Tousey-Sinclair: Arbiter (Step 3)

Arbiter: weighs physicist.md, mathematician.md and contrarian.md (all dated 2026-10-09) for Sinclair Tousey, Medical Electricity, Rontgen Rays and Radium (W. B. Saunders, 3rd ed., April 1921), with the Investigator's 2026-10-08 finds (sites.md, research/). Date: 2026-10-09.

Tags: [read] project documents (I re-read the source text or a page image where reviewers differed); [web] with URL; [memory] with confidence; [calc] my own arithmetic (project venv).

Source gap note

No source gap for the key work: all 1361 pages of the book have OCR text (library-copy scan). Limits that later documents must repeat: - The folder scan lacks printed pp.282-283, 286-287, 290-295. The Investigator located a second scan (IA medicalelectrici00tousiala, UCLA copy) holding pp.282-283 (electrotonus) and p.287 (concentration-cell membrane passage). The three reviewers read those passages from noisy OCR [read: research/pages/ia-1921-ucla.txt]. No deep reader has read them from page images, and pp.290-295 are still unread. Claim 7's membrane account is therefore read, but only at OCR quality. - Digest-only chapters: static and dynamic electricity, muscle physiology, high-frequency currents and diathermy (PDF 544-647, except goitre, impotence, neuritis, and the p.585 diabetes passage, which the physicist read), tube design, dosimetry, x-ray effects, radium. Claim 3 rests partly on digests. Every verdict on claim 3 below is low confidence for that reason. [Updated by the QE revision: PDF 544-647, 557-681, 297-349, 749-773, 1130-1180 and 1252-1297 have since been deep-read; the evidence description for claim 3 is now medium confidence, efficacy verdicts remain low.] - Everything about Tousey's own cases comes from his own text. Much of the book is a compilation; sentences credited to "the author" are Tousey's. - No conclusion in this file rests on [memory] alone. Where a figure is [memory] it is marked.

Where the reviewers differ, and what I recomputed

D1. Dose of the "Tousey dose"

Inputs [read: PDF p.1181-1182]: 6-inch spark (the book equates this to about 70 kV peak), 3 mA, 10 min, 10 in from the focus, 3 mm Al; table value "(5) H" printed in parentheses; row text "Author's dose for every kind of therapy. Never causes dermatitis. Not to be repeated inside of four to six weeks."

• Charge Q = I t = 3 mA x 600 s = 1800 mAs. Distance 10 in = 0.254 m, so inverse square factor (1/0.254)^2 = 15.5 [calc].
• Output of a modern 70 kV tube at 1 m with 2-3 mm Al: 0.023 to 0.030 mGy per mAs [web: https://www.rug.nl/education/courses/other-education/radiation-protection/edu/ts-mt/ts-mt-data-examen/mcts-mtfiguura?lang=en, via a search summary; the table's column alignment is ambiguous, so treat as +/- factor 2]. This resolves the reviewers' disagreement about the base figure: the physicist's 0.03-0.05 is above what I could source; the mathematician's 0.03 is at the top of the sourced range.
• Free-in-air dose: 0.023-0.030 x 1800 x 15.5 = 0.64-0.84 Gy. Skin backscatter about x1.3 [memory, medium] gives 0.83-1.09 Gy.
• The mathematician then multiplied by a further 0.5 for the 3 mm Al, but the modern figure already includes 2-3 mm Al, so that is partly double counting. The physicist added no correction for the period apparatus. The reasonable correction: a coil with a gas tube has a peak kV above its effective kV. If the effective voltage is 55 kV instead of 70, output (proportional to V squared, Tousey's own rule [read: p.741]) falls by (55/70)^2 = 0.62. That gives about 0.5-0.7 Gy.
• Arbiter's range: about 0.3-1.1 Gy at the skin per session, central value near 0.7 Gy, uncertainty a factor of 2-3 either way (gas-tube vacuum drift, which Tousey himself warns makes the milliammeter "dangerously misleading" [read: p.1181]).
• Yardsticks: the IAEA says a skin reaction is unlikely below about 2 Gy and gives 2-5 Gy for transient erythema [web: https://iaea.org/resources/rpop/health-professionals/radiology/erythema]; the early-transient value of 2 Gy and the "main erythema" near 5-6 Gy are confirmed by search summaries only. So the estimate is below the modern erythema threshold, by a factor of about 2-7.
• The old dose units cannot be converted: Holzknecht's "1 H" was an arbitrary colour change, a light erythema was "3 H", and Sabouraud-Noiré "Tint B" meant a full erythema dose; observers disagreed on what a full erythema dose was (a 1926 Radiology paper: one observer's full dose might be 80 percent or less of another's) [web: search summary of RSNA papers, e.g. https://pubs.rsna.org/doi/full/10.1148/6.3.236; I found no source for a roentgen value of an H, so I give none].
• [Superseded in part by QE revision R1: the book does hold period calibrations, and its own conversions of these settings span about 0.16-1.5 erythema doses.]
• Conclusion. Both reviewers are right in kind: the physical dose was probably below erythema (mathematician), but not by the large margin he claimed, and the physicist's 1-2 Gy is too high on sourced output. Tousey's own table prints the dose as "(5) H" and defines 5 H as "an erythema dose" (and p.713: 5 H mild, 7 H severe dermatitis [read]). The parentheses suggest a computed rather than measured value. The label and the "never causes dermatitis" statement conflict in his own units; my independent estimate says the real dose was most likely sub-erythema, which is consistent with "rarely a visible reaction in one sitting" and says nothing about "never".

D2. The "output greater than input" passage [read: page images pp.740-741]

Tousey's wording: 100,000 V x 2 mA "would be 500 watts" and x 15 mA "1500 watts"; primary about 500 W; "nearly half of the power in the secondary is wasted in the form of inverse discharge which must be suppressed"; "the meters indicate the maximum of the successive currents rather than their average strength". - Arithmetic [calc]: 100 kV x 2 mA = 200 W (printed 500); 100 kV x 15 mA = 1500 W; 56 kV x 10 mA = 560 W (printed 746). All reviewers agree. - The three explanations are not rivals; they stack. (i) A moving-coil milliammeter reads the average current [memory, high], so the mathematician is right that the current factor is not a peak. (ii) The voltage is read from a spark gap, which gives the breakdown (peak) value, and Tousey measured the voltage falling "to less than one-half during the passage of the spark" [read: p.740]. (iii) Tousey himself says about half the secondary power goes to inverse discharge. - Check [calc]: with the voltage at one-half of peak throughout, 50 kV x 15 mA = 750 W, still above a 500 W primary. The explanation needs the mean voltage during conduction at or below 33 kV (500 W / 15 mA), i.e. a third of the 100 kV figure, or less than 15 mA actually flowing. So Tousey's p.740 "less than one-half" alone does not close the books; the 100 kV figure is also inconsistent with the book's own 6-inch spark = 70 kV (which would give 1050 W at 15 mA). The conclusion (no over-unity) is not in doubt; the quantitative reconciliation is not in the book. - Oudin resonator, 10 kV x 0.25 A = 2500 W: the physicist's reactive-power point is correct and is the main factor there. For a sinusoidal V_peak of 10 kV, V_rms = 7071 V, apparent power 7071 x 0.25 = 1768 VA; at a power factor 0.1 [memory, medium] the real power is 177 W [calc]. The power factor is an assumption, not data. Tousey explains it as peak versus average, which is incomplete for this device.

D3. "Falls off slightly faster than inverse square"

Air attenuation [calc, mass attenuation coefficients for air recalled, memory medium]: over 25.4 cm (10 in) a 20 keV photon loses 2.4 percent, 30 keV 1.1 percent, 50 keV 0.5 percent; over 50 cm 4.6, 2.1 and 1.0 percent. So the beam does fall slightly faster than 1/d^2 for the soft part, by a few percent at most. Tousey's wording, "rather more rapidly than the square", is physically right and trivial in size. His worked example (a plate corner at "almost twice" the centre distance, receiving a quarter) is geometrically wrong: 9 in plate distance, 8-9 in off-centre gives 12.0-12.7 in, ratios 1.34-1.41 [calc: sqrt(81+64), sqrt(81+81)]. Inverse square gives 0.56-0.50 and with the obliquity factor (cos^3) 0.42-0.35; the anode heel effect lowers it further [memory, high]. The contrarian's "internally inconsistent" is too strong; the example is wrong but the statement is not contradicted by anything else in the book.

D4. Ultraviolet from the glass vacuum electrode

[Superseded in part by QE revision R3. The quotation below is out of context: the lines are PDF p.578, and the sentence says it is "hardly to be supposed" that the ultraviolet takes a direct part. The book says glass and mica stop the UV and puts its source in outside sparks. The glass-absorption data below remain correct.]

The book says the glass vacuum electrode produces "violet and ultraviolet radiations", ozone, and that "ultraviolet rays absorbed by the body" matter [read: PDF p.585 region, text lines 32139-32166, 34087, 35107]. Soda-lime glass is fully UV-absorbing below about 300 nm, with a 50 percent cutoff near 337 nm for a 3 mm sheet [web: patent text via search, https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/6878653 and related; approximate, thickness dependent]. The strongest UV line of an air or nitrogen discharge is the 337.1 nm band [web: same search, summary]. So: short-wave germicidal UV (254 nm) cannot leave a soda-lime electrode; near-UV at 337-390 nm leaves, attenuated near 337. The book mentions quartz windows and lenses for mercury-vapour lamps (text lines 37175-37429) but I found no statement that the vacuum electrode is quartz. The physicist is right about the strong form; the contrarian is right that the discharge makes UV and ozone. Whether near-UV through the glass does any good was not tested. "506,000 Gauss" [read: text line 32548]: the book says a meter inside the cage shows the field, "Moutier's field averages 400,000 Gauss". The physicist's induction estimate (B of 1 mT, 500 kHz, r 0.15 m gives E about 236 V/m [calc: confirmed]) shows 50 T is impossible. I could not check Doumer's original, so "unit slip in the sources, unresolved" is the correct status.

