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

Tyndall-John: Science Contrarian review (STEP 2)

Reviewer role: judges the seven claims in claims.md by experimental evidence, builds the strongest honest case, and examines how each was received. Tags: [read] from the project documents (deep reads, and raw text spot-checked by me), [web] with URL, [memory] with confidence, [calc] my own arithmetic. Source note repeated from claims.md: all five folder documents are Tyndall's own (an interested party on his priority and on how his results were received). No independent replication or critique is in the folder; the Investigator's finds fill part of that gap. Where a conclusion rests on [memory] alone I say so.

Corrections to Step 1

Mayer's number. claims.md (claim 6, and the "771.4" bullet) says "Mayer's method reconstructed to 771.4". The deep read already says 771.4 is Tyndall's reconstruction, but nothing says it is not Mayer's published value. Mayer's 1842 figure is 365 kilogram-metres per kilocalorie (about 3.58 J/cal), about 14 per cent below the modern value [web: https://en.wikipedia.org/wiki/Julius_von_Mayer ; search snippet of a Greifswald exercise sheet, https://physik.uni-greifswald.de/fileadmin/uni-greifswald/fakultaet/mnf/physik/BiOx/Blatt_14.pdf]. In Tyndall's units (ft-lb per lb per degree F) that is 365 x 3.2808 / 1.8 = about 665 [calc]. Tyndall's 771.4 (2,160 / 2.8, using air heat capacity 0.24 and the ratio 1.421) is therefore a modern-data reconstruction, within 0.1 per cent of Joule's 772 by construction, not Mayer's own result. I did not find a source that shows Mayer's exact 1842 arithmetic. Anyone quoting "Mayer got 771" is wrong. The source of the error is that Heat presents the calculation under Mayer's name; the 1.421 ratio is Mayer's but the 0.24 heat capacity and the pressure units are Tyndall's.
The "sun about 2" ratio (claims.md claim 3, from Muller via 1866 p.3 footnote). I computed ideal black-body ratios of invisible (>0.78 micrometre) to visible (0.38-0.78) power [calc, numpy]: 2,500 K 15.9; 2,800 K 9.3; 3,000 K 6.9; 3,500 K 3.8; 3,800 K 2.9; 5,800 K 0.93. The arc value 7.7 (8 by filter) corresponds to an effective source of about 2,900 K, which fits the glowing carbon points of an arc fed by 50 Grove cells and is physically unremarkable; the visible fraction 1/8.7 = 11.5 per cent also matches "one-tenth luminous" in the Rede lecture (black body at 3,000 K: 12.7 per cent). The solar value does not fit as well: a 5,800 K black body gives about 0.9, and at ground level the atmosphere removes infrared, so a ground-level ratio near 2 looks too high. Muller's curve (Fig. 2) and the introduction are not in the folder (journal p.1 missing), so I cannot check it. Treat "sun about 2" as an unverified figure of Muller's, reported by Tyndall, not as a result.
Claim 4 omits a contemporary objection that is in the record. Stokes to Tyndall, 13 Jan 1865 [web: https://epsilon.ac.uk/view/tyndall/letters/Tyndall2215]: "I am glad you have changed the term, which I don't think was quite correct", and heating a solid to glow by concentrated rays "is not a little different from fluorescence". claims.md states claim 4 as "the reverse of Stokes's fluorescence rule" without this. It should be listed beside the claim.
Strachey series: unresolved, not an error yet. Heat's raw text reads "of March 1850" (Heat text line 15097 [read]); the 1882 Bakerian lecture describes the same July 1866 Phil. Mag. paper as Madras data for 1841-44 (Investigator, archive.org note). My own search for the 1866 paper failed [web: no hit]. Both could be in the paper. Until someone reads Phil. Mag. July 1866, the place and date of the series quoted in Heat (place not stated in the extract) are unknown.
Small internal tension in the Investigator's notes. The Epsilon note quotes the editors as saying Tyndall "did not write or publish on the subject" between Nov 1864 and 1865, while listing Tyndall's 'On the History of Calorescence', Phil. Mag. 29 (1865) 218-31. The editors probably mean part of 1865; the month is not given. Check before citing.
Rede lecture versus Heat on water. claims.md claim 2 and the claim 6-adjacent text give "water 1/10 inch passes the sun's rays" as one of Tyndall's unsourced figures. The deep read shows the Rede statement ("half this thickness stops every terrestrial ray", credited to Melloni) is stronger than Heat's own data (water 0.27 inch opaque to a hydrogen flame; about 97 per cent of a lamp's radiation cut off at roughly 0.04 inch). I keep that comparison in the review below; Step 1 had it only as a deep-read aside.
Confirmed, no change: the gas-table corrections (1590, 6030, 372) are independently confirmed by the Wikisource transcription [web: https://en.wikisource.org/wiki/On_Radiation_(Rede_Lecture)]. I also checked the raw text for "ten feet" (On Radiation, lines 1319-1323) and "March 1850" (Heat) and the Mayer arithmetic (2,160 / 2.8 = 771.4 [calc]).

