: The Field, 26 March 1870, https://people.wku.edu/charles.smith/wallace/zCarpenter1870Field1.pdf, via research/2026-10-10-wku-wallace-online.md]. Wallace's "£500 I received from Hampden" (My Life, p.15) is Hampden's share of that £1,000. Hampden's leaflet charge that Wallace "obtained possession of a cheque for £1,000" (Wallace p.13) is therefore the whole pool, not an unexplained second sum. The Wallace deep read's remark that the £1,000 is "Hampden's claim; Wallace does not confirm or explain" is resolved by the Field text.
3. sites.md "Leads", Walsh "was stakeholder and Wallace's nominee". Soften. Wallace says he first proposed Walsh as referee, Hampden later agreed that Walsh be umpire (Wallace, My Life, p.4 and p.9 [read]); Walsh's letter says he was "authorized in writing by Mr. Hampden" to decide (Epsilon WCP5561 [web: https://epsilon.ac.uk/view/wallace/letters/WCP5561]). "Wallace's nominee" is Hampden's later characterisation. It stays in the record as Hampden's complaint; it is not established as fact.
4. claims.md claim 2 counter-records, "press reports of the Knodel gyroscope drift". Say what it is: a documentary-press report (Earth.com, Kottke) that two flat-earth proponents measured a signal of "15 degrees per hour" with a roughly $20,000 laser gyroscope; the number is the sidereal Earth rate (about 15.04 degrees/hour), not an instrument fault; Kottke says later videos tried to debunk it. No instrument record or paper is in the folder. It is single-source press evidence of a detection by people who wanted a null; I weigh it as such (claim 2).
5. Dubay deep read, section 3b ("target's own horizon not subtracted"). Correct but understated. The pattern is systematic: Dubay's hidden amounts reproduce 8 in x (d - eye horizon)^2 and omit both refraction and the target's own horizon (items 68-79), which is why items 78-79 (17,400-ft and 20,320-ft peaks) come out as 7,700 and 9,200 ft hidden when their own horizons place them in plain view. Items 81-93 repeat Rowbotham's 1865-81 table and inherit its errors (see claim 1). No change of verdict by the earlier reader; recorded because the arbiter should cite it.
6. Rowbotham 1865 vs 1881 (deep reads). One point the earlier readers did not join up: the 1865 flag was 3 ft with the observer "seated in the water" and the 1881 flag is 5 ft in the opposite direction, both described as the 1838 observation. Combined with the 1872-preface admission that some first-edition evidence was "not entirely satisfactory" (1881, p.v [read]), this is an inventor correcting himself, which is to his credit; but it also means the central 1838 observation exists in two inconsistent versions by one author (claim 1).
7. Investigator, Clifton letter ("a witness on the inventor's side, not an interested party"). Clifton was Dallmeyer's operator hired by Lady Blount; the zetetic account calls him "a globularist" (Blount 1905 [web via sites.md]); Watson's English Mechanic letter (80:160) says the negative was taken "on behalf of J.H. Dallmeyer, Ltd" for Watson (Bedford deep read). Which person commissioned the photograph is therefore reported two ways. The text of Clifton's own letter quotes him conceding only "unless rays of light will travel in a curved path". He is a witness of what he saw, not of what it means; his own parenthesis names the refraction explanation.

No correction changes a central verdict; items 1 and 5 make the quantitative case against the zetetic arithmetic slightly stronger.

2. Web verification (budget 6; four used)

Remaining [memory] items that a verdict leans on are tagged at point of use.

3. Claim 1: "Water is level and the earth shows no curvature"

Mathematics (a)

Rowbotham's rule is the standard small-angle result: drop below a tangent at distance d is d^2/(2R), which is 8.0 inches per square mile [calc, R = 3958.8 mi; confirmed in deep/1881 section 0]. It is correct for the geometry in which it is used: eye at a small height, target at distance, subtract the eye's horizon first. For Experiment 1 (eye 8 in, flag 5 ft, 6 mi) the flag top is 11.7 ft below the line of sight at k = 0 [calc]. His arithmetic is mostly right (deep/1865 section 7: exceptions are Whitby 102 for 112, Lower Farn 67 for 21.7, and the Grimsby refraction step 206 for 216.7).

Recurring errors, all in the same direction (they enlarge the "missing" amount): - Applying the refraction allowance to the wrong quantity: 1/12 "of the altitude of the object" rather than a fraction of the curvature drop (1881, p.31 and 34; 1865, p.179). On his own premise the standard allowance of about 1/7 of the curvature would remove 14% of the drop, not a few inches (deep/1881 section 2). - Treating ambiguous "miles" as nautical whenever that helps (Spurn, Ballycotton, Grimsby: "always understood when referring to distances at sea"). - Treating tabulated computed lighthouse ranges (Findlay, which already include curvature plus refraction and 10 ft of deck) as observations. Eddystone, Calf of Man, Cromer: Findlay's 13 and 23 nmi equal the standard geometric ranges 13.3 and 22.5-22.8 nmi (deep/1865 section 7). Three British rows (Ryde, Bidston, Whitby) exceed the standard range; the reason (misreading, luminous vs geographic range, favourable refraction) cannot be settled without Findlay. - Confusing the drop below the tangent over the whole distance (24 ft at 6 mi) with the sagitta of the chord when both ends are at equal height (6 ft). Carpenter's referee report uses it ("24 feet in six miles" vs Wallace's 5 ft), and Dubay mixes both forms (item 9 uses 600 ft for a 30-mile canal, items 10-11 use the mid-chord value) (deep/Dubay section 3a, 7; sites.md Leads, Carpenter). - Skiba's 2016 "circumference of the 41.5 degree parallel is 18,630 miles so the drop is 1,622 ft" (Part 9): a sight line across a lake is a chord on a sphere of radius 3,959 mi; the length of the circle of latitude is irrelevant [calc; memory high]. Not a curvature test. - Rowbotham Exp. 13, Mont Fréjus: 8^2 x 8 in is 42.7 ft, then the formula is credited with 52 ft (deep/1881). Exp. 7 (levelled board on the horizon) treats the horizon as a straight 20-mile chord; the horizon on a sphere is a circle about the observer (deep/1881). - Koreshan: same 8 in x n^2 read concavely; the survey result rests on the sign, not the size.