D5. "Almost all physiologic effects of electricity are electrolysis and electric osmosis" [read: p.297]

• Mathematician's Faraday calculation [calc, confirmed]: a threshold charge of 0.2 microcoulomb corresponds to 0.2e-6 / 96,485 = 2.1e-12 equivalents, about 5e-11 g of sodium. That is too little chemical change to be the cause of a twitch if "electrolysis" means chemical decomposition at an electrode.
• But ions moving across a membrane is also an ionic process. Charging a nanofarad-scale membrane capacitance by 0.1 V moves about 2e-10 C of charge [calc, capacitance assumed]. Tousey's own concentration-cell (Bernstein) passage describes this kind of mechanism [read: ia-1921-ucla.txt, OCR]. So the claim is wrong in its strict chemical reading and partly right in a loose ionic reading.
• Nernst check [calc]: K+ 5 mmol/L outside, 140 inside, 310 K gives -89 mV, within the usual -70 to -90 mV [memory, high].
• Verdict split accordingly (see claim 7). The mathematician's "Refuted by calculation" applies to the strict form only.

D6. Ion speeds on printed p.401

The page image shows, as printed: silver 0.166, lithium 0.094, sodium 0.926 "cm. per hour" for 1 V/cm [read: page image PDF p.405]. This is not an OCR slip. The numbers are not in the order of real mobilities and not mutually consistent: modern values are Ag+ 6.4e-8, Na+ 5.2e-8, Li+ 4.0e-8 m^2/(V s), which is 2.3, 1.9 and 1.4 cm/h at 1 V/cm [calc from the mobilities; Na+ value [web: https://en.wikipedia.org/wiki/Electrical_mobility], via the mathematician]. The printed sodium value is within a factor of two of modern; silver is 14 times low and lithium 15 times low; the ordering is wrong. The deep read's "probably about 0.1 cm/h" for sodium (OCR garbled) is wrong: the page reads 0.926. The conclusion Tousey draws (ions drift only millimetres to a couple of centimetres per hour per V/cm, so action is superficial) survives, and the modern numbers are the ones to use. The 39 m/s statement [calc, confirmed]: Na+ drift at 100 V/m = 5.19e-6 m/s (1.9 cm/h); 39 m/s would need 7.5e8 V/m. Charge relaxation time epsilon/sigma = 80 x 8.854e-12 / 0.5 = 1.4e-9 s [calc; sigma [memory, medium]].

Claim-by-claim

Claim 1: the "Tousey dose" never causes dermatitis; x-ray cures or controls surface cancers and skin disease; only temporary benefit in leukaemia, little cure in pulmonary tuberculosis

Agreement: all three accept that superficial x-ray therapy of basal and squamous skin cancers is mainstream and effective; the leukaemia and tuberculosis statements are consistent with later experience and are Tousey's own; his cure counts have no denominators; combined treatments blur the effect of x-ray alone; the lupus case lacked microscopy and may duplicate the nose epithelioma (both a woman of 67 whose father lost a nasal ala) [read: deep read]; "never causes dermatitis" sits against his own "(5) H = erythema dose" label. Disagreement: how to label "never". Physicist: Internally inconsistent. Mathematician: Never properly tested, tension minor. Contrarian: Never properly tested. Dose estimate: see D1. Stronger argument: the contrarian and mathematician. "Never" is an empirical claim, and the contradiction is in a label (H) whose meaning the book itself says varies by a factor of four with the pastille position and which the table prints in parentheses. The contradiction shows the table mixes writers' units; it does not show the author's clinical statement false. Evidence against "never" sits in the book itself: the hyperidrosis patient (about 10 H, 28 min) with blister, keratosis and fissures who "would have preferred the disease to the after-effect" [read: p.1190-1191]; deliberate x-ray burn in hypertrichosis [read: p.1182 region]; his remark that severe skin change "produces in some cases a tendency toward malignancy" [read: p.1191]. And evidence from outside: his own hand keratoses cured by radium (Arch. Radiol. Electrother. 20(10):344, 1916 [web: https://academic.oup.com/bjrbir/article-pdf/20/10/344/57486712/are.1916.0084.pdf; PDF not opened, snippet only]) and his 1922 statement "3600 units of soft x-rays will produce severe dermatitis" [web: Google Books snippet via the Investigator]. Those two are about operator exposure and a different dose, so they qualify the claim rather than refute it. What would settle it: a dose reconstruction from a published spectrum table for a coil-driven gas tube (no 1921 spectrum on record in the project), or a case series with follow-up of his patients (Garden City, St. Bartholomew's Clinic records, not obtained).

Final verdicts [QE revision R1 revises the dose-label and "no reaction in one sitting" lines; the list below is the first pass]: - Superficial skin cancer and lupus respond to filtered low-kV x-ray: Established. - "The author's dose never causes dermatitis": Never properly tested (not dismissed untested: it was a clinician's report, no systematic series). Separate note: his dose label is Internally inconsistent in his own units (minor; a units issue); my estimate of the delivered dose (about 0.3-1.1 Gy) is below modern erythema thresholds, so "usually no visible reaction in a single sitting" is Plausible, untested. - His specific cure counts (non-adherent epithelioma "almost certain to be cured", lupus at three years, breast recurrence at five years): Never properly tested (case reports, no denominators or controls). - Leukaemia temporary benefit only: Established (his own statement). - Pulmonary tuberculosis little permanent cure: Tested, not supported (as treatment; he reports the negative literature himself).

Claim 2: x-ray dermatitis is partly a "condenser" effect avoided by contact; the Sinclair Tousey tube beats Geyser's Cornell tube; static charge may help ulcers heal

Agreement: the dose-distance arithmetic is right (contact puts the pastille at the skin, so "Tint B" is four times the usual dose [read: p.1232]); Kienböck 1900 attributed the skin effect to dose, not electric phenomena [web: https://pubmed.ncbi.nlm.nih.gov/25012643/, via the Investigator]; the tube comparison has no neutral test; both Tousey and Geyser are interested parties (Geyser's later Tricho hair-removal system was criticised by the AMA in 1925 [web: https://en.wikipedia.org/wiki/Albert_C._Geyser]; that bears on how much weight his 1908 record deserves, and does not refute it). Tousey's own position is hedged: "has not been proved" but "probably contains a certain element of truth" [read: p.1231]; and burns occurred where distance alone was relied on [read: p.1218]. Disagreement: the mathematician adds "Refuted by calculation" using i = C dV/dt = 10 pF x 7e8 V/s = 7 mA and P = i^2 R about 0.05 W [calc, confirmed; the 10 pF and 0.1 ms are his assumptions]. That bounds only one pathway (Joule heating by displacement current). It does not touch corona, ozone and nitrogen oxides at the surface, dust attraction, or a sensation of discharge. The experimental test (Kienböck) is the stronger basis. The calculation is a supporting bound. Not found: the contrarian did not find a passage where Tousey cites Kienböck; the book's x-ray burns chapter was digest-only. "Not found", not "absent". [Superseded by QE revision R2: p.1158 cites Strater, Kienböck, Oudin and Scholtz against electric causes.] Priority: the book dates his contact tube's demonstration to 17 Oct 1905 [read: deep read, Fig. 832]; the Cornell tube is dated 1905 on Wikipedia [web, via Investigator]; his patent filing (US1100037, 1911-02-23) postdates both and decides nothing. Both dates are inventors' own accounts. This corrects sites.md, which called the accounts "conflicting". What would settle it: a matched-dose clinical comparison (not now possible), and for priority Dental Cosmos 1906, NY Odontological Society minutes of 17 Oct 1905 and Geyser's JAMA 1908 paper (none read).

Final verdicts [QE revision R2 replaces this list]: - Condenser (electrostatic) cause of x-ray dermatitis: Tested, not supported (Kienböck 1900; also his own distance burns). - Heating by tube capacitance as a burn mechanism: Refuted by calculation (this pathway only). - Static charge speeds ulcer healing: Never properly tested. - Tousey tube superior to the Cornell tube, and priority: Never properly tested. The lead-glass bulb that cuts off-focus radiation (his US1100037) is sound physics [memory, high].

Claim 3: high-frequency currents (d'Arsonval, Oudin, Tesla), usually with x-ray or ultraviolet, give lasting benefit by raising metabolism, lowering blood pressure, via ozone and ultraviolet from the vacuum electrode

Reading limit: digest-level except where stated. Confidence on every verdict here: low-medium. Agreement: no controlled trial; nearly all of Tousey's cures combine x-ray and high-frequency treatment; the book records disagreement (Moutier's fall against German authors' rise; Delherm and Laquerrière's smaller, inconstant effects; Cohn, Boedeker and Vinaj no change in urine sugar; Denoyes could not separate diet) [read: digests; PDF p.585 text line 32625]; the 1911 reviewer judged many cures "not borne out by facts, but which the author quotes in good faith" [web: https://archive.org/details/paper-doi-10_1097_00000441-191107000-00023, via Investigator]; nerves are not excited at these frequencies (the book says so, via Boruttau [read: p.292]) while heating continues. Diabetic sugar "from 2700 grains (180 grams) a day to a trace" is his own case [read]; 2700 grains is 175 g [calc]; insulin arrived 1921-22 [memory, high]. Disagreement: only on the UV mechanism (D4) and on whether the "506,000 Gauss" is a unit error (D4). Stronger argument: the physicist's, because it is backed by glass-absorption data and the contrarian does not dispute the glass. What would settle it: a blinded sham-controlled trial in a defined condition (the contrarian's suggestion; the physicist's heating estimate of about 14 W/kg [calc, confirmed, assumptions: B 1 mT, 500 kHz, 0.15 m, sigma 0.5 S/m] tells the trial designer that the cage current is strong whole-body heating, a confound); a deep read of PDF 544-647.

Final verdicts [QE revision R3 and R4 revise the blood-pressure, diabetes and ultraviolet lines]: - Radio-frequency deep heating, and no nerve stimulation at these frequencies: Established. - Lasting lowering of blood pressure, raised metabolism, diabetic sugar fall, goitre, impotence, ataxia, sciatica cures: Never properly tested. - Germicidal UV from a soda-lime vacuum electrode: Refuted by calculation (glass absorbs below about 300 nm). Near-UV and ozone from the discharge: real; any benefit through the skin: Never properly tested. - "506,000 Gauss" as a magnetic field: Internally inconsistent as a unit (unresolved; source not checked).