What the evidence base is

Tyndall's method is a differential thermopile (a stack of thermocouples whose voltage tracks absorbed heat) with a compensating second source, galvanometer calibrated against non-linearity, rock-salt windows (transparent to infrared), tubes 3-4 feet long [read: Heat deep read, ch. X]. Controls are better than a reader might expect: he moved the apparatus out of the laboratory, took air from seven country and seaside sites, removed the end plates and used an open tube, then removed the tube altogether (free-air cylinder, 5 and 10 degree swings), and he checked rock-salt plates for brine films [read: Heat deep read ch. XI, section 2.3]. Weak points are real too: a broadband source, an "air = 1" baseline that is itself a small noisy number, no error bars anywhere in the folder, and the "eminent men" who verified the work are mostly unnamed.

Claim 1: compound gases absorb radiant heat, elementary gases do not; radiating power equals absorbing power

a. Mathematics: there is little mathematics, only ratios. Internally consistent. One caution that is arithmetic, not algebra: every ratio is "times air", and air is set to 1 by definition. If the true dry-air signal were 0.5 or 2 units, every number in the table would double or halve. Tyndall's own humid-air ratios rise from 15 to 30, 40, 50, 60, 70 as the dry reference was purified [read: Heat deep read 2.5], which shows the denominator moved, not the numerator. The table is a ranking at one source temperature, not a set of constants. Nitrous oxide appears as "1800" in the prose and 1860 in the table [read: On Radiation deep read, section 1]. b. Consistency with experiment: strongly consistent with modern infrared spectroscopy. N2, O2 and H2 have no changing dipole moment in vibration, so they do not absorb infrared; CO, CO2, N2O, NH3, H2O, hydrocarbons do [memory: high]. Equality of emission and absorption is Kirchhoff's law (1859-60), established [memory: high]. Tyndall's reciprocity experiment (hot gas against a screened pile) is a direct test of it for gases [read: On Radiation deep read, section 8]. The one outlier is Macdonald's unrefereed viXra paper (2015) arguing N2 and O2 "are also greenhouse gases" because they are Raman-active [web: https://vixra.org/abs/1504.0165, author's claim, unrefereed repository]. Raman activity needs a changing polarizability, infrared absorption a changing dipole; those are different selection rules [memory: high]. Weak collision-induced absorption by N2 and O2 does exist [memory: medium], and is far below the compound gases, so "practical vacuum" stands as a statement of degree. c. Tests: Tyndall's own (1859 onward), Franz 1855 and Foote 1856 earlier on absorption by gases (priority question, Jackson 2019 [web: https://royalsocietypublishing.org/doi/10.1098/rsnr.2018.0066], via Investigator), Lecher and Pernter 1880 "supporting the lecturer in the matter of gases" [web: https://en.wikisource.org/wiki/Popular_Science_Monthly/Volume_20/March_1882/Sound_and_Radiant_Heat], Rontgen 1881/1884 (different method, pressure rise of the irradiated gas), Graham Bell and Tyndall's photophone sounds (1880-81), and a century of spectroscopy. Replicated by independent hands. Magnus (1861-62) and Buff dissented, mainly on vapours (claim 2); Franz found 3.54 per cent apparent absorption in dry air in a black-lined 3-foot tube [web: Very 1900, https://archive.org/details/atmosphericradia00veryrich, via Investigator], which is the kind of baseline artefact discussed above. d. If true: nothing in mainstream physics changes; it is part of it. Priority: Franz and Foote, and Tyndall's own date (first positive result 18 May 1859, RI journal MS JT/2/13/1158, not seen), are matters for historians; Tyndall's later accounts of his own priority are those of an interested party.