Internal inconsistency inside the movement: the TFES wiki says Bedford diagrams show a plane; the Koreshan survey reads the same arithmetic as a concave surface; Blount-Smith 1904 say meridian degrees "converge" in the south and need refraction to lengthen northern degrees (deep/1904 section 6). The same sightings cannot be proofs of flat, concave and plane-with-converging-degrees at once.

Consistency with established experiments (b)

• Wallace 1870: three equal-height marks, both sides agreed beforehand what flat and convex would look like (Wallace p.5-6), and both referees signed the other side's sketches (p.6-7). Predicted rise of the middle mark 5 ft 8 in, "perhaps about five feet" with refraction; umpire found curvature "five feet, more or less" [web, Field 26 Mar 1870]. My check: the sagitta of a 6-mile chord is 6.0 ft at k = 0, 5.2 ft at k = 0.13 [calc].
• I verified Wallace's own logic and find Carpenter's reading ("equal steps below the cross-hair mean a straight line") wrong. For a level at A and marks B (3 mi), C (6 mi) at equal height above a spherical surface, B sits about 78 arcsec below the instrument's horizontal and C about 156 arcsec, so C appears one equal step below B [calc]. If the water were any straight line, level or tilted, B and C would be collinear with A and would coincide in the eyepiece. That conclusion does not depend on the cross-hair being exactly level, which answers Carpenter's complaint that the large telescope "was not levelled". It does depend on refraction being roughly uniform over the six miles, which nobody measured.
• Oldham 1900-01 (BAAS Section E, 1901): same design on the Welney-Denver reach, middle mark "about six feet above the line of sight", geometry 6.0 ft [read]. It omits instrument, trials, refraction, weather and whether the intended photographs were obtained; it is a 23-line abstract (deep/Oldham section 6).
• Hirt et al. 2010 [read]: near-ground refraction coefficient over grass at 1.8 m ranges from -4 to +16 in 800 m on sunny summer days, not the textbook +0.13; cited literature gives +1 to +18 about 5 m above water (Kabashi 2003) and -14 to +10 over ice. This is the key fact for the "strongest honest case" below, and equally the key limit on any single sighting.
• Koreshan survey (1897): see below.
• Modern geodesy (NGA STND 0036, GRACE-based EGM2008, GPS) assumes and measures an oblate ellipsoid with a 1/298.257 flattening (deep/NGA section 1). These are not independent tests of Bedford, and the NGA is an interested party.

Has it been tested; replication (c)

Tested, with sides, by: - Rowbotham himself, 1838, 1844, 1849, 1856, 1862, 5 April 1870 (the turf-boat telescope, 18 in, six miles, board bottom 6 ft 6 in). Witnesses: an unnamed boatman, an unnamed local gentleman, unnamed neighbours (1881 witnesses table: no independent named witness for any Chapter II experiment). The 1870 record is also in his own Experimental Proofs [web, saved in research/files]. No refraction measured; his barometer/thermometer/hygrometer comparison measured air at the two end stations at one height, which does not measure a vertical temperature gradient along the path (deep/1881 section 2). - Surveyors who challenged him with levelled theodolites: he reports that "in every instance" the horizon appeared below the cross-hair, and attributes this to lens collimation (1881, p.41). That is a reported experimental result against his position that he explains away without a test of the lens error beyond Heather's figure. - Wallace/Coulcher/Carpenter 1870 (sketches only; no numeric readings). - Oldham 1900-01 (curvature value). - Blount-Clifton 11 May 1904: a Dallmeyer photo-telescopic camera "less than two feet above the ground-level", a sheet at Bedford Bridge six miles away (Blount 1905; Schadewald: 15 ft square, bottom just above the water) [web via research notes]. The negative is said to show the whole sheet and its reflection. The negative is not in the folder; the English Mechanic 80:40, 160-161 exchange (HathiTrust) records an "abnormal refraction" reply the same year. - Koreshan Geodetic Survey (Naples, Florida, March-May 1897; Morrow, the Koreshan Unity; interested party). - Skiba, Lake Michigan, June 2016 (camera 3 ft, 42.6-46 mi; the lower part of the skyline missing; called "inconclusive" by Skiba himself); second balloon launch Nov 2017 ("a very slight curve ... hard to tell", lens distortion, clouds).

Replications: none of the Bedford tests was repeated with simultaneous refraction measurement. There is no neutral replication of the Rowbotham anomaly and no hostile-lab replication of Clifton's photograph in the record I have.

What would have to change (d)

If water were level over tens of miles, geodesy's arc measurements, tide theory, satellite orbits and the whole ellipsoid would have to be reinterpreted. If refraction instead explains the sightings, nothing in mainstream physics changes; the Bedford sightings become sensitive probes of near-water atmosphere.

