Arbiter: Hieronymus-TG (US Patent 2,482,773, "Eloptic" element analyzer)

Role: Arbiter (Step 3). I read claims.md, facts.json, the three Step 2 reviews and sites.md, recomputed the numbers where the reviewers differed (venv numpy), and web-checked the remaining medium/low-confidence [memory] facts that verdicts rest on. Tags: [read] project document, [web] URL, [memory] with confidence, [calculation] mine.

Source gap note: none for the patent. It is the only document in the folder and is fully readable. But the patent publishes no measured data. Everything about the inventor's later quantitative work (1988 autobiography), Campbell (1956), Hite (Pathoclast study), the 1972 notes and the Huntsville/Apollo accounts comes from the Investigator's web notes; no reviewer re-read those sources, and none were downloaded to research/files/. Where a conclusion rests on those, it is flagged "via Investigator". The Fig. 3 chart numbers are read by eye from a small drawing (about 10 percent uncertain).

Interested parties: the inventor (Hieronymus; his patent, his later instruments, books and associated organisations), his circle (Tompkins & Bird, the autobiography's preface), Campbell (editor who promoted the topic; no commercial stake found), Hite (a critic with no stated commercial stake; his own characterization of the Pathoclast as a "hoax"). Wikipedia-cited critics (Gardner, Sladek, Williams, Sampson & Vaughn) are quoted as their characterization only.

Corrected claim list

Corrections to claims.md and facts.json, accepted after checking against the reviews:

1. Claim 4 (resistance method): claims.md follows the specification ("maximum resistance that still keeps the reaction"). Patent claim 6 says "until the surface frictional characteristics ... are unaffected", which reads as the opposite end point [read: p.5]. Both wordings stay in the record. Agreed by all three reviewers.
2. Claim 6 (plants): the patent says "the control plant" (singular) and treated plants "relatively healthy" [read: p.3], with no species, count, duration, light measurement or replication. "Stayed green / lacked chlorophyll" is a paraphrase of a qualitative report. The 1931 date, aluminium-lined boxes, water-pipe ground and "seed grain" are from the inventor's later telling and Tompkins & Bird, not the patent [web, via Investigator].
3. Hall US 2,115,578 (facts.json "not checked"): checked. A 1938 thermal-pneumatic "Radiant energy receiver", a detector of ordinary heat radiation [web] https://patents.google.com/patent/US2115578A/en . It is not support for element emission; it shows the only cited prior art was a conventional detector.
4. Worked example (79.2 called calcium): the number 79.2 is what the patent's own Fig. 3 line returns for a 30 degree reading: angle = 10 + 0.25 A gives A = 80 at 30 degrees [read: page image p.1, approximate; calculation]. Calcium is 40.08 [memory: high], which would read about 20 degrees. The example is conditional ("If atomic weight 79.2 is calcium"), so it is a slip in the illustration, not proven fraud. 79.2 is near selenium (78.97) and bromine (79.90) [memory: high].
5. Hite's device was a Pathoclast 1X-C, a medical rate instrument sharing the amplifier and "stick plate" detector, not the prism analyzer. His negative result bears on claim 5 only, and on claims 1-4 not at all. Anyone summarising "Hite showed it is a hoax" as a test of the patent's element claims misstates the evidence.
6. Mixtures (part of claim 3) are undefined: "composite frequency" has no rule (sum? mean? set of lines?). One scale reading cannot encode two elements and no deconvolution is given [mathematician].
7. Hydrogen is not at zero: Fig. 3's line has a built-in 10 degree offset the text never explains [mathematician, read].
8. The patent itself hedges what the radiation is: "violet to ultraviolet", also "probably electrons" [read: pp.2-3]. Claims 1 and 2 are correctly stated, but the nature of the entity is left open by the patentee himself.

9. Date conflicts left open: GB 663,978 published 2 Jan 1951 (Google Patents) vs "1952" (Hite); Advanced Sciences corporation 1967 (Florida Sunbiz no. 712451) vs "1975" (Hite, probably a Georgia registration, unverified).

10. [QE revision] The patent states a converse: "the flow of electrons along a conductor produces a radiation having characteristics of the radiation from each respective element" [read: deep read addendum (a), p.3]. One sentence, no support, not used by the apparatus. It is listed as part of claim 2 (see QE revision).