Verdict: Established (qualitative pattern and reciprocity). Split: the numerical table (972, 5,460, 6,030 and so on) is Established only as a ranking for his source; the numbers are not transferable, because they depend on source spectrum, window and baseline. Classification of the 1859-65 reception: tested and replicated for gases; the dissent was about vapours.

Claim 2: water vapour controls terrestrial heat loss (blanket); at least 10 per cent of terrestrial heat is intercepted within ten feet; CO2 only minor

a. Mathematics: the arithmetic checks. Humid-air table: observed 16, 32, 49, 64, 82, 98 against proportional 16, 32, 48, 64, 80, 96 [read: Heat deep read, page image p.388]; 62 of 780 units = 7.9 per cent [calc]; 80 x 200 = 16,000 per molecule is a product of two rounded numbers (and uses 80 where he measured 70-90 and 1 in 200 where he gave 0.45 per cent) and is an order-of-magnitude statement only. The step "at least 10 per cent of the earth's heat within ten feet" is an extrapolation: the measurement was a 4-foot polished tube and a laboratory source near 100 C, then stated as "no doubt" for the earth. The Rede lecture words it differently (a column of ordinary air ten feet long, 10 to 15 per cent, "I think it certain that the larger of these numbers fails to express" the true value) [read: On Radiation lines 1319-1323]. Ratios 15, 70, 72, 90 and "90 to 100" are explained by him as improving purity of the reference, which I accept as plausible and as a limit on their meaning. The Strachey table is monotonic except one step (12.6 then 12.1) and was chosen by Tyndall as "representative" [read: Heat deep read, section 3]. b. Consistency with experiment: the qualitative claim is now Established. Water vapour is the dominant absorber of terrestrial long-wave radiation in clear air, via the rotational and 6.3 micrometre bands and the continuum [memory: high]. The GISS attribution gives water vapour about 50 per cent of the present clear-and-cloudy greenhouse effect, clouds about 25, CO2 about 20 [web: Schmidt et al. 2010, https://www.giss.nasa.gov/research/briefs/2010_schmidt_05/], while the authors warn this is not the same as which gas drives climate change. The ten-feet figure is of the right order for humid air and a broad source (a Hottel-type emissivity estimate for about 0.14 atm-feet of vapour gives several per cent to ten per cent) [memory: medium; not checked on the web, flagged]. The frost, desert and mountain range arguments (Strachey, Hooker, Livingstone, Sturt, Martins) are consistent with the dryness-range relation, though in each case dry air also means clear sky, high altitude and surface conditions, and Tyndall gives no control for those. "CO2 only minor" was a fair reading of his data in 1861-68 but is superseded for the long-term question: water vapour condenses and follows temperature, CO2 does not, so CO2 sets the level on which vapour acts [memory: high; Schmidt et al. above]. Tyndall's data do not say CO2 is unimportant: Heat ch. XII (CO2 percentage tables, a CO flame source) was covered only by digest, so I cannot judge his CO2 numbers (48.0 to 74.3 per cent at 1 to 10 inches, from digest/OCR only). No split of effect between vapour and CO2 is given in the 1868 pages read [read: Heat deep read, section 5]. c. Tests and rejection. This is the claim with the longest contested record, and it matters how it was rejected. Magnus (Phil. Mag. 1861, 1862) and Buff found no vapour action and blamed films on the rock salt (Tyndall's own report, via Epsilon letters 1815, 1941; interested-party account). Tyndall's answer was to remove the plates, use an open tube, then no tube [read: above]. Lecher and Pernter (1880) agreed on dry gases but found "moist air a little more transparent than a vacuum" [web: Wikisource, PSM 1882, above; Tyndall 1882 as quoted, so a hostile paraphrase is possible]. Moist air cannot transmit more than a vacuum, so I read that result as an instrument artefact (vapour hesion on walls or windows, or source temperature at 100 C and a narrow band) rather than as a measurement of the gas [inference]. Very (US Weather Bureau, 1900) defended Tyndall, called Magnus's 14.75 per cent from dry air with glass plates "the fallacy of Magnus" (heated end plates acting as secondary sources) while asking for "further elucidation" of Tyndall's humid-air proportionality table [web: https://archive.org/details/atmosphericradia00veryrich]. Rontgen (1884), by an independent method meant to avoid hesion and rock-salt errors, concluded "moist air absorbs rays of lower refrangibility" and wrote that many physicists still regarded the question as open [web: https://archive.org/details/sim_annalen-der-physik_1884_23_9]. The record does not look like suppression; it looks like a hard experimental problem with three or four careful critics, resolved by method change and spectroscopy. The disputes were conducted in the open and Tyndall's reply was printed (although the Geneva Archives printed Magnus's memoirs and not his reply, per Soret, Epsilon letter 2040). Support Tyndall cites from Airy and Herschel is in his own letters (interested-party reporting); Airy's own letter is on Epsilon. The "70 to 90 times" figures were never independently reproduced in those units that I could find. d. If true: nothing changes; the modern greenhouse picture is built on it.