Claim 4: dental infection from blind abscesses causes systemic disease including endocarditis, ulcer and cancer; the permanent arch can be predicted from the x-ray width of the unerupted upper central incisor (R = 2w + 0.24 in)

Agreement: the construction is mathematically right (ED = 4R/sqrt 3 = 2.309 R [calc]; with w = 0.36 in, R = 0.96 in, molar distance 1.44 in, ED = 2.22 in [calc, confirmed]); table slips are in the book (ratios 1.30 and 1.25 should be 1.57 and 1.29; "7 cases" for 9 children); the radiographic width check (9 teeth, 7 children, within 0.01 in) tests image fidelity only, not the arch prediction, and the only follow-up is one untreated child; focal infection was widely accepted in 1910-1930 and later questioned (a 1916 orthodontia review called his reports "more convincing" [web, via Investigator]); no counts or controls stand behind "every case of cancer". Disagreement: arch formula labelled Plausible, untested (mathematician) or Never properly tested (physicist, contrarian); cancer statement Not testable as stated (mathematician) or Never properly tested (others). Stronger argument: the physicist and contrarian. The statement "has always found dental infection in every case of cancer" is testable (count dental foci in cancer patients against matched controls) but untested; it is not "not testable". The mathematician's base-rate point is correct and worth keeping: dental pathology is very common in adults [memory, medium], so "always" is expected by chance. For the arch formula, the mathematician's check against published tooth widths (central 8.7, lateral 7.0, canine 7.8 mm gives 1.85 in for six teeth; formula 1.94 in at w = 0.343 in, within 5 percent [web: https://search.bvsalud.org/gim/resource/en/emr-109906, one study via search summary]) shows the formula is anatomically sensible at an average w. It says nothing about outcome. It therefore stays "Never properly tested" in the predictive sense (no outcome data), and mixed-dentition analyses of Moyers type descend from the same idea [memory, medium; not web-checked]. The 0.24 in constant and the 1.5 factor are not derived in the book. Third-party counts for the cardiac claim exist (Hartzell, Henrici and Leonard: endocarditis in 23 of 31 cardiac cases [web, via Investigator: https://archive.org/details/roentgenographic00tousiala]); I did not read them. What would settle it: dental radiographs and matched controls for cancer patients; for the arch method, a longitudinal series comparing predicted with measured adult arches.

Final verdicts: - Dental radiography of root abscess and unerupted teeth: Established. - Oral infection causing endocarditis in susceptible hearts: Established. - Focal infection as a cause of ulcer, cancer, insanity, cured by extraction: Historically superseded but instructive (ulcer: H. pylori, Marshall and Warren [memory, high]). - "Dental infection in every case of cancer": Never properly tested. - Arch formula: geometry consistent; as a predictor of good orthodontic outcome Never properly tested.

Claim 5: much plate density comes from tissue secondary rays (about half, by the Bucky grid); fall-off slightly faster than inverse square; doubling the coil's primary current cuts exposure about tenfold; the "Tousey" unit

Agreement: scatter is large and real; a grid needs about double exposure (grid factor 2 is in the low range of modern values [memory, high]); his "fluorescence of the body" mechanism is labelled by him "the author's theory", and modern physics calls it mostly Compton scatter that fogs rather than forms the image; the Tousey unit (the effect on Kodak film of 1 candle-power carbon-filament light) is non-transferable and the book says it "is no longer required" [read: claims.md]; the 76:1 gaslight comparison needs the rounded "five times" distance (the stated 60 in / 11 in gives 3 x 5.45^2 = 89 [calc, confirmed]). Disagreement: the tenfold rule. Contrarian: Internally inconsistent with "twice the milliamperes would require only half the exposure" [read: p.821, p.824]. Physicist and mathematician: different variables. The rule is "doubling the primary amperage reduced the time tenfold": time proportional to I^-3.32 [calc: log 10 / log 2]. Output proportional to mA x kV squared (his own rule, p.741 [read]); if doubling primary current doubles both the tube mA and the voltage, output rises 2 x 4 = 8, close to ten. That is a coil assumption, not in the book, but it removes the contradiction because the two rules hold different things fixed (primary current in a coil, versus tube mA at fixed voltage). The contrarian himself says "plausible but undocumented". I side with physicist and mathematician. On scatter share: "about one-half" is right for limbs and low for the trunk (scatter-to-primary of 3-5 or more [memory, high]); his lead-cylinder and grid tests are qualitative. Further: the Tousey unit was published (Archives of the Roentgen Ray 17:427, 1913), abstracted in JAMA 60:1664 (1913), used in 1916 and by himself in 1922, and noted by historian R. F. Mould [web: Google Books snippets via Investigator, research/2026-10-08-google-books.md]. His radium assay through the sealed tube (NYT 7 Jan 1914) uses the penetrating gamma method [web: https://www.nytimes.com/1914/01/07/archives/suggests-testing-tube-of-radium-strength-can-be-determined-without.html via Investigator; the method is standard [memory, high]]. What would settle it: a measured exposure-versus-primary-current table for an interrupter coil; a scatter-to-primary measurement on a phantom.

Final verdicts: - Large share of plate density from scattered rays; grid and diaphragm help: Established. The figure "about one-half": Plausible, untested. - Tissue "fluorescence" forming the useful image: Refuted by calculation in the strong form (attenuation differences form the image; scatter is fog); as his labelled theory: Not testable as stated. - Falls off slightly faster than inverse square: Plausible, untested (air absorption gives 1-2 percent, D3). His worked plate example: Refuted by calculation (geometry wrong). - Doubling primary current cuts exposure about tenfold: Plausible, untested. - Tousey unit: Historically superseded but instructive.

Claim 6: coil output exceeding input is a meter illusion; the ideal generator is constant potential, which Cabot and Shaw built after reading Tousey

Agreement: not an over-unity claim; energy conservation holds; two products in the text are mis-multiplied (D2); x-ray output scales as mA x kV squared; constant potential is the engineering ideal; his hope that constant voltage gives "x-rays all of the same wave-length" [read: p.746] is wrong because bremsstrahlung stays continuous up to the Duane-Hunt limit [memory, high]; the priority claim is from Tousey himself (book, and NYMJ 97:648-649, 29 Mar 1913: Cabot and Shaw "found three or four pages which suggested the present discovery"; "Doctor Williams, of Boston" sent them to x-rays [web: https://archive.org/details/newyorkmedicaljo9719unse, via Investigator]); Cabot's 1912 converter patents (US1112436, US1191759; filed 1912-08-22) do not mention x-rays or Tousey [web: https://patents.google.com/patent/US1112436A/en, via Investigator]. Silence of the patents is not contradiction; patents rarely credit consultants. Disagreement: mechanism (D2), and the label for the single-wavelength hope. The contrarian says "Refuted by calculation"; the physicist and mathematician treat it as a hope. Tousey's words were "might very probably be able" [read, per physicist]. It is a hope, and the physics refutes the monochromatic version. Priority is documented in one dated contemporary publication, by the claimant, who has a stake in the credit; it is plausible in chronology (his 1910 edition predates Cabot's filing), names a third party, and has not been corroborated or contradicted by Cabot, Shaw or Williams. It is neither dismissed nor confirmed. What would settle it: Cabot and Shaw papers (Brookline, Massachusetts, and their company), Williams' correspondence, a copy of Tousey's 1910 edition with the pages in question (the Investigator located the 1910 edition on IA), and any Cabot-Shaw publication on x-ray supply.

Final verdicts [QE revision R5 and R6 add to this list]: - Output above input is a meter artefact, no over-unity: Established (mechanism is a combination, D2; the Oudin case is mostly reactive power). - Constant potential is the ideal x-ray supply: Established. - Monochromatic x-rays from constant potential: Refuted by calculation (offered only as a hope). - Cabot and Shaw built it from his pages: Never properly tested (one interested-party statement, dated 1913; not dismissed).

Claim 7: body acts as NaCl electrolyte; almost all effects of electricity are electrolysis and electric osmosis; medicinal ions act at the surface; Leduc's interrupted current can produce electric sleep (unestablished in man)

Agreement: ionic conduction in tissue, correct [memory, high]; surface action of ionic medication is supported by the experiments Tousey cites (Iscovesco and Matza; Tuffier and Maute; Gautrelet [read: p.415-416]) and by my mobility figures (D6); the 39 m/s statement is internally inconsistent with his p.281 text (180,000 mi/s) and confuses ion drift with the speed of the field (D6); the "288,000 mi/s" for metals exceeds light [read]; electron charge 1.34e-19 C against 9e23 per faraday do not agree [read; 96,500 / 1.34e-19 = 7.2e23 [calc]]; the electromagnet chapter says "no discoverable effect" for "unbiased observers", with no trials cited [read: printed p.453]; electric sleep rests on one human self-experiment (Leduc, 20-minute sessions, 90-110 interruptions a second, up to about 4 mA) plus unnamed animal work, and Tousey himself says it "has not been tried in a sufficient number of cases" [read: p.534-536]; Corrado's dog experiments are cautious as Tousey reports them [read: p.537-540]. Membrane account: now readable in the second scan (OCR): tissue currents "principally ... concentration cells since no metals are present", "two electrolytes with different degrees of concentration separated by a semipermeable membrane" (Ostwald, 1890), the injured place negative, Tschagowetz's calculated EMF said to agree "quite closely" [read: ia-1921-ucla.txt lines 17352-17430, OCR]. Bernstein's pre-existing-gradient view is the one modern physiology kept (Nernst example above, -89 mV). It is a quotation of other authors with no equation printed, not Tousey's derivation. Disagreement: see D5 for the electrolysis-only claim, and the label for "deep action of ionic medication" (contrarian: Tested, not supported, as if Tousey claimed deep action). Tousey does not claim deep action; he reports surface action with the deeper effect due to the current. The contrarian's verdict applies to the version that a reader might assume, not to Tousey's text; I correct it. What would settle it: for electric sleep, a sham-controlled trial (modern cranial electrotherapy reviews found insufficient evidence for depression and low-strength evidence for anxiety, none for anaesthesia [web: https://en.wikipedia.org/wiki/Cranial_electrotherapy_stimulation; primary reviews not opened]); for the membrane account, a deep read of pp.282-287 from page images.