The strongest honest case for the claim

1. The sightings are real. Wallace and Carpenter agree what was seen on 5 March 1870 (they signed each other's sketches); Clifton, a self-described globularist (Blount's word), reported the sheet visible and wrote that "unless rays of light will travel in a curved path" the six miles are level [web via research notes]. Skiba's meteorologist (Coomes) and optics professor (Rennie) concede that Chicago is visible over Lake Michigan from 50-60 miles through "looming" (clip, deep/Skiba Chicago section 5). All sides accept the object is seen when geometry says the lower part should not be.
2. The mainstream explanation is refraction with a large coefficient, and it was invoked rather than measured in every case. The 1904 English Mechanic reply, the BAAS abstract, Proctor's remark about Eddystone ("refractive power ... less efficient that morning") all assert it. Blount-Smith's retort ("they can blow both hot and cold", 1904 Preface) is a fair logical point: if refraction explains the sighting that was claimed against curvature, it is a free parameter in the sinking-ship effect too. A tested model needs the coefficient measured in the run.
3. The required coefficient is large. For Clifton (eye 2 ft, sheet bottom at the water, 6 mi) the bottom of the sheet is visible only at k about 0.9 (hidden height 12.1 ft at k = 0, 10.0 at 0.13, 4.2 at 0.5, 0.6 at 0.8, 0.0 at 0.9) [calc]. Rowbotham's Nab light-ship hull needs k about 0.46 (deep/1865 section 1). Hirt's inversion formula gives the local coefficient from the vertical temperature gradient: k = 0.5 needs about +0.05 K/m (an inversion), k = 1 needs about +0.13 K/m [calc from Hirt eq. 2, p = 1013 hPa, T = 288 K]. Cold water under warmer air makes such inversions; Hirt's own literature review gives +1 to +18 near water. So refraction at k about 0.9 is physically possible over cold, still, flat-calm water ("smooth as a molten mirror", Rowbotham 5 April 1870; Clifton's "shimmering vapour", Schadewald), but it is a special state, and it is a state that those experimenters selected by choosing still, clear days.
4. Rowbotham's own report that the visibility varies with water state (Nab: whole hull on calm days, none on rough days) is the signature of variable refraction as much as of waves.
5. The flat-earth side did the one thing mainstream commentary skipped: it published numbers that can be rechecked. Where those numbers are wrong (above) they can be corrected. That is how science should proceed.

How the claim was rejected

• Wallace's trial: sound in design (pre-agreed, three equal heights, level and telescope, signed sketches). Weaknesses on the record: sketches rather than measured offsets; no cross-hair in the large telescope (Carpenter's complaint); no refraction measurement; one week; and the umpire's decision rested on referees' reports that both sides signed. Hampden refused to look through either instrument ("trusted Mr. Carpenter", Wallace p.9). Hampden's complaint about the umpire was procedural; the court later ruled only on wager law (Hampden v. Walsh, 1 QBD 189, 17 Jan 1876: the agreement was a wager; the depositor could recover his stake if he demanded it before payment; no finding on the experiment or on Wallace's conduct [web: https://people.wku.edu/charles.smith/wallace/zAnon1877ReportsCases.pdf]). The TFES statement that a court found Wallace "cheated" is unsupported: the words "cheat and swindler" are Hampden's, in letters that were the subject of the libel proceedings (WCP6734 [web]). Hampden was convicted/imprisoned for libel and threats (Wallace p.10-12; Epsilon WCP1947: a week's imprisonment for want of sureties, July 1871 [web]; 1875 Chelmsford sentence per Wallace and Schadewald).
• Counter-claim, equally on the record: Wallace profited from a wager he later called "an ethical lapse" and spent 15 years in litigation; Wallace's account is the winner's, written 35 years later (deep/Wallace). Carpenter's account is the loser's referee's. Both are interested parties; the Field of 26 March 1870 prints both referees' reports and the umpire's decision, which is the better source.
• Oldham, Blount-Clifton: dismissed by assertion of refraction and, in Clifton's case, a published rebuttal; I find no experiment that retested Clifton's setup with refraction measurement.
• Rowbotham: no Royal Society or university response is recorded; Proctor's public reply to Carpenter's 100 Proofs was a postscript ("One hundred difficulties for young students of astronomy"); Baird (Smithsonian) said the proofs "have doubtless been thoroughly investigated by the inquisitive astronomers and geodesists of the last four centuries"; Johns Hopkins' Gilman is not recorded as replying to five letters (Carpenter 1885, pp.33, 37-38 [read]). That is dismissal without engagement, not tested-and-failed. Carpenter's practice of offering space only to those who "append his name" and calling non-repliers "cowards" is rhetoric, not method.
• Funding or publication blocking: nothing in the folder shows it. Rowbotham, Blount and the Koreshans published and lectured freely; the Review lists their own societies. The Investigator found no institutional archive of the Universal Zetetic Society.

Koreshan survey (variant of claim 1)

Facts [read]: a mechanically stepped "rectiline" from a tide-referenced 128-in datum; 1,045 adjustments over 2.5 miles; the remaining 1.6 miles to the water by a telescope and a boat hull; measured drops 8.02 in at 1 mi, 30.62 at 2, 75 at 3, 128 at 4 1/8 (table) against 8, 32, 72 (calculated); witnesses mostly colony members; no independent surveyor. Quadratic growth is the right discriminator between a tilt (linear) and curvature (quadratic).

My calculation (not in the book): a systematic bias common to every 12-ft step would also give a quadratic path. Reproducing a radius of 3,959 miles requires a cumulative angular drift of 12 ft / (3,959 mi) = 5.7e-7 rad per step, about 0.12 arcsec, or roughly 1e-4 inch of lateral offset at a 4-ft arm [calc]. That is the same size as the "0.0001 inch" the colony claims as its closure tolerance on the 228-ft trial. The "System of Reversals" cancels alternating cross-arm errors; it does not cancel errors that act identically on every step (sag, thermal gradient across the arms in beach sun, settlement of the 8-ft standards in sand, plumb error). The colony's stated checks (return over 0.375 mile within 0.25 inch; "settling" rebuttal) are consistent with, but do not exclude, such a bias. The last 1.6 miles are optical, with refraction "so small a factor" asserted, then used to explain why the line met the water at 4 1/8 rather than 4 miles. Conclusion for this variant: a serious, documented, never independently replicated self-test; original notebooks may survive (Koreshan Unity papers N2009-3, State Archives of Florida, Box 6 [web via sites.md]).