Verdict: split. Water vapour absorbs radiant heat and is the main absorber in the lower atmosphere: Established (it was Tested, contested for about twenty years, then settled by independent methods). "At least 10 per cent within ten feet": Plausible, untested as stated (consistent in order of magnitude, extrapolated from a 100 C source; Tyndall's lab figure is 7.9 to over 10 per cent). The numerical ratios 70-90 and 16,000 per molecule: Not testable as stated in his units. "CO2 only minor": Historically superseded but instructive (a fair reading in the 1860s; wrong as a guide to climate control).

Claim 3: the invisible radiation of luminous sources exceeds the visible; iodine in carbon bisulphide passes the heat, blocks the light; dark rays ignite things

a. Mathematics: Table I's second column is the first times 1.37, with small mismatches (15.5 gives 21.2, printed 21.0; 44.0 gives 60.3, printed 60.0) [read: 1866 deep read]: rounding or calibration, not an error of substance. 7.7 against 8 (area against filter subtraction) is two methods agreeing within 4 per cent. 8 / 9 = "one-ninth" and "one-tenth of the total" are consistent [calc]. Consistent. b. Experiment: the black-body check above shows 7.7 corresponds to an effective 2,900 K. Iodine in CS2 passing infrared and blocking visible is a standard filter [memory: high]. The ignition of paper, zinc, magnesium and charcoal in oxygen at the dark focus is a demonstration still performed. Tyndall himself notes combustion alone is not decisive for his theoretical point and used non-combustible platinum foil and carbon in vacuo [read: 1866 p.17]. His controls are good: a burned-through diaphragm, an alum cell (passes light, blocks dark heat) to detect any stray visible light, and an "undefended" eye test (the 1.2 inch cell failed; the 2 inch cells passed) [read: 1866 deep read]. Risky: he put his own eye at a focus capable of heating platinum to red (footnote warns against copying). c. Tested by: Tyndall, with demonstrations at the RI from early 1862 [read: Rede]; Melloni earlier (90 per cent of an oil-flame's emission, 98 per cent of incandescent platinum's, obscure [read: 1866 p.3]); Akin and Griffith, independently, in sunshine with a 2-foot mirror (1863) [web: Phil. Mag. 29, via Investigator]. Replication: yes in outline; the specific 7.7 and 8 figures were not independently reproduced in the sources I have. d. If true: nothing changes. It is the Planck/Wien picture before Planck.

Verdict: Established (arc and incandescent sources emit most of their power in the infrared; filter works). The solar "about 2": Never properly tested here (cited from Muller, not checked, and apparently high; see Corrections 2).