Final verdicts [QE revision R7 revises the electrolysis and 39 m/s lines]: - Body is an ionic (NaCl-type) conductor; tissue currents originate in membrane concentration cells: Established (the concentration-cell account is Historically superseded but instructive in its details, kept in the resting-potential picture). - Medicinal ions act chiefly at the surface; deeper effect is the current: Established (broad form); specific ion cures (lithium for gout, salicylate for rheumatism, and others) Never properly tested. - "Almost all physiologic effects are electrolysis and electric osmosis": split. Strict form (chemical decomposition at electrodes explains stimulation): Refuted by calculation (Faraday, D5). Loose ionic form: Historically superseded but instructive (membrane excitation, Hodgkin-Huxley, and Joule heating [memory, high] explain most effects). - 39 m/s as the speed of electricity in the body: Internally inconsistent, and Refuted by calculation as a signal speed. - Leduc electric sleep and anaesthesia in man: Never properly tested.

Corrected claim list

(Updated by the QE revision of 2026-10-09; changed lines are marked "revised".)

1. The "Tousey dose" (6-inch spark, 3 mA, 10 min through 3 mm Al at 10 in) is said by him never to cause dermatitis. His own units label it an erythema dose, and the book's own conversions of those settings span about 0.16-1.5 erythema doses (Internally inconsistent in his units, not minor; revised). My estimate of the delivered skin dose is 0.3-1.1 Gy, below modern erythema thresholds, in line with the lower period calibrations (Cabot and Dodd; the 57 kV figure). Skin-cancer x-ray treatment is Established; "never" is Never properly tested; "usually no visible reaction in one sitting" is Plausible, untested. Cumulative operator injuries and the pregnant-cat result stay in the record (revised).
2. The condenser theory of x-ray burns is Geyser's, not Tousey's; as the sole cause it is Tested, not supported, and Tousey cites the experiments against it (p.1158) and rejects it (p.1162); a small extra electrostatic component is Never properly tested. Burns within 24 hours from tube ultraviolet (p.681): Never properly tested. Tube superiority and priority: Never properly tested (revised).
3. High-frequency benefits (blood pressure, diabetic sugar and the rest): Never properly tested, with Apostoli's 500 cases and the named small series listed beside the contrary reports; heating Established. The book says glass stops the ultraviolet and puts its source in outside sparks (Established); the claimed benefits of UV and ozone from those sparks are Never properly tested (revised; the first-pass "germicidal UV through glass: Refuted by calculation" is withdrawn as a verdict on Tousey).
4. Focal infection: Historically superseded but instructive; "every case of cancer": Never properly tested; arch formula: geometry internally consistent, anatomical fit Plausible, untested, outcome Never properly tested (revised wording).
5. Scatter and grids: Established; tenfold rule: Plausible, untested; his plate example: geometry wrong; tissue "fluorescence" image formation: Refuted by calculation in the strong form; Tousey unit: Historically superseded but instructive (and it does not transfer to radium: R8).
6. No over-unity: Established, and Tousey's own p.156 experiment explains the apparent excess by arc-voltage collapse (revised); constant potential: Established; monochromatic hope: Refuted by calculation, and the book reports a GE test that did not support it (revised); Cabot-Shaw priority: Never properly tested, restated as the x-ray application of a generator Cabot was already developing (revised).
7. Ionic body and surface-acting ions: Established; electrolysis-only: Refuted (strict) or superseded (loose), and contradicted by the book's own pp.301-302, 306, 340 (revised); 39 m/s: Internally inconsistent (180,000 mi/s on p.289, a conditional 39 m/s on p.296), Refuted as a signal speed (revised); electric sleep in man: Never properly tested.

What would settle it (one deciding check per open point)

(Additions are in QE revision R11: Cabot and Dodd's paper, the NYHS and Pennsylvania archival holdings, the original French reports, Strater, Oudin and Scholtz, Bierry, Henri and Ranc.)

• Dose of the "Tousey dose": spectrum and output of a 6-inch-spark coil-and-gas-tube set at 10 in (a replica or a published 1920s calibration), converted to skin dose. Or his own patients' follow-up records.
• Cabot-Shaw priority: Cabot and Shaw company papers, Williams of Boston, Tousey's 1910 text on the relevant "three or four pages".
• Contact-tube priority: Dental Cosmos 1906, NY Odontological Society minutes (17 Oct 1905), Geyser JAMA 1908.
• High-frequency effects: a sham-controlled trial; a deep read of PDF 544-647 now.
• Arch formula: arch outcomes in children whose temporary arch was regulated by the formula, or a comparison with measured adult arches.
• Dental infection in cancer: dental radiographs of cancer patients against matched controls.
• Electric sleep: a sham-controlled trial of interrupted low-current head stimulation.
• Membrane passages: deep read of pp.282-295 of the second scan from page images.
• Oudin power: measure V, I and phase on a resonator to get real power.

Open questions

• Was the lupus case the same patient as the nose epithelioma (both a woman of 67, father lost a nasal ala)? Source: Dr. Jarman's case notes, not obtained.
• What did the "(5)" H in his table mean (measured, computed, quoted)? [Partly answered by QE revision R1: the book gives eight conversions of those settings that disagree with each other; which, if any, is the measured one is still unknown.]
• Does the book anywhere cite Kienböck 1900 against the condenser theory? [Answered by QE revision R2: yes, p.1158, with Strater, Oudin and Scholtz; Tousey rejects Geyser's theory at p.1162.]
• What does "506,000 Gauss" in Doumer's cage refer to? Doumer's original is needed.
• Who is "Doctor Williams of Boston"?
• Is 0.926 cm/h for sodium and 0.166 for silver a misprint of Kohlrausch's table? Only Kohlrausch's source can say.
• Did the book's reader community (Energetic Forum, Firstenberg, de la Peña [web, via Investigator]) use it for claims it does not make? That explains the Energy folder, not the physics.

Absence-of-evidence analysis

The record does not go dark for the man. It runs from 1864 to his death on about 25 Sept 1937 (NYT obituary 26 Sept 1937, radium gift to Morrison Hospital, late fiction in 1934-35) [web, via Investigator]. It goes quiet in five places where more would be expected.

1. Cabot, Shaw and Williams on the generator priority. Expected: a reply, a company history, or a Cabot-Shaw paper on x-ray supply. Found: none (Investigator did not reach Espacenet, Massachusetts archives or the Medical Record/Archives of the Roentgen Ray full text). - Explanations: (a) ordinary neglect of a minor credit dispute, since Cabot was a wireless and converter engineer, not a medical author; (b) the generator was a side product of Cabot's converter work, so the x-ray application was not documented separately; (c) the company's papers were lost or are in a private collection; (d) Tousey overstated his role (his own stake); (e) no one chose to contest it. Evidence for (a) and (b): the patents are generic "electric conversion" documents with no x-ray text [web: Google Patents, via Investigator]. For (d): the only account is his own. Nothing found for suppression. - Deciding record: Cabot and Shaw papers, Williams' correspondence, 1912-1913 x-ray trade press.
2. The Tousey unit after 1922. Published 1913, abstracted by JAMA, used by others in 1916, used by Tousey in 1922, then absent. Explanations: replaced by the roentgen (adopted 1928 internationally [memory, medium]) and the erythema dose, ordinary succession; the book itself says it "is no longer required". Nothing found suggesting suppression. Deciding record: radiology dosimetry histories of the 1920s (Mould is cited, not read).
3. Geyser's side of the contact-tube dispute. Only reported through Tousey and Wikipedia. Explanations: a minor dispute between two interested practitioners; Geyser's papers exist (JAMA 1908) but were not read. Deciding record: Geyser's own 1908 paper.
4. Commercial fate of Tousey's tube and patents (seven US patents 1911-1925, five probable Canadian and French family members [web: Google Patents, via Investigator]). No sales, licences or assignee history was sought. Explanations: manufacture by an established maker (the book says the tubes were made by the same manufacturer as the Cornell tube [read: p.1232], name not carried here), commercial failure, quiet absorption into later products. Deciding record: assignee data in the patents, trade catalogues.
5. Wartime fire-control and firearm-aiming patents (filed 1917). No war contract record sought. Explanations: a private invention not adopted; secrecy; a patent that went nowhere. Deciding record: the patents and contemporary War Department files. No sign of suppression for any of the five. These are leads, not a conclusion. The contrarian's finding stands: this is mostly ordinary medical-science succession, not a documented suppressed anomaly. The one place where the establishment was first wrong and later corrected is the focal-infection doctrine.

Corrections to earlier documents

(The QE revision adds items R10.1-R10.9 at the end of this file; apply them with the items below.)