11. [QE revision] Lens variant (Fig. 4; patent claims 10-12): a lens "having a focal point for each known element", scale along the axis, no experiment, no lens data [read: deep read addendum (d), pp.4, 6]. Listed as part of claim 3.
12. [QE revision] The plant paragraph's own interpretation is that "the radiations of the elements from the light" were "apparently" conducted to the plants through the wires [read: deep read addendum (b), p.3]. The radiation source in the plant claim is therefore sunlight, not a sample at 40-90 F. Only one control box is described, and it had no apparatus at all; there is no wired-but-not-exposed control. The patent says the zone "has as yet not been fully explored" [read: p.2].
13. [QE revision] Fig. 3 is drawn only to about atomic weight 120; anything above is extrapolation (the text says the full-size chart continues to all known elements) [read: page image p.1; deep read]. Attorney: C. Earl Hovey (signature on p.1; matches autobiography p.89 [web, via Investigator]). Patent family GB 663,978 and CA 486,008 are now recorded in facts.json.

Corrected central claims (as final-verdict targets):

1. Every element emits room-temperature (40-90 F) violet-to-UV "radiations" with electrical and optical character.
2. These travel along wires like current (responding to L, C, R) and refract, focus and diffract like light; conversely, electron flow in a wire produces a radiation with each element's character (stated once, unsupported).
3. A prism exit angle rises with atomic weight (isotopes included), so scale plus chart identifies the element; mixtures give composite signatures. A lens variant puts a focal point per element along an axial scale (patent claims 10-12).
4. Quantity is read from the largest series resistance that still preserves the detector reaction (spec and claim 6 word the end point differently).
5. Detection is by the felt change of "drag" on a coated plate; the patentee says the mechanism "is not known".
6. Wired dark-room plants stayed "relatively healthy" and green; the single unwired control lacked chlorophyll; the patentee concludes that "the radiations of the elements from the light" were conducted by the wires.

Recomputed numbers (where reviewers differed or where verdicts depend)

Web checks done (budget 8; 5 used)

• Light-dependent vs dark-operative chlorophyll synthesis: LPOR is "primarily found in plants and oxygenic photosynthetic bacteria, as well as in some algae"; DPOR is "present in photosynthetic bacteria, cyanobacteria, green algae and gymnosperms" [web] https://en.wikipedia.org/wiki/Protochlorophyllide_reductase . This confirms the physicist's [memory: high] statement that conifers and some algae green in the dark, and shows species matters. Angiosperm light requirement: the Etiolation page confirms etioplasts, yellow colour and de-etiolation but does not itself name the enzyme step [web] https://en.wikipedia.org/wiki/Etiolation . I keep the angiosperm statement at textbook level ([memory: high]) with partial web support only.
• Electrovibration: the original 1954 observation used "110 V at 60 Hz", "dragging a dry finger over a conductive surface covered with a thin insulating layer", producing a "rubbery feeling" [web] https://en.wikipedia.org/wiki/Electrovibration . This upgrades two reviewers' [memory: medium] claim. The effect is established, and the 1954 geometry (finger, insulating coat over a conductor, 60 Hz mains) closely resembles the patent's detector and Hite's finding of 60 Hz-pulsed signal on the plate. This is an important and fair point for claim 5.
• Ti-50 and Cr-50 masses (above) [web].
• Hall patent content [web, via physicist] https://patents.google.com/patent/US2115578A/en (accepted).
• Not rechecked: skin depth in copper at optical frequency (physicist, medium; I computed no replacement; not load-bearing, since the circuit-frequency argument B of the mathematician makes the same point), flash-greening light dose (medium; not load-bearing), ultra-weak photon emission rates (medium; not load-bearing), the statement that patent examination does not require demonstrated operation (medium; context only). Budget remaining: 3 calls unused.

Verdict per claim

Claim 1. Room-temperature violet/UV radiations from all elements

• Agreement: all three agree the Boltzmann tail makes thermal emission in the band zero for any practical purpose (1e-52 or smaller) and that no excitation source is named. All three agree no neutral instrument test of the emission itself is on record.
• Disagreement: the mathematician and contrarian give "Never properly tested" as the headline and a refuted-by-calculation split for the thermal reading; the physicist headlines "Refuted by calculation" for the photon form. Contrarian adds a weak form ("unknown weak non-photon influence").
• Stronger argument: the physicist's, for the stated form. The patent states violet-to-UV, at 40-90 F, without special treatment [read: p.2]; that is a thermal-or-unexcited claim and the calculation excludes it. The contrarian is right that the calculation does not exclude a different, non-photon entity. That is a different claim from the one written.
• Final verdict: Split. (i) Violet/UV photon emission from unexcited elements at room temperature: Refuted by calculation. (ii) An unspecified non-photon, element-specific influence: Never properly tested (no neutral test of an operator-independent detector; not dismissed without testing, simply never taken up on record).
• What would settle it: place a sample and a blank in front of a photomultiplier or UV spectrometer in the dark, and the pickup coil with a spectrum analyser; count signal against blank over hours. A null falsifies (i) directly. For (ii), only a blind operator trial (see claim 3) can say anything.