Claim 4: calorescence is a transmutation of heat rays into higher-refrangibility rays, the reverse of Stokes's rule; blue glasses make the platinum glow pink

a. Mathematics: none to check; the argument is verbal. The logical issue is the analogy. Stokes's rule concerns a cold body re-emitting a longer wavelength after absorbing a shorter one. Tyndall's case is a solid heated by absorbing long waves and then glowing at all colours, which Stokes called "not a little different from fluorescence" (letter, above). Tyndall's own private statement (to Croll, 14 Jan 1865): you hit the mass with "waves of slow recurrence, and ... obtain, in return from the mass, waves of quick recurrence" [web: https://epsilon.ac.uk/view/tyndall/letters/Tyndall2216]. Read literally, that is absorption, heating, and thermal re-emission, which is the modern account and does not need a ray-by-ray transformation. The public name (negative fluorescence, then calorescence) and the phrase "transmutation of heat-rays into light-rays" suggest more. b. Experiment: the observations are real and reproducible. Platinized platinum foil and carbon in vacuo glow white at the iodine-filtered focus, and the glow shows a complete prism spectrum [read: 1866 p.17]. Modern reading: hot solid, Planck spectrum, no violation of energy conservation, since the glow is fed by absorbed power. The "pink glow through certain blue glasses" is the only unresolved sub-claim [read: 1866 deep read, section 15; Rede footnote; Wikisource]. The deep reader notes the glow is faint and dim glow is poor for colour judgement. I add that his own coloured-glass table lists some blue glasses as passing a blue/green band plus "extreme red", so a thermal glow filtered by such a glass would plausibly look pink or purplish with no gap in the calorescent spectrum at all [inference from the table; no spectrometer data]. Tyndall himself says only "if subsequent experiments should confirm". No later check was found (Investigator). c. Tests: Tyndall's; Akin and Griffith's independent sunshine experiment; no controlled test of the pink glow. Priority is disputed. Akin's 'Account of Preliminary Experiments on Calcescence' (BA Report 1863) and Phil. Mag. 29 (1865) 28-43 describe the same principle (platinum made incandescent by "Herschellic" rays, the converse of fluorescence) [web via Investigator]. Tyndall's published history of the matter is Phil. Mag. 29 (1865) 218-31; the dispute ran Nov 1864 to early 1865 in nine letters; Sabine and Armstrong (BA) sided with Tyndall, Armstrong writing that the BA "could not ... confer upon Dr. Akin any exclusive privileges" [web: Epsilon letters 2075-2208, via Investigator]. Tait (1868) credits Akin with "Calcescence or Calorescence". Tyndall's letter of 3 Nov 1864 mentions an experiment held back "nearly nineteen months" (Epsilon 2167): the document that might show who published what first. Absence-of-evidence note: Akin's physics record stops after 1865; he later held Hungarian academic and political posts and died in poverty [web: Nevpont, via Investigator]. Competing explanations for the silence: loss of the dispute with the Royal Society establishment; move to other interests; no laboratory post. Records that would decide: Akin's letters in the Tyndall Correspondence vols 8-9, the Hungarian Academy archive, and the 1863 BA Report text. I do not find evidence of suppression in what I have; I also cannot rule out a lopsided contest between an outsider and the RI professor, and every account in the folder of the dispute is Tyndall's. d. If true as stated (one-step conversion by individual rays): nothing mainstream could allow it, since photon up-conversion by a single linear absorption would violate energy conservation. As corrected (thermal re-emission), nothing changes.

Verdict: split. Dark infrared rays make platinum and carbon glow, with a full spectrum: Established. "Transmutation of rays" as a distinct process, reverse of Stokes: Historically superseded but instructive (Stokes objected at once; modern account is heating and thermal emission, and Tyndall's private wording already says so). Pink glow through blue glass: Never properly tested (probable filter effect, untested). Priority between Tyndall and Akin: not settled by the folder; Tyndall's account is that of an interested party.