1. deep read (Emailing...42fec2.md), section on ionic medication, "sodium (OCR garbled ... probably about 0.1 cm/h)": the page image reads sodium 0.926 cm per hour, silver 0.166, lithium 0.094 [read: page image PDF p.405]. The printed values do not match modern mobilities (D6). Use 1.4-2.3 cm/h per V/cm from modern mobilities.
2. deep read section 5.4 ("both factors are peak-like"; true power about D x V_peak x I_peak): the mathematician is right that a moving-coil milliammeter reads the average. The overstatement comes from peak voltage, mean voltage during conduction, inverse discharge and, for the Oudin, reactive power (D2). Replace the duty-cycle explanation with the stacked one.
3. deep read section 3 and claims.md claim 5 ("the author does not reconcile" the tenfold and halving rules): they apply to different quantities; the tenfold rule is plausible (D3 and claim 5). Delete "unreconciled".
4. claims.md claim 1: add that the "Author's dose" prints "(5) H" and "(10) X" in parentheses, that 5 H is labelled an erythema dose, and that my estimate of the delivered dose is 0.3-1.1 Gy (D1).
5. claims.md claim 3: add that on printed p.585 Cohn, Boedeker and Vinaj found no change in urine sugar, that Denoyes could not separate diet, and that his diabetic cases also had x-ray over pancreas and liver. 2700 grains is 175 g, not 180 g.
6. claims.md Source note: the "folder contains no patent" line and the "cannot be read" line on the missing pages are out of date. Seven US patents exist (first US1100037, filed 1911-02-23); the missing pages are in a second scan (IA medicalelectrici00tousiala). Both are the Investigator's finds, confirmed by all three reviewers.
7. sites.md "Leads for review", Geyser: it sets Tousey's 1911 patent date against the Cornell tube's 1905. His book dates his own tube to 17 Oct 1905; the patent date does not decide priority.
8. facts.json Part 11: add von Laue, Friedrich and Knipping (1912) to the x-ray diffraction conflict note [memory, high].
9. Deep read "x-ray energy proportional to the fourth power of velocity": likely a garbled Whiddington range law (range proportional to v^4, 1911) [memory, medium]; a source slip, not a new claim.
10. deep read PDF 698-748: "Rontgen discovered x-rays April 30, 1895"; the usual date is 8 Nov 1895 with publication 28 Dec [memory, high]; a source error to flag.
11. digest p.1171 "X = 4 H" against the deep read "1 H = 2 X": prefer the deep read (a page-image reading); unresolved here.
12. Contrarian, claim 5 verdict "Internally inconsistent" for the tenfold and fall-off statements: not sustained (above). Contrarian, claim 7 "Tested, not supported" for deep action of ions: Tousey does not claim it (above).
13. Physicist, claim 1 dose "1.1-1.8 Gy": the base output of 0.03-0.05 mGy/mAs is above what I could source (0.023-0.030); revised to 0.3-1.1 Gy. Mathematician, claim 1 "0.25-0.5 Gy": double-counted the 3 mm Al filter (D1); revised.

Web verification log (budget 8; used 5; see also QE revision R9)

1. Output of a 70 kV tube per mAs: 0.023-0.030 mGy/mAs at 1 m [web: https://www.rug.nl/education/courses/other-education/radiation-protection/edu/ts-mt/ts-mt-data-examen/mcts-mtfiguura?lang=en]. Checked (search summary; column alignment ambiguous).
2. Holzknecht and Sabouraud-Noiré units; roentgen equivalent of an H: not found; 1926 paper on observer disagreement [web: https://pubs.rsna.org/doi/full/10.1148/6.3.236]. Check failed for the conversion; stays unconverted.
3. Skin erythema threshold: about 2 Gy lower limit, 2-5 Gy transient erythema [web: https://iaea.org/resources/rpop/health-professionals/radiology/erythema]. Checked (IAEA statement; ICRP primary not opened).
4. Glass UV absorption and the 337 nm nitrogen band [web: https://image-ppubs.uspto.gov/dirsearch-public/print/downloadPdf/6878653]. Checked (patent text via search; approximate).
5. Air attenuation (QE revision R9): NIST XCOM, https://physics.nist.gov/PhysRefData/XrayMassCoef/ComTab/air.html. Checked; D3 figures stand. Not web-checked, still [memory]: (air attenuation coefficients now checked, item 5) power factor of an Oudin resonator (medium), scatter-to-primary ratios (high), Duane-Hunt limit (high), Nernst potential (high), 1928 roentgen adoption (medium), H. pylori (high). None changes a verdict.

---

QE revision (2026-10-09)

Arbiter, second pass. Inputs: qe.md (verdict MINOR), and the eight new deep reads appended to deep/Emailing-3a-Medical-electricity-and-Rontgen-rays-42fec2.md (violet and ultraviolet, PDF 557-681; the coil, meters and Tousey's apparent-power experiment, PDF 144-217; the Cabot and Shaw generator and erythema calibrations, PDF 749-773; radium, PDF 1252-1297; the mechanism of stimulation, PDF 297-349; clinical results for high-frequency currents and diathermy, PDF 544-647; dose conversions, burns and the pregnant-cat experiment, PDF 1130-1180). The earlier sections above are kept as written; passages this section replaces are marked "superseded" in place. I re-checked the quotes on which the QE findings turn against the extracted text (p.156 "10,000 volts to originate"; p.289 and p.296 on 30 m/s, 180,000 mi/s and 39 m/s; p.578 "hardly to be supposed"; p.603 Apostoli; p.763 and p.759 "3400 milliampere seconds"; p.1158 and p.1162 on Geyser) [read: _text, lines 8009, 15858, 16265, 32156, 33665, 41590, 41876, 59903, 60195]. Tags as above; [calc] is my arithmetic in the project venv.

Result. The QE finding stands: no claim moves to a different place on the scale in its main form. Seven verdicts are re-targeted, split or relabelled, because the book says more than the first pass credited (and in three places says the opposite of what the first pass attributed to it). The changes:

R1. Dose of the "Tousey dose": the book's own conversions (replaces the "no calibration on record" statement in D1 and in the open questions)

The first pass said no period calibration of the "(5) H" was on record. That was wrong. The book converts settings to erythema doses in at least eight places. Author's dose A: 6-inch spark (7 Benoist, about 70 kV), 3 mA, 10 in; unfiltered 5 min = 900 mA-s; or 10 min through 3 mm Al = 1800 mA-s [read: p.1184 note, p.1181 table]. Fractions of an erythema dose (E), taking E = 5 H where the book gives H:

Notes on the rows that need care. - Row 7 and the voltage. The deep read shows QE's table cell was wrong: p.759 does not name Dodd and gives 90 kV only as an illustration of the rule "10,000 volts = 1 Benoist"; the experimenters, the hospital and the measured penetration (4 Benoist, about 40 kV) appear only on p.763 [read: deep read of PDF 749-773; _text lines 41590, 41876]. The unscaled 0.26 hides the dependence on kV. Scaling to 70 kV by Tousey's own output rule (output proportional to V squared, p.741) gives 0.81 if the 3400 mA-s applies at 40 kV (the better-attributed reading) and 0.16 if at 90 kV [calc]. The V-squared rule is crude for skin dose, so I give the range, not a point. - Row 6 and the hardness. The Tousey scheme is stated "of about 6 Benoist"; the author's spark is 7 Benoist. The book itself says doubling the voltage doubles the pastille (skin) reaction and quadruples the film effect (p.1171), so a film-based erythema time reads long for hard rays. That is the direction of the discrepancy, but the book does not make the argument or measure it (inference). - Row 3 and 2 agree that the filtered form is just under an erythema; rows 1 and 4 put it at or over one; rows 5-8 put the unfiltered form at a quarter to a half. - The book warns against all of these: "milliampere seconds do not form a reliable measurement of x-ray dosage" (p.763); the Tousey unit "is no longer required" (p.1173); the milliammeter on a gas tube is "dangerously misleading" (p.1181); McKee's factor of 4 between pastille positions (p.1168); and p.1177 (Bordier) shows the dose changing with interrupter rate (19 min per tint at 1938 breaks a minute, 10.5 min at 3544).

Weighing. The book's own scales put the author's dose between about 0.16 and 1.5 erythema doses, a spread of a factor of 6 to 9. The two conversions that carry a named experimenter and a measured setting (rows 7 and 8, Cabot and Dodd; the 57 kV figure) and my independent modern estimate agree that the dose was well below an erythema dose in one sitting. My first-pass estimate was 0.3-1.1 Gy at the skin, central 0.7 Gy [calc, D1]. That is 0.15-0.55 of the 2 Gy "reaction unlikely below" figure [web: https://iaea.org/resources/rpop/health-professionals/radiology/erythema], and 0.05-0.2 of the 5-6 Gy "main erythema" figure that I confirmed only through search summaries, with a factor of 2-3 uncertainty either way. The conversions that put the dose at or above an erythema (rows 1, 4) are Tousey's own table and directions, and they are the ones he qualifies himself in the same chapter (rows 3, 5). I cannot settle between them from the book alone.

Verdicts for claim 1 (revised): - "The author's dose never causes dermatitis": Never properly tested (unchanged; a clinician's report, no series). - "Usually no visible reaction in a single sitting": Plausible, untested. Support is now two period calibrations, the book's own p.1173 observation, and my modern estimate; against it, rows 1 and 4. - The dose label "(5) H" and the units: Internally inconsistent, no longer minor. In his own units the same settings read from about 0.16 to 1.5 erythema doses. The inconsistency is in the book's conversions, not in the clinical statement. - Cumulative injury is a separate matter that "never" does not cover. The book records chronic operator dermatitis "produced by frequently repeated exposures which individually would produce no effect" (pp.1154-1155), his own hand keratoses (twelve years; Plate 14 caption, February 1915, January 1920; cause given as his "former practice of holding the film in the patient's mouth in dental radiography", p.1159), deaths of named x-ray workers from x-ray cancer (Kassabian, Caldwell, Leonard, Greene, Bauer, Baker; p.1155), a patient who needed amputation after a general x-ray application for warts (p.1180), and the worst case he saw, gangrene of the hand of an apparatus seller (p.1154). He also writes that a burn is no more likely to be followed by cancer than any lesion of similar severity (p.1154), and, one page later, that x-ray warts carry "great danger" of epithelioma (p.1155). The book does not reconcile these. All stay in the record next to the verdicts [read: deep read of PDF 1130-1180].

Pregnant-cat experiment (retained; the first pass omitted it) [read: p.1142-1143, with the settings table]. In 1904 Tousey radiographed a cat that had borne several healthy litters, seven exposures from 6 Oct to 17 Dec 1904 (3-8 in spark, 8-11 A primary, 15-19 in to the plate, 10-65 s each, 276 s in all; no tube current given), holding her with his own hands in the beam ("this was a dangerous experiment"). Three mature, perfectly formed kittens were born dead on 30 Dec; with exposures stopped, four healthy kittens followed on 22 Mar 1905. Limits: one animal, one litter, no control litter, dose not recoverable (no mA), other causes of stillbirth not discussed, and cat gestation of about 63-65 days [memory, high] puts the October exposures before conception, so about 170 s fell in gestation [calc]. Verdicts: that rapidly dividing and embryonic tissue is more radiosensitive than the adult tissue: Established (the book cites Bergonie and Tribondeau's 1905 rat-testis work, p.1145; modern radiobiology agrees [memory, high]). Tousey's single-animal result as evidence: Never properly tested.