Claim 2. Radiations travel on wires and refract like light

• Agreement: all three say a lumped broadcast-band circuit and a ray-optics prism cannot handle one entity at one frequency without a new wave theory the patent never offers (coil and 365 pF capacitors resonate near MHz; violet light is 7.5e14 Hz). The patentee's own "probably electrons" adds a third identity. The 1972 "opaque black Plexiglas prism refracts like glass" and "works with power disconnected" statements [web, Hite appendix pp.33-34, via Investigator] fit no light-like behaviour.
• Disagreement: the physicist says "Never properly tested" with the opaque-prism form internally inconsistent; the mathematician and contrarian say "Internally inconsistent".
• Stronger argument: the mathematician's. The inconsistency is in the patent's own picture (a lumped RF circuit tuned by broadcast condensers, and a ray at optical frequency), independent of any test. The physicist's reservation (no neutral test) is also true and is kept as a second part.
• Final verdict: Internally inconsistent as stated (single frequency regime impossible); the underlying experimental question (does a real prism, a dummy prism or no prism change the response under blinding?) is Never properly tested. Campbell's ink-drawn "symbolic" prism machine and the unplugged results, reported by an enthusiast and the inventor's circle, point to the response not depending on the stated components; they are kept in the record unweighed beyond that.
• What would settle it: blinded, randomised comparison of real prism, dummy prism and no prism, powered and unpowered, same operators and samples.

Claim 3. Prism angle rises with atomic weight; chart identifies the element; mixtures

• Agreement: no formula; the chart is per-instrument and calibrated on the same elements it later claims to identify (so no independent prediction until a blind unknown is identified); dispersion in the stated materials spans roughly 1 to 5 degrees against 10 to 65 degrees needed; real spectra order by electronic structure (and X-ray lines by Z via Moseley), not atomic weight; no blind identification on record; the inventor's two data points (Be about 186, Bi about 1097 of 1600 divisions) fit the 1949 chart's slope within a few percent.
• Disagreement: the physicist splits the mechanism (Refuted by calculation) from operator-based identification (Never properly tested). The mathematician headlines Never properly tested with the mechanism and the isobar extension Refuted. The contrarian says Never properly tested with the UV-prism mechanism refuted on dispersion.
• Stronger argument: all three give essentially the same split, so I adopt it. One caution I add: the "refuted by calculation" applies to the claim only if the entity is violet/UV light refracted by a glass or quartz prism, which is how the patent states it. If the entity is something else the dispersion calculation does not apply, and then the claim is not testable by optics at all and rests on operator trials. The inventor's later nuclear-content hypothesis (1941: refraction "in the order of the contents of the nucleus") is a restatement, not a mechanism [web, via Investigator].
• Final verdict: Split. (i) Prism dispersion of violet/UV light giving an atomic-weight scale: Refuted by calculation. (ii) The extension to isobar separation (Ti-50/Cr-50): Refuted by calculation against the inventor's own half-graduation resolution (0.0007 deg needed vs 0.028 deg stated, 40 times too fine) and, additionally, the free-nucleon "nuclear weight factor" orders the pair opposite to their real masses (see recomputation). (iii) Operator-based identification of elements and isotopes by the device as built: Never properly tested; reported positive by the inventor (about 35 elements run, Analyzer No. 508 peaks matching isotope counts for Ti, V, Cr, Mn, Fe, Ni) and by Campbell (operator-dependent, positive), documented only in the inventor's circle and one editor; no tables, names, blinding or date seen in the passages read. Mixture signature: Not testable as stated (no rule).
• What would settle it: blind identification trial. A helper prepares coded sealed vials (for instance 10 elements plus blanks, random order); the operator records scale readings; compare with chance. With ten samples and a tolerance of, say, 1 element in 10 per reading, 5 or more correct has probability about 1 in 1000 under chance guessing (physicist). Repeat with power off, a dummy prism, a second operator and a second instrument. A published table of angles for two instruments would also test it. No such trial appears in the record.

Claim 4. Quantity from the largest series resistance that keeps the reaction

• Agreement: all three: no units, no calibration chart published ("a chart must be prepared" [read: p.5]), endpoint set by operator perception, dependence on amplifier gain; the 1972 notes record calibration drift (normal tissue 15-20 on a 40 scale, later near 100) [web, Hite appendix, via Investigator]; claim 6 and the specification define different endpoints.
• Disagreement: the physicist adds "stronger form Never properly tested"; the contrarian "Never properly tested" too; the mathematician does not.
• Stronger argument: the same. The wording is too open for a verdict beyond not testable until units and endpoint are fixed.
• Final verdict: Not testable as stated. Once a chart and a stated endpoint are supplied it would become Never properly tested.
• What would settle it: a stated endpoint rule, a calibration against known masses of a single element, repeated by blinded operators.