Verdicts, claim 1

• Rowbotham's quantitative claim that standard curvature is excluded by his Bedford, sea and lighthouse numbers, as presented: Refuted by calculation for the lighthouse table and for the several arithmetic errors (Whitby, Farn, Fréjus, refraction misapplication, computed ranges used as observations); for the 1864 Portsmouth/Plymouth sea observations the discriminating power is low (Eddystone "entirely out of sight" fits neither model without haze or swell; deep/1865 section 4).
• Rowbotham 1838/1870 and Clifton 1904 anomalous visibility (is the water surface a plane?): Never properly tested — there is no controlled repeat with refraction measured. Not dismissed without testing (several tests were run), and the tests that were run by the other side (Wallace, Oldham) came out for curvature. The anomaly has a calculated, physically possible explanation (k about 0.5-0.9 over cold calm water) that nobody confirmed on those days.
• Plane earth (no curvature at all) as a physical claim: Tested, not supported by Wallace 1870 (both referees' signed sketches show an equal-step, parabola-like profile), Oldham 1900-01, and the broader evidence in claims 2-3. Reading Wallace's design strictly, the test discriminates only if refraction was roughly uniform over six miles; that is plausible and not verified.
• Koreshan concave earth (8 in x n^2 concave): Never properly tested (independent) — self-test by the party with a stake; unreplicated; systematic-bias question open (calc above).
• Sinking ship effect as inconsistent, therefore not curvature (TFES): Not testable as stated; the page itself attributes the effect to swells, resolution and refraction and states no flat-earth prediction.

Investigator findings kept here: Hewitson's 1866 diary (neutral hearer of Rowbotham; self-luminous moon, tides, plane); Rowbotham's 1870 own pamphlet (procedural objection that the agreed three discs were not used, and his list of instrument errors); Lady Blount's 1905 account and Clifton's reported letter; Carpenter's Field report ("equi-distant in the field of view"); Walsh's decision ("five feet, more or less", after consulting the optician Solomons "at the request of Mr Hampden"); Hampden v. Walsh 1876; the libel record; the English Mechanic 80 exchange (pp. 40, 160, 161 read; 138, 169, 184, 205-209 unread); the Koreshan Flaming Sword March/June 1897 reports and the Florida archive; Ouse Washes local history (the bridge-name confusion: "Welney Bridge" is often used for the Old Bedford Bridge at Welney, which explains why Carpenter and Walsh name different ends; Rowbotham used a different stretch in April 1870); Goudey 1930 and Winship/"Rectangle" 1899 (not read; the Internet Archive misattributes Zetetic Cosmogony to Rowbotham).

4. Claim 2: "The earth is stationary"

Mathematics (a)

Internally consistent where it is quantitative, with slips. Examples: Blount's Review prints equatorial speed as 1,250 ft/s where 465 m/s is 1,526 ft/s [calc]; Schoepfer's pendulum table (20,000 vibrations in about 2,000 s) is incompatible with a clock pendulum (period 0.1 s), and Richer's 1867 should be 1672 [deep/1893 section 3]; Carpenter's proofs 42-44 and 54 treat orbital motion as uniform rectilinear motion; the Review quoting an RAS Fellow that Foucault's plane at latitude 5 degrees turns 1.307 degrees/hour (11.47 days per turn) is correct for 15 degrees/hour x sin(5 degrees) [calc] and, contrary to the editor's "JUST WHAT THE Flat-earthites SAY", shows an effect that is slow, not absent. The Review's thermal explanation for pendulum-length variation with latitude accounts for roughly half of the printed equator-to-pole difference by its own numbers (deep/1893 section 5.1).

Consistency with experiment (b)

• Foucault 1851 predicted the sine-of-latitude law (rate 360 degrees x sin(latitude) per sidereal day: about 11.3 degrees/hour, one turn in 31.8 hours at Paris [calc]); his note says only that the trace "does not make a full turn in 24 hours" and invites geometers to examine; no rate is measured in that note (deep/Foucault). His release method (burning a thread) is designed to avoid lateral impulse.
• The TFES page has no quantitative rate analysis at all and cites real, documented artefacts: Longden 1919 (wire anisotropy reversing the sense when the wire is rotated), Airy's difficulty starting a free pendulum, Dowden/Kitchell's elliptical motion, a South Pole station trivia page reporting rotation "at twice what is expected" before modification. These are honest systematic-error evidence. They show the pendulum is hard to do well; they do not show a rate that disagrees with sin(latitude) in controlled work, and the page does not claim to.
• Falling bodies: Reich 1831, 158 m, eastward deflection 2.84 cm vs 2.75 cm predicted, with large scatter and an unexplained southward component [web: https://en.wikipedia.org/wiki/Ferdinand_Reich]. Benzenberg's results scattered widely. The Coriolis page quotes both fairly and then draws the opposite conclusion from the one its own quotes support (deep/Coriolis section 8).
• Michelson-Gale-Pearson: observed 0.230 +/- 0.005 fringe vs calculated 0.236 +/- 0.002 [read]; I reproduced the calculated value (0.2363) from the paper's inputs [calc]. The TFES/Correa +0.26 mean and the "range -0.04 to +0.55" are not in the paper (the range may be single readings; Fig. 3 is not text). The paper's set sizes sum to 269 (11 x 20 + 26 + 23; set 6 is 26 on the page image; corrected per arbiter section 4 item 1, the TFES-page deep read gave 263). Both sets' means are all positive and within about 0.04 of 0.23.
• Ring lasers: G (Wettzell) beat frequency 348.6 Hz, which I reproduced from the area, wavelength and latitude [calc; web]. The TFES RLG page claims signal "pulled from noise" but proposes no mechanism for a Sagnac-type frequency split that depends on cos(angle to the rotation axis), and misreads the Brillouin-laser caption (flipping the axis N to S gives +15 to -15 degrees/hour, a change of 2 x 15; E-W gives near zero, as expected) (deep/RLG section 6).
• Airy 1871: the water-filled telescope null is what Fresnel's partial-drag coefficient (1 - 1/n^2) predicts for a moving Earth [web]; Jones's "dragging of starlight ... shows it is the stars which move" is not an inference the quoted sources make (Planck-Institute and Johnson quotes give Fresnel's aether explanation) (deep/Airy section 4).
• Michelson-Morley: the Zetetic Review era read it as "no motion". Mainstream reading: a null for a stationary ether. A null is compatible with a fixed Earth and with a moving Earth in special relativity; it does not by itself select either. The discriminators are annual aberration, parallax (Bessel 1838) [memory, high], Sagnac, MGP and the ring-laser beat. Dubay's item 18 describes Sagnac as a null; Sagnac 1913 reported a rotation-proportional fringe shift [memory, high] and the TFES RLG page itself says light's velocity "is indeed affected by detector motion".
• Coriolis/drains: Shapiro 1962 (a 6-ft tank, 30-millionths of g at 42 N, counterclockwise) and Trefethen 1965 (southern hemisphere; citation confirmed [web], outcome unverified here). The Veritasium trial (three trials per hemisphere) is anecdotal, as Di Liberto says. Artillery: the 1967 DTIC firing tables contain a rotation-of-Earth correction (Table H); the TFES page's reply (artillery misses by tens of metres anyway) is a dispersion argument it never compares with the table's size. Without the table values, this stays unsettled in the folder (unknown from the documents).