R2. Claim 2: the condenser theory is Geyser's, and the book rejects it (answers the open question)

• p.1158: "The experiments of Strater, Kienbock, Oudin, and Scholtz seem to leave no room for doubt that radiodermatitis is due to the x-rays and not to electric discharges and other physical phenomena which take place around the tube." The same page cites Bordier and Salvador (1899) on electrolytic action in erythema, adding that "the present author believes that the x-ray is the exciting cause of the electrolytic action" [read: p.1158].
• p.1162: "Geyser has suggested that the cause of dermatitis is not the x-ray, but an electrostatic condition which may be avoided by placing the tube directly in contact with the part to be treated. The author does not regard this as probably correct, and recommends avoidance of overdosage when the Rontgen ray is applied in this way" [read: p.1162].
• p.1231 (cited in the first pass): the theory "has not been proved" but "probably contains a certain element of truth" [read]. The book does not reconcile p.1162 with p.1231. The consistent reading is that p.1162 rejects the electrostatic condition as the cause in place of the x-rays, while p.1231 leaves room for a small electrostatic addition. That is my reading, not the book's statement.
• The book names the four workers and gives no detail of their experiments. The Kienbock 1900 result is documented by the Investigator [web: https://pubmed.ncbi.nlm.nih.gov/25012643/]; I have no source on Strater, Oudin or Scholtz beyond Tousey's sentence. Tousey is an interested party (his own contact tube competes with Geyser's), and so is Geyser.
• The first pass stated that the book was digest-only here and that Kienbock was "not found". Both are superseded. Open question 3 is answered: yes, p.1158.
• Third mechanism, p.681 [read]: "A burn appearing within twenty-four hours after an x-ray exposure is doubtless due to ultraviolet rays generated by the x-ray tube. This source of injury may be suppressed by interposing any screen opaque to ordinary light." The cited work (Bierry, Henri and Ranc, C. R. Soc. Biol. lxx, 523, 1911) concerns UV from a Crookes tube and its effect on carbohydrates; the text does not say that paper reported the burn claim. No experiment of Tousey's is given. The book's own glass test (p.678) says ordinary glass abolishes UV fluorescence, so UV made inside a tube would have to leave through the glass wall; the book does not discuss that. A mainstream alternative exists for a burn within a day or so: early transient erythema is a dose effect of x-rays at about 2 Gy [web: IAEA page above, as used in D1]. I do not have a source that tested the tube-UV explanation either way.

Final verdicts for claim 2 (revised; the earlier list is superseded where it differs): - Electrostatic ("condenser") condition as the cause of x-ray dermatitis instead of the x-rays (Geyser's theory): Tested, not supported (Strater, Kienbock, Oudin, Scholtz as reported in the book itself; Tousey rejects it). - A small electrostatic contribution in addition to the x-ray dose ("an element of truth", p.1231): Never properly tested. - Heating by tube capacitance as a burn mechanism: Refuted by calculation (this pathway only; unchanged). - Early (within 24 h) burns from ultraviolet generated by the tube (p.681): Never properly tested. - Static charge speeds ulcer healing: Never properly tested (unchanged). - Tousey tube superior to the Cornell tube, and priority: Never properly tested (unchanged).

R3. Claim 3 and D4: what the book says about ultraviolet from the vacuum electrode (replaces the D4 quote and the "Refuted by calculation" verdict as applied to Tousey)

The first pass quoted the book as saying that "ultraviolet rays absorbed by the body" matter, from "the PDF p.585 region, text lines 32139-32166". The lines are PDF p.578, and the sentence is: "It is hardly to be supposed that the ultraviolet rays absorbed by the body during a treatment by means of vacuum electrodes takes any more direct part in the physiologic or therapeutic effect" [read: p.578, line 32156]. The first pass inverted its meaning. What the book says [read: deep read of PDF 557-681]: - p.557-558: "there is really no such treatment" as "violet-ray" treatment; "the ultraviolet ray will not penetrate the thinnest piece of mica or celluloid, the results ... are due to the form of current applied, and not so much to the ultraviolet rays, which are given off from the sparks outside of the electrode." - p.578: the visible light has "no demonstrable physiologic or therapeutic effect"; UV arises in the tube gas and, especially, in minute sparks along the outside of the tube; it liberates ozone and is absorbed in the most superficial skin; its direct role is "hardly to be supposed". - p.579: the role he claims is indirect, through ozone and nitrous acid: oxidation, nitrogenous matter leaving as urea rather than uric acid, "vacuoles filled with ozone" in tissue. No source or experiment is given on the page. - p.611, 631, 635: for gout and acne he tells the operator to choose the electrode giving "the greatest amount of ultraviolet radiation, as detected by the fluorescence caused in a piece of Willemite"; for acne "a germicidal effect from the ozone ... and from the ultraviolet ray produced around the vacuum tube"; for skin disease the effect is "attributable in part" to UV. p.658 says the tube's own light contains "an appreciable amount" of UV. The book does not reconcile p.557/578 with p.631/635/658. - p.678: "if the fluorescence is due to ultraviolet rays, the glass will cause the fluorescence to disappear entirely"; the solio-paper test shows glass blocking and quartz passing UV.

Weighing. The strong claim that I refuted (germicidal short-wave UV leaving a soda-lime vacuum electrode) is not the book's claim; the book says glass and mica stop the UV that excites Willemite, which agrees with the glass-absorption data in D4 (soda-lime absorbs below about 300 nm; near-UV at 337-390 nm is attenuated, not blocked). One consequence follows from his own test: a Willemite glow near the electrode reports the glass-blocked UV, so it reports the outside sparks, not UV through the wall. That is my inference from his p.678 test, not his statement. What the book does claim is a germicidal and tonic benefit from ozone and from UV made by sparks outside the electrode, in acne, skin disease and as a systemic "oxidation" effect. That is testable and untested.

Final verdicts (revised): - UV through the glass wall of a vacuum electrode: the book does not claim it and says glass stops it; consistent with accepted absorption data: Established. The calculation in D4 stays as the explanation, and "Refuted by calculation" applies only to the proposition, not to Tousey. - Ozone and nitrous acid from the discharge: Established as products of an air discharge [memory, high]. Their therapeutic effects as he describes them (oxidation, urea in place of uric acid, ozone "vacuoles"): Never properly tested. - Germicidal or tonic effect of UV from outside sparks (acne p.631, skin p.635): Never properly tested. - The "506,000 Gauss" figure: unchanged, see D4 and R4.

R4. Claim 3: the clinical evidence, both sides (deep read of PDF 544-647; replaces the "digest level" limit for the tabulation)

All of Tousey's reporting on others is second hand: a journal citation, not the data. I list what the book gives, with named sources, case counts and co-treatments. None of it is controlled, blinded or randomised [read: deep read, section 12].

Blood pressure: - Positive reports: Guilleminot, 15 cases, 21 to 17 cm Hg after 15-20 treatments, "permanent effect" asserted, no interval (p.559); Moutier, "many" cases, first-treatment falls of 3-9 cm to about 15 cm, lasting "several weeks" (p.582-583); Gidon, 5 cases, "permanent reduction of arterial tension from 240 down to 160 mm", with diet unmodified in case 1 and mechanotherapy, sinusoidal current and Oudin effluve added in case 5 (pp.593-594); Gay, 12 neurasthenic cases of which two are printed (BP 160 to 120 and 180 to 135 mm Hg; pulse 70 to 80 and 80 to 72; urea 12 to 16 and unchanged; uric acid 3 to about 1.5 and 5 to 2), the rest summarised as "similar changes", several seen again "some months after" with the improvement permanent (p.595); Doumer, 4 cases, 27 to 13.5, 26.5 to 16, 29.75 to 17.75 and 26 to 16 printed as "mm" (page image confirms), which read as cm Hg magnitudes, so the unit is unresolved (p.584). - Contrary reports: Delherm and Laquerriere, "considerable experience", Moutier's technique followed "as closely as possible": the immediate fall "not as constant or as marked", the permanent fall "decidedly less marked and less constant", no sure prevention of arteriosclerosis (p.584) (an uncontrolled replication attempt with a weaker result, number of patients not given); Boedeker and other German authors: a rise averaging about 5 cm Hg for about twelve minutes (pp.582-583), which Tousey attributes to apparatus differences; Carvallo, no effect from the cage (p.579); d'Arsonval: "some patients show a material fall and others a material rise" (p.580); Wertheim Salomonson: the temperature rise (surface up to 6 F at 250 mA and 14 F at 600 mA, p.582) is simple Joule heating, which he thinks accounts for bipolar, couch and Oudin use. Tousey's own cases showed "a more gradual improvement", not Moutier's abrupt fall (p.584). - The heating explanation matters: it is the physicist's estimate (D4, claim 3 settle-it) arriving from inside the book. The book also contradicts itself on heating, saying the cage gives "No important rise of temperature" and, a few lines later, that the therapeutic effect is "due to the heating of the body" (p.646).