Claim 5: absorption is a property of the individual molecule, selected by resonance with the molecule's periods and by structure; conservation of molecular action and thermal continuity of liquids and vapours

a. Mathematics: arithmetic checks. 38 / 10.8 = 3.52, matching "3 1/2", and a 1 mm ether layer against 7.2 inHg over 38 inches is 9.6e-4 against 9.7e-4 mol/cm2 (deep reader's check); for hydride of amyl 8.8e-4 against 8.9e-4 [calc by deep reader; the density and molar mass values are his [memory]]. Equivalent matter is therefore correctly set up. Differences between pairs: 0.1 to 0.5 percentage points [calc]. b. Experiment: resonance of radiation with molecular vibration periods, and absorption proportional to the number of molecules in the beam at low pressure (Beer-Lambert), are Established [memory: high]. Tyndall's "cavities and indentations" and "ether condensed and entangled" explanations of why shape matters are superseded: the modern criteria are vibrational modes and the change of dipole moment [memory: high]. The strong form of "conservation of molecular action", that absorption is unchanged under any compression and between vapour and liquid, does not hold in detail. Pressure broadening changes line shapes, and liquid water's bands differ from vapour's (hydrogen bonding shifts and broadens the O-H stretching band) [memory: high]. His tests used broadband sources and a thermopile that integrates over wavelength, which would hide band shifts, and ether and hydride of amyl are not strongly hydrogen-bonded; at 66.7 against 67.2 per cent the weak form is supported. His own caveat applies: single tabulated examples, no repeated-measurement error, lime-light liquid value unreadable ("883") [read: NF deep read]. The inference from liquid water's opacity to water vapour's, offered in the Rede lecture "even if no direct experiment on the vapour of water had ever been made", is an inference from continuity and the vapour was in fact measured later (claim 2). The Investigator's Bakerian 1881 evidence is relevant here: sounds from the photophone, and the argument that results "of such precision and consistency" are "only compatible with the conclusion that the veritable radiators and absorbers are the molecules of the vapours" (Tyndall 1882, interested party). Photoacoustic spectroscopy descends from that method [web via Investigator: Manohar 2016]. c. Tested by Tyndall; Rontgen's independent pressure-rise method (1881, 1884) tests molecular absorption in vapours; no independent test of the 66.7 / 67.2 liquid-vapour pair in the folder. Forbes asked in 1859 whether absorption "follows the degree of condensation" (Epsilon 1627) and the long-tube/short-tube experiment is an answer to that. d. If true in the strong form: absorption spectra would be independent of phase and pressure, contradicted by spectroscopy. The weak form changes nothing.

Verdict: split. Molecular origin and resonance: Established. Conservation of molecular action and thermal continuity, weak form (integrated broadband absorption roughly independent of density and state for these test vapours): Plausible, untested independently (supported by his own data, 0.5 per cent). Strong form: Tested, not supported by later spectroscopy [memory: high; no source fetched]. Geometric and "ether entanglement" explanations of molecular structure: Historically superseded but instructive.

Claim 6: heat and light are the same wave motion differing in period; heat is a mode of motion (Rumford, Davy, mechanical equivalent)

a. Mathematics: the cubic-foot calculation is internally consistent step by step (273 x 1.8 = 491.4 F, text 490; 144 x 15 = 2,160; 1.29 x 0.24 = 0.31 oz; 0.31 x 490 = 152 oz; 9.5 / 1.42 = 6.7; 2,160 / 2.8 = 771.4) [read and calc]. Sensitivity is large: the answer rests on a difference of two nearly equal heats (9.5 and 6.7), so a 1 per cent change in the ratio changes 2.8 by about 10 per cent. With unrounded intermediates it is about 769; with the modern ratio 1.40 about 797 [read: claims.md; deep read]. So 771.4 agreeing with Joule's 772 is partly luck of rounding. Modern value 4.186 J/cal is 778 ft-lb per Btu-per-lb-F [calc]; Joule's 772 is 0.8 per cent low. b. Experiment: identity of light and radiant heat (reflection, polarization, double refraction, and the visible-to-infrared continuum of the electromagnetic spectrum) is Established [memory: high]. Heat as molecular motion and the equivalence of heat and work, with Joule, Mayer, Rumford: Established. Corrections to detail: Rumford's "107" is the water temperature after one hour, not a final result; the 23.5 against 28 minute figure measures a coating on goldbeater's-skin vessels, not cannon boring [read: NF deep read, E]. Tyndall's account of Mayer is Corrected above. c. Tests: Joule's paddle-wheel experiments, Rumford's cannon boring (1798), Davy's ice rubbing (I have not assessed how sound Davy's experiment was), Hertz in 1888 for electromagnetic waves [memory: high]. All replicated. d. Nothing: this is mainstream. Honest note: the credit given to Mayer's method is Tyndall's.