Tests and replication (c)

Foucault: many demonstrations since 1851, with the artefacts noted. MGP: one run (Clearing, 1923-25); the ring-laser lineage (C-II, G, UG) repeats the measurement continuously, though those papers are not in the folder; Knodel/Campanella report an Earth-rate signal with a commercial gyroscope (press, single source, flat-earth proponents). Replication by non-believers is routine in geodesy; replication of a null by flat-earth proponents is absent from the folder.

What would change (d)

A stationary Earth would require Foucault, MGP, ring-laser beat frequencies, Coriolis in meteorology and artillery, aberration and stellar parallax to be reinterpreted without rotation. The TFES pages offer Mach's principle (stars dragging the pendulum) as the alternative; that is a coordinate-choice argument, and it is honest to say a rotating-universe description in an Earth-fixed frame is kinematically equivalent for the pendulum. It does not predict the observed 348.6 Hz or the sin(latitude) law without a full set of fictitious forces whose values are the Coriolis ones.

Strongest honest case

Foucault-type demonstrations are fragile (Airy's own private view, Longden's wire effect); early press reports "to the horror of the spectators" show wrong-way results; falling-body tests scatter widely (Benzenberg up to nine times mean variation, per the TFES page); the MGP paper and TFES numbers disagree; ring-laser outputs are heavily processed. In a Machian or Tychonic description the Earth can be taken as the frame; what is then needed is not "no rotation" but "everything else rotates". Blount's own quoted "Dr. Schoepfer" lecture (1878, Scientific American) contains real pendulum observations; the tabulated heavier-ball slowing of the period is not in Galileo's theory and could be a real finite-amplitude/bob effect [memory, low-medium; not checked].

Rejection

Carpenter's Hopkins challenge, Smithsonian reply and Proctor's postscript are non-engagement. Airy's "fraud" remark about the pendulum is private (Conlin, quoted by TFES). For the physical claims the experiments were run and published openly.

Verdicts, claim 2

• Earth does not rotate (physical claim, in an inertial-frame sense): Tested, not supported (Foucault law; Reich; MGP 0.230 vs 0.236; ring lasers, 348.6 Hz verified by me).
• Earth does not orbit (Airy, Michelson-Morley as proof): Tested, not supported; the cited nulls are predicted by Fresnel drag or relativity, not by a fixed Earth alone, and annual aberration and parallax are not addressed by the sources.
• "Earth is at rest" as a coordinate choice (Machian/geocentric frame): Not testable as stated.
• The "1,000 mph / 19 miles per second is absurd" argument: Not testable as stated (argument from incredulity; Carpenter and Dubay treat orbital motion as straight-line speed, which is not what is claimed).
• "Coriolis never demonstrated" (TFES): Tested, not supported: the page's own quotations (Reich's mean, Shapiro, Trefethen, Table H) show demonstrations; the page contests their strength, not their existence.

Investigator findings kept: none were added specifically for claim 2; the Investigator searched for Knodel and Michelson-Gale material and recorded them as "next run" keywords (sites.md).

5. Claim 3: "Sun and moon are small and near; sunrise and sunset are perspective or curved light over a plane"

Mathematics (a)

Sun 32 miles at 3,000 miles subtends 32/3000 rad = 0.61 degrees (Rowbotham, Voliva); the real sun subtends 0.53 degrees; Blount-Smith give 30 miles at "six or seven thousand" miles (deep/1904), an inconsistent height. On the Blount-Smith figures Dubay's item 147 "equal size and distance" is also off. Perspective on a plane cannot produce a sunset at a finite height: a sun at height h = 3,000 miles over horizontal distance D has altitude atan(h/D); at D = 6,000 mi it is 26.6 degrees, at 10,000 mi 16.7 degrees, and it never reaches 0 [calc]. Its angular size at D = 6,000 mi is smaller by 2.2 times, which is not seen (the setting sun is not smaller; its size at the horizon is within about 1-2 percent of noon, ignoring the optical illusion, as Skiba's own pixel comparison "302 to 290" suggests, deep/Skiba Chicago section 9). Blount-Smith's reply (water droplets magnify the setting sun, "focussed image") is qualitative and unmeasured. G.M.'s equinox-noon test in the Review (altitudes 90, 60, 45, 30, 0 at latitudes 0, 30, 45, 60, 90) is self-consistent, and with the Review's own 69-mile degrees the implied sun heights are 3,585, 3,105 and 2,390 miles for latitudes 30, 45, 60, which do not coincide; 0 degrees at 90 N needs infinite height [calc; deep/1893 section 4]. This is a measurement-quality test the flat model fails on its own scale, and G.M. presents it as a challenge that no reply in the folder meets. Harpur's "unit circle" reply gives Bordeaux 93 million miles from the equator, which is a misuse of the diagram by both sides (deep/1893).

Electromagnetic Acceleration (EA): y = (3/4)(beta x^4/c^2)^(1/3) from a forum author; no value of beta, a stated validity limit (y much greater than x), and a 4/3 power that is not derived (deep/EA section 2). It cannot yet predict a number.