Diabetes: - For: d'Arsonval, Charrin, Boinet and Poncy, Apostoli and Berlioz, Reale and Renzi, Vinas: general condition "uniformly" better, sugar markedly reduced and in several cases gone (p.585). Apostoli at the Twelfth International Medical Congress reported success "in over 500 cases", average 25 treatments in the cage or on the couch, followed by his pupils Williams, Allen and Tousey (p.603). No sugar figures are given for the 500; this is documented in one secondary account (Tousey), and the congress paper is not in the folder. The first pass listed the contrary authors and omitted this; that was a one-sided condensation (QE retention defect) and is corrected here. - Against: Cohn (sugar unchanged), Boedeker (sugar unchanged, acetonuria cleared during treatment and returned on stopping), Vinas (cage, 4 cases, sugar not cleared), Denoyes (diet such that the effect on sugar "could not be determined") (p.585); Boedeker, Cohn, Loewy, Doumer and Halfon found "no reduction in the amount of urine" (p.603); DeRenzi: sugar returned in cases of exaggerated denutrition (p.603); Freund gives no opinion. The p.585 contrary reports are about sugar and the p.603 list is about the amount of urine, and Apostoli's endpoint is unspecified. - Tousey's own cases: sugar "from 2700 grains (180 grams) per diem to a mere trace" (p.585; 180 g is 2778 grains, 2700 grains is 175 g [calc]); a lady of about 40, sugar 10 percent to under 1 percent in three weeks with urine volume 56 to 55 oz; a gentleman of 55 with 10 percent sugar, whose weight loss slowed, and who "made a good recovery" under static electricity from his regular physician (pp.603-604). His caveat: "It has not been the author's habit to ask for a discontinuance of medicinal treatment, and so his own cases can hardly be cited as absolute proof that the electric treatment alone would have been effective" (p.604). Arsenic, x-ray over the pancreas and liver, and likely diet were also in play. The first pass left this hedge out. - Context: insulin arrived 1921-22 [memory, high].

Other items in the record, kept: chronic rheumatism, three cases, in the first two the improvement came "before she had begun taking salophen" (pp.605-608); a case of tuberculosis of lung and larynx treated 1903-04 with x-ray, mercury-vapour light and the electrode, improved, relapsed after treatment stopped, and died May 1905 (pp.613-614); a gonorrhoea case in which x-ray plus the urethral electrode did not clear the gonococci (p.591); "There is nothing magic about the high-frequency currents" (p.619); Keating Hart fulguration: "cannot recommend" (p.567); Tousey's statement that the energy gives "a tonic effect upon any person within 10 feet" and that he enjoyed "extraordinarily vigorous health" for four to five years (p.581, his own report); Millener's 22 power-house workers (pp.302-303); the 1904 radiated-saline experiment (R8).

Final verdicts for claim 3 (revised; replace the earlier list where they differ): - Radio-frequency deep heating, and no nerve stimulation at these frequencies: Established (unchanged; the in-book Joule-heating account of Wertheim Salomonson fits). - Lasting lowering of blood pressure by cage or couch: Never properly tested. Several small uncontrolled series report it; one uncontrolled replication attempt (Delherm and Laquerriere) found it weaker and less constant, and German workers reported a rise. Not dismissed untested, and not controlled. - Diabetic sugar fall: Never properly tested. - Raised metabolism, goitre, impotence, ataxia, sciatica, rheumatism and the other listed cures: Never properly tested. - UV and ozone: see R3. - "506,000 Gauss": Internally inconsistent as a unit (unresolved; source not checked). Doumer's BP units show the same kind of slip in transcription. - Confidence: the evidence description is now at deep-read level (medium); efficacy conclusions remain low confidence because nothing in the chapter is controlled.

R5. Claim 6 and D2: Tousey's own "output greater than input" experiment

Pages: PDF p.156 (printed 154), p.155 (printed 153), p.214 [read: deep read of PDF 144-217]. Tousey's experiment: a hot-wire milliammeter rated to 650 mA on one pole of a 12-inch coil, Wehnelt interrupter, 110 V and 6 A primary (660 W; "about 600 watts", part lost in the interrupter), a 4-inch flame, "a voltage of 40,000 and the milliamperemeter indicated an intensity of 150 milliamperes. Here there was an apparent power of 6000 watts produced by the application of about 600 watts ... The apparent discrepancy is due to the fact that the voltage is greatly reduced during the passage of the continuous flame, just as 100 volts will maintain an electric arc 1 inch long which it would take 10,000 volts to originate" [read: p.156, line 8009]. - Arithmetic [calc]: 40,000 x 0.150 = 6000 W, a ratio of 9-10 to the input. The 40 kV is read from the spark length (4 in at 10,000 V per inch, p.217; p.151 gives 10,000-30,000 per inch, so 40-120 kV), not measured. To balance 600 W at 150 mA, the mean voltage across the running flame must be at most 4000 V, one-tenth of the striking figure, and less once the interrupter loss is allowed for. His own analogy implies a factor of 100. So his explanation is quantitatively sufficient on his own numbers, and by his own rule on p.155 (a coil cannot return more than goes in) 600 W at 40 kV allows at most 15 mA. - This is a second, documented passage of "apparent output above input", and it comes with its own explanation. The first pass said that on the book's evidence the ledger did not close for an averaging meter. Superseded in part: with a voltage collapse of tenfold to a hundredfold, as in p.156, the ledger closes for any meter. My earlier check (half the voltage throughout gives 750 W against 500 W) still shows that "less than one-half" on p.740 alone is not enough; the p.156 factor is. - Meters, in Tousey's own words: the D'Arsonval meter "indicates in a general way the average strength and direction" and can read zero while the tube gives a good picture (p.155); a hot-wire meter is "not influenced by the amount of inverse discharge" (p.156) and, on alternating current, reads "somewhat more than the true figure", so that volts times amperes "will often be found to exceed the amount of power expended in running the dynamo" (p.214). The p.741 sentence ("the meters indicate the maximum of the successive currents rather than their average") is therefore inconsistent with his own p.155 for a moving-coil meter and is loose even for a hot-wire meter (which reads a heating-weighted value, not a maximum). The mathematician's point is borne out by Tousey himself. - The book gives two explanations of apparent excess (arc-voltage collapse, p.156; hot-wire weighting, p.214) and does not say which applies to a given device. For the Oudin resonator the reactive-power point in D2 stands; no phase measurement is in the book.

Verdicts (unchanged): output above input is a meter and voltage artefact, no over-unity: Established; constant potential is the ideal supply: Established.

R6. Claim 6: the Cabot and Shaw generator, as the book tells it (deep read of PDF 749-773)

• Tousey states that in the first edition (page 663) he expressed the opinion that a constant, unidirectional, regulable high-voltage generator would be preferable to the coil, static machine, rectified transformer or Grissonator, and that the apparatus "came about from the perusal of these paragraphs by Messrs. Cabot and Shaw", adding: "Mr. Cabot, having already been at work upon a means of generating high-tension direct current for electro-metallurgy, they decided to apply their invention, in the first place, to the generation of current for x-ray purposes" [read: p.755]. So his own account is narrower than "built from his pages": he credits his pages with the x-ray application of a generator Cabot was already developing. That fits the generic wording of Cabot's patents (claim 6 above; Investigator). The maker's account is not in the folder; the makers (Cabot, Shaw, Electric Conversion Co., Brookline) and Tousey all have a stake.
• Stages: the "heterodyne" (stepped-up 220 V alternating current, condensers and choke coils; an oscilloscope showed a continuous band; the rays were "believed to be approximately homogeneous"), then the Cabot constant-voltage generator (220 V direct current in, 10,000-140,000 V continuous direct current out, loss "about 1 or 2 per cent.", Gramme-ring arrangement of three transformers with advancing brushes) [read: pp.756-758]. Tousey says the heterodyne wasted too much power and was hard to regulate (p.757).
• Hope against test: on p.764 Tousey reports that a Hull apparatus built by General Electric found the spectrum from constant potential "the same as that of the ordinary transformer with a high-tension rectifier. This does not confirm my hope in regard to securing homogeneous x-rays", with tungsten giving much the same spectrum whatever the momentary voltage; but that the unvarying current spares the Coolidge focus spot. So the book reports a test that did not support the hope. The method of that spectral comparison is not described; GE had a stake.
• mA-seconds. Tousey uses 3400 mA-s at 10 in as an erythema calibration (R1), states that equal mA-s give equal results at a given voltage (p.760, p.762), and says in the same place that "milliampere seconds do not form a reliable measurement of x-ray dosage" unless voltage, tube and distance are standardised (p.763). The book also reports GE needing less mA at a given voltage than with a rectified transformer (p.764) while Tousey says the same mA is needed (p.760). Both are recorded.

Verdicts (revised): - Monochromatic x-rays from constant potential: Refuted by calculation (continuous bremsstrahlung up to the Duane-Hunt limit [memory, high]); and the book itself reports a spectral test (GE) that did not support it. - Cabot and Shaw applied an existing high-tension direct-current generator to x-ray after reading Tousey's call for one: Never properly tested (one interested-party statement, 1913 and the book; compatible with the patents; not corroborated or contradicted by Cabot or Shaw).

R7. Claim 7 and D5: signal speed, ion drift, and the stimulation mechanism

• Signal speed [read: p.289, line 15858 ff.; p.296, line 16265 ff.]. PDF p.289: "The rate of progress of the wave of low potential, not over 30 meters a second ... does not indicate the rate of transmission of the same electric current through the tissues ... The velocity at which the current flows, possibly 180,000 miles a second". PDF p.296 (printed 300): the body "according to many observers conducts only by electrolysis ... This would mean that electricity can travel through the body at a rate not exceeding 39 meters a second, instead of at the rate of 180,000 miles a second." The premise ("only by electrolysis") is attributed to many observers; the inference "this would mean 39 m/s" is a conditional that is Tousey's own. QE's wording ("the observers' view as reported") is a little too generous to Tousey; the book states the correct distinction (p.289) and then draws the wrong conditional (p.296) without reconciling them. Verdict: Internally inconsistent (the book gives both numbers); as a signal speed the 39 m/s is Refuted by calculation (D6).
• Mechanism of stimulation [read: deep read of PDF 297-349]. The sweeping sentence (p.297) is hedged: "It is probable that in the last analysis almost all the physiologic effects ... should be considered as due to electrolysis and to electric osmosis". On p.330 "electrolytic" is "used in its broadest sense", meaning a migration of ions and a difference in osmotic pressure inside cells and fibres, with Nernst's membrane view named and no formula given. The book itself then lists stimulation without ion transport: condenser discharges "produce hardly any electrolytic or polarizing effect" (p.306); the d'Arsonval method "prevents polar or electrolytic effects" and the muscle still contracts (p.340); static sparks and high-voltage discharges act "by means of vibration of the tissue-cells, rather than by transportation of ions" (pp.301-302); high-frequency currents do not excite because "the transportation of ions through the semipermeable membrane ... cannot keep such rapid rhythm" (p.302). It offers three unreconciled accounts of why high frequencies do not excite (ion transport too slow, p.302; a 5000 per second cut-off from latency, p.326; a refractory gap of 2-12 ms against reversals shorter than a microsecond, p.339). Mainstream reading: the membrane behaves as a low-pass filter and excitation is threshold depolarisation [memory, medium; not web-checked].
• Condenser energy law. Tousey states that equal contractions need equal one-half C V squared, so the voltage varies as the inverse square root of the capacity, within "quite wide limits" (p.306), and reports in the same place that Lapicque's experiments to confirm it as "the law of stimulation" gave "negative results" and that "No one simple formula suffices" (pp.307, 330). One-half C V squared is the energy stored in the condenser, not what reaches the tissue [memory, high]. Verdict: Historically superseded but instructive (a descriptive fit that the book itself reports as not general).