Verdict: Established. Mayer's number as given in Heat: a reconstruction, not Mayer's result.

Claim 7: glacier ice moves without viscosity, by fracture and regelation; the Ice Age needed abundant vapour and strong sun (Heat ch. VI)

a. Mathematics: the energy bookkeeping is consistent: 1,000 units latent heat x iron specific heat 0.1 gives 10,000 F, over a 2,000 F melting rise gives 5 lb of iron per lb of vapour (arithmetic [read: Heat deep read, section 7]; his 1,000 is a round figure, about 970 Btu per lb at 212 F [memory: medium]). It is a rhetorical comparison, not a model of glacial climate. The 3 feet a year over the Christian era equals 5,580 feet only at 1,860 years, not shown in the text. b. Experiment: regelation (Faraday, 1850) is real and Tyndall's press demonstrations (a cup of ice, compacted scrapings) are reproducible [read]. As the explanation of glacier flow it is largely superseded: ice deforms by creep under sustained stress (Glen's law, a power law with an exponent near 3, 1955) and Forbes's "viscous" view is accepted as the closer one; regelation survives as part of basal sliding [web: summary at https://tc.copernicus.org/articles/15/4589/2021/ and the Kamb and LaChapelle sliding study at https://www.cambridge.org/core/services/aop-cambridge-core/content/view/A032752BFEA9BB2628AC95966D72C332/S0022143000028756a.pdf; the Forbes-Tyndall outcome is from a search summary, not a primary reading]. Tyndall himself conceded that ice acts as a viscous body under pressure and "accounts, certainly, for half the facts" (Heat para 235) [read: Heat deep read, section 5]. His premise "the most delicate experiments ... have failed to detect this stretching power" cites no experiment, date or person [read]. The Ice Age cause (vapour supply plus a better "condenser") is not a model; the orbital explanation developed after Croll [memory: medium, not checked on the web]. c. Tests: Forbes's field measurements (2 feet a day at the Mer-de-Glace centre, per Tyndall) and Tyndall's; later laboratory creep experiments (Glen, 1955). The historical record, as told by the Science History Institute (2025), has Tyndall attacking Forbes in terms such as "utterly annihilated", Forbes holding the two views were "complementary" and complaining in 1864 of being "disheartened and hopeless" [web: https://www.sciencehistory.org/stories/magazine/under-pressure/]. Forbes's claims were not tested and failed; they were rejected in a prominent rhetorical contest, and later vindicated in substance. This is the one claim in the folder where the interested party (Tyndall) won the argument in his lifetime and was wrong on the central point. d. If Tyndall's strong form were true: crystal-plastic deformation would not exist. It does.

Verdict: split. Regelation as a phenomenon: Established. "Glacier ice is not viscous; flow is fracture plus regelation": Historically superseded but instructive (Tested and not supported by later creep experiments; Tyndall's concession on pressure was nearer to the truth than his headline). "More vapour and strong sun, not a weaker sun" for ice ages: Plausible, untested as a mechanism and superseded by later orbital accounts.

The strongest honest case for Tyndall

The core experiment was hard, adversarial and survived. A careful critic (Magnus) challenged it, Tyndall answered each objection with an experimental change, and the result outlasted Magnus, Buff and Lecher-Pernter; independent methods by Rontgen and a century of spectroscopy agree with the qualitative result. The "anomalous" result of 1859 (that a few hundredths of a per cent of a gas could stop a large part of a beam, per Forbes "surprised at the magnitude of the effects") was true.
The ranking of gases, the reciprocity of emission and absorption, the breath-CO2 measurement (4.66 chemical against 4.58 physical per cent, and 5.33 against 5.22, agreement within 2 per cent [calc]), and the invisible-light filter are all repeatable.
On the pattern of the record: no evidence of suppression; the dispute was open and long. The weak spots are Tyndall's own unsourced figures (ten-foot column, 35 per cent ether shell, Sahara frost, "most delicate experiments", "distinguished authority"), which I report as his statements, not as results.