Consistency with experiment (b)

• Southern observers' reports (Revell, New Zealand 1893; "R.A." East Australia; "G.A." Auckland, 1904): sun rises north of east in winter, circumpolar southern stars, Southern Cross never sets. All of this follows from a southern latitude on a sphere (Revell's own noon altitude of "not more than 30 degrees" on 11 July at about 37 S is 31 degrees [calc; latitude memory medium]). The Review's own editors answered that light "travels faster in some places than the sun itself".
• Antarctic midnight sun: accepted by Blount-Smith 1904 with Belgica/Discovery evidence (an explicit change from the earlier line), but Dubay (items 56-58) says it contradicts a globe without calculating. The 1904 pamphlet's figure-8 sun path and two vortices fit a southern midnight sun but contradict the equinox-altitude table above.
• Moon: Blount-Smith's moon "not reflecting sunlight" and its light "turned towards the sun by currents" rests on one 29 Oct 1904 observation (the bright limb not pointing at the sun); the mainstream explanation is perspective of a great-circle terminator seen over a wide field. Both are observation-compatible without measured angles in the pamphlet (deep/1904 section 9). Skiba cites a friend's Theodolite-app triangulation giving a moon "less than 300 miles" away, second-hand and unpublished (deep/FEIC 2019, item 15). With a baseline of order 2,000 km, a moon at 480 km would show tens of degrees of parallax; the lunar parallax actually found would be about 0.3 degrees; the folder has no angles. Unreplicated and unrecorded; kept in the record as a single-source claim, no data.
• Eratosthenes (TFES): re-read as the diameter of the sun-lit region; the page's own arithmetic is off (500 x 25 = 12,500, not 12,250).

Tests/replication (c)

Rowbotham's sun/moon model is argued by perspective, not by a measurement program. No triangulation of the Sun from two stations that produces 3,000 miles is recorded in the folder. Skiba's "40 mi across, 3,000 high" repeats Rowbotham's numbers (deep/Skiba Chicago section 9).

What would change (d)

Optics (rays curving upward over thousands of miles at a rate that has no laboratory support), astronomy (parallax, eclipses, transits), and the whole dynamics of the sun-planet system.

Strongest honest case

The pre-1840 evidence for the Sun's distance spans 3 to 104 million miles in the quoted lists (Carpenter proof 60, Leo Castle 1896, Dubay item 123); the Sun's distance was uncertain until Venus transits in 1761/1769/1874-82 [memory, high]. Ptolemy-era eclipse prediction does not require the globe, as Blount-Smith note. Perspective is a real effect (parallel rails meet; sun-rays seen "converging" through cloud). Honest, but none of this yields a sunset.

Rejection

Dismissed by Proctor, Ball, Lockyer; replies are quoted by the Zetetics. No test of the plane sun by the other side was needed because the model fails its own geometry.

Verdicts, claim 3

• Sunrise/sunset by perspective over a plane: Refuted by calculation (altitude never reaches 0 at finite height; angular size would shrink 2x).
• Sun/moon 30-32 miles across at 3,000-7,000 miles: Tested, not supported (G.M.'s own equinox table gives inconsistent heights; southern observations are inconsistent with a single northern centre); the lunar triangulation claim by friends is Never properly tested (no data).
• EA (upward-curving light, beta undefined): Not testable as stated until beta and the full expression are given; the page itself says "not valid for short-range experiments".

6. Claim 4: "Gravity is not a pull from the Earth's mass"

Mathematics (a)

UA: 9.8 m/s^2 constant. After one year in the Earth frame, gt/c = 1.03 (Newtonian), so relativistic treatment is required; the page does it (Lorentz factor, limit c) but the equations are images (deep/UA section 3). The Equivalence Principle is local; the page's own words ("not possible to locally discern") mean UA and gravity are indistinguishable locally and differ non-locally. The Davis plane is listed as both an explanation of UA (repulsive exotic matter) and an alternative to it (finite attractive g) (deep/UA section 4). Tides need "Celestial Gravitation", which restores a small gravitational pull. Blount-Smith's moon speeds: their own calculation (2,281 mph) is correct, then added to the Earth's orbital speed to get "67,300 mph ... unthinkable", and compared with the quoted 10,800 mph; the claim that the moon "would be practically standing still" misreads relative motion; their "cycloid gap in the ellipse" is stated without computation (deep/1904 section 9).

Consistency with experiment (b)

• Measured gravity varies: NGA normal gravity 9.7803253 m/s^2 at the equator to 9.8321849 at the poles (0.53%) [read: Table 3.6, deep/NGA section 6]; UA predicts the acceleration of a body and its direction, not a latitude-dependent value, without extra terms.
• Cavendish and torsion-balance tests are dismissed by link captions only ("inconsistent short range attraction experiment"); no argument.
• Eötvös: the TFES page quotes the equivalence-principle tests (1 part in 10^9, Dicke 10^11) as showing "no effects from external forces" which is not what those tests measure (deep/Coriolis section 6).
• Planetary orbits, GPS orbit determination using GM = 3.986004418e14 (NGA table), and Neptune's discovery (1846) rest on the inverse-square law; Blount-Smith say the Uranus perturbation "might be accounted for by electric currents", without a calculation (deep/1904 section 4).

Tests/replication (c)

Blount-Smith: none. UA: none quoted; "Evidence for Universal Acceleration" is on another page not in the folder. The TFES claim "Gravimetry: gravimeters are seismometers" is a link caption.

What would change (d)

Newton's and Einstein's gravitation; the equality of inertial and gravitational mass; the orbit of every artificial satellite.

Strongest honest case

The equivalence of uniform acceleration and a uniform field is textbook (Einstein 1907) and the TFES page is right that locally they cannot be told apart. Newton himself wrote that action at a distance without a medium is "so great an absurdity" (Bentley letter, quoted by Blount-Smith, p.25 [read]); Blount and Smith were using Newton's own caveat. Tides and altitude variations are the places where the claim would be tested, and the page assigns them to a separate model.