Verdicts (revised): - "Almost all physiologic effects are electrolysis and electric osmosis", as the unqualified sentence reads: Internally inconsistent with his own pp.301-302, 306, 340 (and hedged by "probable" and "in the last analysis"). Split as before: strict chemical form Refuted by calculation (D5); loose ionic form Historically superseded but instructive. - 39 m/s as the speed of electricity in the body: Internally inconsistent; Refuted by calculation as a signal speed. - All other claim 7 verdicts unchanged. Add to D5: Tousey's own mechanism of stimulation (ion migration, osmotic differences, Nernst's semipermeable membranes, p.330) supports the "loose ionic" reading from the main text, not only from the second scan.

R8. Claim 5: the Tousey unit and radium (deep read of PDF 1252-1297); other experiments retained

• The Tousey unit: one Tousey meter-second is the effect on Kodak film of a 1 candle-power carbon-filament lamp at 1 m for 1 s (p.1259). For radium, the author's applicator No. 4 (about 20 mg, "activity" 2,000,000 times uranium) makes the same film density in 5 s bare, so 0.2 Tousey meter-seconds per second; 5 minutes is about 60 [calc from p.1260]. The book then says 5 minutes of contact gives "a pronounced dermatitis". Against the x-ray scale, where 500 meter-seconds is an erythema dose (p.1173), that is a factor of about 8 [calc]. The unit does not transfer between radiation qualities, which fits the book's own warning that film and skin respond differently to hardness (p.1171). Verdict unchanged: Historically superseded but instructive.
• Radium to H conversions (p.1281): 5 H in 12 h at activity 7,000; in 15 min at 300,000; in 5 min at 1,800,000. The rate is proportional between the first two (x43 activity, x48 rate) and not between the last two (x6 activity, x3 rate). Oudin's specimen of 1,500,000 gives 3 H in an hour, about 20 times slower than the 1,800,000 figure (p.1282), yet produces a month-long erythema from 15 minutes. Tousey writes that the H measurements "are to be cautiously verified" (p.1281). The radiation split of the bare salt (alpha 64, beta 24, gamma 10 percent on pp.1255-1256; alpha 90, beta 9, gamma 1 percent for the applicator on p.1260) is stated twice without reconciliation, and the "55 per cent." after glass does not follow from either [calc]. These are internal inconsistencies in a side topic, recorded.
• Forearm self-experiment [read: p.1270, Plate 15 page image p.1272]: 20 mg applicator in glass, one to five minutes on his own forearm; nothing at one minute; redness, itching and some desquamation at two to five minutes, no blistering; about two weeks' incubation. One subject, no dose in any unit, no repeat. Never properly tested as a calibration.
• 1904 radiated-saline experiment [read: pp.1282-1283]: 0.1 g radium (activity 20,000) in a sealed tube in 6 oz of saline for fourteen days; the solution was not radioactive by any test tried (no luminosity, fluorescein bag over silver coins on film for 18 hours, no Willemite glow); cat and rabbit injections without ill effect; a dog died days later "evidently in consequence of the operation". A sealed glass tube stops alpha rays (p.1255) and transfers no radium to the liquid, so a null result is what the arrangement would give; the book reports it without that comment [reader's inference].
• Also retained from the QE list: shoe-conductivity measurement of 12 May 1919 (p.360); oyster-pearl radiography (p.1141); Millener's 22 power-house cases (pp.302-303); thiosinamin, which Tousey introduced in 1894 and used in keloid and scar treatment (pp.634, 1189-1190; the Investigator found the 1896 monograph).

R9. Process and scale corrections

• Tuberculosis line (claim 1): two propositions. Tousey's statement that x-ray gives little permanent cure in pulmonary tuberculosis is consistent with later experience: Established (as the leukaemia line). X-ray as a treatment for pulmonary tuberculosis: Tested, not supported.
• Arch formula (claim 4): geometry Internally consistent (ED = 4R/sqrt 3; w = 0.36 in gives 2.22 in [calc, D claim 4]); anatomical sense at an average tooth width Plausible, untested; as a predictor of adult arch or orthodontic outcome Never properly tested.
• Web checks: the arbiter budget is 8; used 5. The fifth check settles the one medium-confidence [memory] fact behind "slightly faster than inverse square": NIST XCOM mass attenuation coefficients for dry air are 0.7779, 0.3538 and 0.2080 cm^2/g at 20, 30 and 50 keV [web: https://physics.nist.gov/PhysRefData/XrayMassCoef/ComTab/air.html]. With air at 1.205e-3 g/cm^3 [memory, high], 10 in (25.4 cm) removes 2.35, 1.08 and 0.63 percent, and 50 cm removes 4.6, 2.1 and 1.3 percent [calc]. The D3 figures (2.4, 1.1, 0.5) stand to rounding. Still [memory] and not web-checked: backscatter factor 1.3, Oudin power factor 0.1, Moyers-type analysis, membrane low-pass filter. None carries a verdict by itself.
• Not done: the second IA scan (printed pp.282-296, including unread pp.290-295) has not been read from page images; no deep reader was assigned to it in this pass.

R10. Corrections to earlier documents (QE items applied, plus two corrections to the QE itself)

1. arbiter.md D1 and open questions ("no period calibration is on record"; "Only his earlier editions or papers could say"): wrong. The book converts the author's settings at pp.1167, 1173, 1179, 1180, 1181, 759/763 and 762 (R1). Both passages are superseded in place.
2. arbiter.md D4: the quote, the page ("PDF p.585 region, lines 32139-32166" is PDF p.578) and the meaning were wrong (R3).
3. arbiter.md claim 2 and open questions ("Kienbock ... Not found"): the book cites Kienbock with Strater, Oudin and Scholtz at p.1158, and rejects Geyser's theory at p.1162 (R2).
4. arbiter.md D2 and the primer's Step 5: the "ledger does not close" conclusion omitted p.156 (R5). The primer needs the same update; claims.md claims 2, 3 and 6 need restating (the QE list stands).
5. arbiter.md claim 3: Apostoli's "over 500 cases" and Gay's 12 cases were omitted while the contrary list was kept; Tousey's p.604 caveat was omitted (R4).
6. arbiter.md claim 7: "39 m/s: internally inconsistent with his p.281 text". The page is PDF p.289, and the 39 m/s sentence is a conditional drawn from the observers' premise (R7).
7. The QE (qe.md) section 3.1 table, row p.759: "at about 90 kV" for Cabot and Dodd. p.759 names neither Dodd nor a measured voltage; 90 kV is an example of the 10 kV-per-Benoist rule, and p.763 gives 4 Benoist (R1). The QE's single figure "0.26" therefore carries a voltage uncertainty of 0.16 to 0.81 after scaling. The QE's "span of about 0.26 to 1.5" stands as the span of the unscaled conversions.
8. The QE (qe.md section 3.6, and the verdict-scale note in its section 6) describes the 39 m/s sentence as the observers' view. The premise is theirs; the inference "this would mean 39 m/s" is Tousey's conditional (R7).
9. deep read, first section, item 4 ("5 minutes unfiltered ... consistent"): the later deep read (section 6) already corrects this; the author's dose cannot be cited as "5 H" without the note in R1. facts.json key claim "500 Tousey meter-seconds is an erythema dose": the book equates it to 5.5 H, not 5 H, and the Tousey scale does not agree with the exposure table for the author's settings (R1).

R11. What would settle it (additions)

• Dose of the "Tousey dose": the Cabot and Dodd paper (Massachusetts General Hospital; cited by Tousey's footnote to Archives of the Roentgen Ray, Aug. 1911, for Cabot's absorption table, not for the calibration) to fix the voltage behind 3400 mA-s; and Tousey's own papers.
• Priority and the contact tube: the archival holdings the Investigator located, the New-York Historical Society typescript of 1911 and the University of Pennsylvania Dental Library, Hellman reprints, box 29, folder 40, "Tousey, Sinclair, 1906-1920" [web via Investigator: sites.md]. They are the nearest records that could decide the contact-tube and dental-arch priority questions.
• Condenser theory: Strater, Oudin and Scholtz, and Geyser's own paper (JAMA 1908), none read.
• Tube-UV burns (p.681): Bierry, Henri and Ranc (1911), and a measurement of UV transmission of a period x-ray tube envelope.
• High-frequency effects: the original reports of Moutier, Gidon, Guilleminot, Gay, Delherm and Laquerriere, and Apostoli's congress paper; a sham-controlled trial.

R12. Open questions changed

• Answered: the Kienbock/condenser question (R2); the existence of an in-book calibration of "(5) H" (R1); who first stated the signal-speed distinction (Tousey, p.289).
• New: which voltage lies behind the Cabot and Dodd 3400 mA-s (R1); why the book's own scales for the same settings differ by 6-9 times; whether the Doumer blood pressures are in cm Hg; what Delherm and Laquerriere's patient numbers were; whether Cabot's electro-metallurgy work preceded or followed Tousey's first-edition paragraphs.