Where the contrarian standard is applied to the inventor

"Air = 1" baseline and no error bars (claim 1, 2).
Rede lecture "tenth of an inch of water passes sun's rays, half that stops every terrestrial ray" is stronger than Heat's own data (0.27 inch opaque, 97 per cent cut at about 0.04 inch) (claim 2).
The 35 per cent two-inch shell of ether vapour is "proved by experiment" with no pressure, source or tube given, and Heat's analogous shell figure uses olefiant gas (33 per cent at 2 inches, 26 at 1 inch) [read: On Radiation deep read, section 7]. Reported, not discarded.
Terrestrial extrapolation from a 100 C source (claim 2); pink glow stated as "not subjective" with no stated control (claim 4); "most delicate experiments" unattributed (claim 7); his own priority narratives (Akin, Foote, Franz, Mayer).
Unfalsifiable language: "ether entangled and condensed" explanations of molecular shape (claim 5).

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

1859 Note (Proc. RS 10, 37-39): first crude test negative, then "Very large effects" with the compensated method; no numbers, no climate [web: https://archive.org/details/jstor-111604].
Forbes to Tyndall, 15 July 1859 (surprise at magnitude; asks about condensed gases) [web: Epsilon 1627]; Tyndall to Herschel and Hooker on Magnus ("he has not a foot to stand upon", Epsilon 1941, interested party); Airy 1863 ("beginning Meteorology"); Soret on Geneva (2040).
Lecher and Pernter dissent on vapours; Buff likewise; Ferrel (as quoted by Very) accepted the film explanation; Very's rebuttal and his request for further elucidation of the humid-air table; Franz 3.54 per cent for dry air; Rontgen 1884; Rontgen "an independent and successful worker".
Foote 1856 and Franz 1855 on gas absorption before Tyndall (priority, Jackson 2019).
Akin priority dispute, Stokes's and Croll's letters, Tait 1868; Akin's silence analysed above.
Macdonald viXra 2015 (unrefereed, author's claim); Babbitt 1878 taking up calorescence in an alternative colour theory (influence); Protons for Breakfast commentary (no replication); RI surviving tube 698 x 85 mm dated 1861; Science History Institute on Forbes.
Not reached by the Investigator and relevant: Contributions to Molecular Physics (1873) for the full 1861-72 memoirs, jstor-108936 for the missing p.1 and plate of On Calorescence, Magnus's papers, Lecher and Pernter's Phil. Mag. paper, Strachey July 1866, Forbes papers at St Andrews, the Royal Institution catalogue.

Verdict summary

Gas absorption pattern and reciprocity: Established (numerical table: source-dependent ranking only).
Water vapour as the main absorber: Established (qualitative, after a contested, open dispute, tested and replicated by independent methods); ten-foot 10 per cent: Plausible, untested as stated; CO2 minor: Historically superseded but instructive; 70-90 and 16,000 figures: Not testable as stated.
Invisible radiation exceeds visible: Established (7.7 matches a source near 2,900 K [calc]); solar "about 2": Never properly tested here.
Calorescence: glow observed, Established; transmutation / reverse-Stokes: Historically superseded but instructive; pink glow: Never properly tested; priority against Akin: unresolved.
Molecular absorption by resonance: Established; weak conservation of molecular action: Plausible, untested independently; strong form: Tested, not supported (later spectroscopy, [memory]).
Heat and light one wave motion, heat as motion: Established; Mayer's 771.4 is Tyndall's reconstruction, not Mayer's result.
Glaciers: regelation Established as phenomenon; "not viscous" Historically superseded but instructive; Ice Age mechanism Plausible, untested.

Web checks done: 4 (Mayer; Strachey 1866, which failed; Glen's law and Forbes-Tyndall; Schmidt et al. 2010). Not checked: the ten-foot emissivity estimate and the phase dependence of water's spectrum [memory], the orbital theory of ice ages [memory], Muller's solar ratio.