Rejection

Dismissed in the pamphlets by mathematical authority; in the other direction the Zetetic literature does not compute a single orbit.

Verdicts, claim 4

• UA ("the Earth plane accelerates upward at 9.8 m/s^2"): Not testable as stated for the local equivalence; Tested, not supported for the non-local predictions (latitude dependence of g, tides, orbital mechanics). The model's internal tensions (Davis plane both ways; tides need Celestial Gravitation; "universe accelerating" without a stated inertial frame) place it close to Internally inconsistent as written, but because the page is a wiki collage I record the weaker verdict.
• Ether-current gravity (Blount-Smith 1904): Historically superseded but instructive as a Newtonian-era dissent (Newton's Bentley letter), Not testable as stated as physics (no equations).

7. Claim 5: "Agencies conceal the shape, and their own documents show it"

Mathematics (a) and consistency with the documents (b)

This claim is tested by reading. The Reddit and Steemit lists quote "flat, nonrotating earth" from about 17 NASA reports and Army documents. The documents say: - NASA TM X-2514 (Preisser 1972): "The method is limited ... to applications where a flat, nonrotating earth may be assumed" (four times); "This assumption limits the analysis to short-range, short-duration flights" (p.7); the report itself says the gyro reads Earth's rotation on the ground (p.10) and counts 0.25 degrees/minute as an uncertainty (p.19). The tangent-plane height error is d^2/(2R): 0.08 m at 1 km, 7.8 m at 10 km, 78 m at 30 km [calc, deep/TM X-2514]. The author can therefore say "flat" for tens of kilometres within the +/- 3 degree angle error he quotes. - NASA CR-4568 (Feeley and Speyer): spherical nonrotating Earth is the actual problem; "flat Earth" is the epsilon = 0 term of a perturbation expansion where the "spherical Earth model" is restored at higher order; the authors state the flat trajectory "deviated significantly from the optimal" (p.70) (deep/CR-4568). - PTTI 2007: a plane troposphere is one of three trial mapping functions, called "too large"; Niell's spherical-shell function is adopted (deep/PTTI). - RP-1207, CR-62898 and the others I weigh from their deep-read summaries only. - TFES WGS 84 page: quotes no WGS 84 definition or IS-GPS-200; its own sources describe a curved Earth ("best modeled as an ellipsoid, not a sphere", "curved surface of the earth", "orange peel"), and the page calls WGS84 both "an ellipsoid" and a "spherical coordinate system". The NGA standard defines a rotating oblate ellipsoid with a = 6,378,137 m, 1/f = 298.257223563 and omega = 7.292115e-5 rad/s; ITRF agreement is by adoption of IGb08, so it is not an independent test; and the standard states these are conventions in places (a is a 1976-79 value; omega a "nominal mean" while actual rotation "fluctuates") (deep/NGA). I verified 2 pi/omega = 86,164 s [calc]; the sidereal day is 86,164.09 s [memory, high]. - Lead Stories (7 Feb 2023): NASA public-affairs officer: the phrase is a simplification and curvature "did not affect the calculations in any meaningful way"; this is NASA's characterisation (interested party), but it agrees with the reports' own text, which I read. - TFES Aviation page: it admits it is "not specifically a disproof".

Dubay's claims on institutions (items 57, 162-178: green screens, composite satellite images, Freemason collusion, the Protocols of the Learned Elders of Zion (a forgery [memory, high])) name no record, witness or dataset except a documentary and his own film; Dubay sells the books and videos he promotes (deep/Dubay section 1).

CIA items in Skiba's FEIC talk: a 1948 geodesy study where "true coordinates ... are unknown to us" (B, L, H are unknown coordinates of surface points, a standard geodesy problem [memory, medium]), and a 1957 sky-brightness formula "on the assumption of a flat earth" (a plane-parallel atmosphere approximation [memory, medium]); the talk himself says "assumption", not "belief". The documents are shown, not quoted in full; I have only his reading.

Tests, replication (c)

No investigation of conspiracy has been done in the folder. Skiba's own conclusion (2016): "I cannot prove any conspiracy at work here"; he concedes "there is a lot of very convincing evidence coming from NASA ... I certainly cannot debunk much of it" (Part 9 [read]). His 2017 balloon: the team's narrator sees "a very slight curve", attributes it to lens distortion and clouds, and the footage is inconclusive (YouTube B0-J7OhcQUk [web via notes]; "Skiba Proves Globe DEBUNKED" 15 July 2017, where he concedes curvature in the footage at about 100,000 ft but argues the same shape appears at under 1,000 ft, which is barrel distortion, [read]). That null-result is the project's most honest self-test. The raw footage is said to be on the "Explosive Truth" channel, unchecked.

Absence of evidence

The README-style record that "the shape of the Earth is concealed" has no document in the folder that says the shape is concealed. TFES itself says there is no "Flat Earth Conspiracy" but a "Space Travel Conspiracy" (deep/Conspiracy page, via facts). Competing explanations for the lack of any witness: no concealment; compartmentalisation; mass participation. The first is the simplest; to decide one would need independent tracking data or insider testimony. I find none in the folder.

What would change (d)

Orbital mechanics, geodesy, every space agency's tracking, and the independent amateur-tracking community [memory, medium: not in folder].

Strongest honest case

The phrase "flat, nonrotating earth" does appear in dozens of NASA/Army reports; a lay reader meeting it is entitled to ask why. The caveats are real but the engineering practice of working on a plane with a frame correction later is in fact how the same agencies do short-range work, which is what the plane model's validity bound means. Lead Stories quotes NASA's explanation; NASA is an interested party.

Verdicts, claim 5

• "NASA/Army documents admit a flat, nonrotating Earth": Refuted by calculation — the plane approximation error budget (7.8 m at 10 km; 0.25 degrees/minute rotation term) shows the phrase is a modeling assumption, and the reports (TM X-2514 p.7 and p.10; CR-4568 abstract and p.70; PTTI "too large") say so themselves.
• "WGS 84 is a flat-map system": Internally inconsistent (the page's own quotations, as above) and refuted by the NGA standard.
• "Space travel is faked; the agencies conceal the shape": Never properly tested in the folder (no evidence offered that could be tested; assertions about green screens and composites); where specific testable photographic claims are made (cloud duplication, Photoshop) the author gives no method; overall closer to Not testable as stated for the conspiracy itself.

8. Claim 6: "Scripture describes a flat, stationary, enclosed earth"

This is a theological claim and I judge no theology. The physical entailments are claims 1-5. Kept in the record: Hampden 1869 ("world" over 200 times, "round" never), Carpenter proofs 17 and 50, Blount-Smith's Ten Commandments creation order, Skiba Part 9 ("The Bible absolutely ... is a Flat Earth book"; Isaiah 40:22 "circle"; "even if all I've said here is totally debunked, I'm still going to put my faith on the side of Scripture"), Voliva (the Bible "his entire scientific library", TFES). Skiba states his reasoning order openly: the scriptural commitment came before the observational test, and the test came back "inconclusive". That candour is worth noting. Hampden's postscript that if Wallace's experiment was founded on fact "the whole of the Scriptures are false" (Wallace p.13) shows the claim being used as a falsifier of the Bible, which is a theological position and not a physical one.

Verdict: Not testable as stated (interpretation of texts; many Christian traditions read them otherwise). It is a motive that explains why the physical tests mattered so much to this movement.

Investigator findings kept: Voliva's AP-wire record (Zion schools adopt a flat earth and "absence of gravitation", 18 Oct 1921); TIME 1933 (bankruptcy) and 1931 (world trip as a test); WCBD radio and the $5,000 standing offer ("nobody ever collected", Schadewald, critic's secondary source); the 1922 New-York Tribune Zion feature.

9. Rigour applied to the inventor as to the critics

Calls against the movement, with evidence: - Single-observer, undated, unwitnessed or unnamed-witness experiments (Rowbotham Ch. II, Carpenter proofs 5, 16, 36, 71, 78). - Computed lighthouse ranges read as observations; unit choices that always enlarge the claim. - Two versions of the central 1838 observation (flag 3 ft vs 5 ft; direction reversed). - Wrong geometry and mixed conventions (sagitta vs tangent drop; circle of latitude as radius; speeds summed as collinear). - Quoted nulls misdescribed (Sagnac; Airy; Eötvös; G-ring figure). - Conspiracy assertions without records; forgeries cited as evidence. - Dubay item 15: an airliner "drops 2,777 ft every minute" is an hourly average of a quadratic; the first minute is 46 ft [calc].

Calls against critics, with evidence: - Ridicule and non-engagement as a response to a published challenge (Gilman; Baird; Proctor's postscript). - Refraction invoked after the fact, with a coefficient that was never measured in the same run (Bedford, 1904-05). - Reading Oldham as confirmation without a stated instrument, trials or photographs. - The 5-ft sheet story told with a "more than 20 ft" hidden height when the right figure is 12 ft (the popular framing helps neither side). - Wallace's account is the winner's, 35 years later; the Field's contemporaneous report is better evidence.

10. What would settle each (single experiment or record)

1. Bedford: a controlled repeat of Wallace's three-equal-height design at three heights above the water, with simultaneous temperature profiles (two or more heights) and reciprocal sights from both ends, in both calm and breezy conditions, with a recorded refractive-index gradient so that k is measured, not asserted. Add a photographic target at night with a reference lamp. Cheap; the Old Bedford River still exists.
2. Koreshan: find the field record books (State Archives of Florida N2009-3), and re-run a stepped line of 12-ft sections on a non-tidal flat baseline with a closed return; a null (zero drift) would test the systematic-bias question.
3. Pendulum/gyro: a 3 m, 20 kg pendulum started by thread-burning, camera-logged plane angle at two latitudes, and a commercial ring laser at two orientations, publishing rates against sin(latitude).
4. Sun height: simultaneous altitude measurements at three latitudes on an equinox noon; the flat model needs one h; the Review's own table already shows it does not.
5. Records not obtained: the Clifton negative and Lady Blount's reproductions; English Mechanic 80:138, 169, 184, 205-209; Oldham's full paper if any; (the Field umpire's report of 26 March 1870 is now held); Skiba's raw balloon footage; Knodel's instrument data; the DTIC firing-table values (Table H) set against the stated dispersion.

11. Verdict summary

Overall: the experimental centre of gravity of the project is the Bedford family of tests. The honest contrarian conclusion is that the flat-earth case contains one genuine open question (how large refraction gets over calm water, and was it ever measured on those days) and no argument that survives the rest of the physics. The best single piece of evidence in the folder is the signed Wallace sketches; the best piece of flat-earth self-criticism in the record is Skiba's own "inconclusive" and Rowbotham's 1872 admission that some of his first-edition evidence was "not entirely satisfactory".

Corrections to earlier documents (for the orchestrator)

1. deep/Related-Research-2022-TFES-wiki-Bedford-Level-Experiment.md, check at "Rowbotham, 1838" and "Points of note" item 4: the wiki's 11 ft is right (16.7 - 5 = 11.7); Clifton's sheet is 12.1 ft hidden (10.0 ft at k = 0.13), not 15 ft.
2. claims.md, Wallace correction bullet: add that the pool was £1,000 (two £500 stakes) per the Field of 26 March 1870.
3. sites.md "Leads", Walsh described as "Wallace's nominee": mark as Hampden's characterisation; Hampden authorised Walsh in writing (WCP5561).
4. claims.md claim 2 counter-records: state that the Knodel 15 degrees/hour is the sidereal Earth rate reported by flat-earth proponents in press coverage, single source.
5. Trefethen 1965: citation verified, outcome not verified in this review; any summary that quotes it should say so.