Claim 5. Touch "drag" detector

• Agreement: the patentee openly says the mechanism "is not known" [read: p.4]. Electrovibration is real and now web-confirmed with parameters matching the patent's setup (finger over insulated conductor, 110 V, 60 Hz [web]). Hite's tracing on a Pathoclast found an RF signal pulsed at 60 Hz on the plate, an amplifier oscillating at the settings where the stick appears, settings moving with posture, and no correlation with samples [web, via Investigator]. Campbell and the 1972 notes say it works unpowered, which does not fit Hite's mechanism; the ideomotor effect (expectation altering finger pressure and skin moisture) fits it better (physicist and mathematician, as an inference).
• Disagreement: the physicist splits "an electrical effect changes finger drag" (Plausible, untested) from "drag encodes the element" (Never properly tested); the mathematician says Never properly tested for the prism analyzer, with "Tested, not supported" only for the Pathoclast; the contrarian splits similarly, with "tested on a related device, not supported".
• Stronger argument: the contrarian's and mathematician's restriction of "Tested, not supported" to the Pathoclast, because the test was on a different instrument. I add that Hite is one non-institutional engineer; his measurement is a specific, checkable engineering result and I weigh it as such, with his stake (none stated) noted. His unsourced statement that Hieronymus admitted wording the patent so it could not be built is not evidence and is conflicted by Campbell building from the patent alone; I keep it in the record as his assertion.
• Final verdict: Split. (i) A felt friction change on a coated plate is a real, established phenomenon (Established for electrovibration in general; Plausible, untested as the cause of the instrument's stick). (ii) The stick encodes the sample's element: Never properly tested for the patented analyzer; Tested, not supported for the Pathoclast only (Hite).
• What would settle it: same operators, same samples, mains connected vs disconnected and shielded, a scope on the plate, and blinded sample presentation. The unplugged operation reported by Campbell and the inventor is the key discriminator between an electrical (Hite) and a human-perception explanation, and it has not been tested under control.

Claim 6. Wired dark-room plants stay green

• Agreement: all three: a single unreplicated, unquantified run, one control, no species, no light measurement; angiosperms normally etiolate in the dark; the conclusion (radiation conducted by wires) does not follow even on the patent's premise without excluding light leaks, differences between boxes and ground paths; Simon Martin's 1988 cress trial (dark yellow, light green, no wires) is a baseline, not a replication [web, via Investigator]; nobody has replicated the wired version.
• Disagreement: none on verdict. The physicist adds the point that conifers and some algae green in the dark (now web-confirmed: DPOR is present in gymnosperms and green algae [web]), so a species-less report cannot be read at all. The mathematician adds the mass-balance argument that a massless radiation cannot supply atoms.
• Stronger argument: the physicist's species point, since it shows the result may need no special explanation.
• Final verdict: Never properly tested (not "tested, not supported": there is no failed replication; not dismissed after a test, simply uncontrolled and unrepeated). The cheapest claim here to test.
• What would settle it: fixed species (an angiosperm such as cress or bean), photometer-verified light-tight room, many boxes randomly assigned to wired or unwired, a grounded-but-unwired group, chlorophyll absorbance measured blind.

Summary of final verdicts

What would settle it (the single most decisive items)

1. Blind, randomised, sealed-sample identification trial with a replica built from Fig. 1 (Campbell showed it can be built from the patent alone), including power-off, dummy-prism and no-prism arms. This one trial bears on claims 1 (ii), 2, 3 (iii) and 5 (ii).
2. A photomultiplier / spectrum analyser run at the pickup coil, sample versus blank (claim 1 (i), claim 2).
3. A published angle-vs-element table from two instruments, or the inventor's 1988 data tables from the autobiography (deep read of pp.228-267; PDF 59 MB, not downloaded) (claim 3).
4. A controlled wired/unwired/grounded plant trial (claim 6).

Where the record goes dark: absence-of-evidence analysis

Observation: Radiation Laboratories, Inc. (Kansas City), which sold the analyzers and the medical instruments, stops appearing about 1950. The Experimental Special Data Bulletin (monthly, Jan 1948 to at least Jul 1949) exists, is wanted by the USPA, and has not been located. No published data set, conference paper or neutral test follows the 1949 patent. The 1944 parent application (Serial No. 555,553) was abandoned. The Huntsville "300-copy report" to von Braun has not been found. The sole experimental critique (Hite) came 45 or more years later and concerns a different instrument. Between 1950 and the 1956 Campbell articles, the record is thin.

Competing explanations, with evidence for and against, and what would decide:

1. Commercial failure. For: the inventor's own reason (a), bank rule changes ended discounting of out-of-state doctors' notes [web, autobiography p.90, via Investigator]; he sold the laboratory and stopped building instruments in 1950; the business model depended on practitioners' credit. Against: none found beyond the absence of records; the inventor continued selling and writing for decades (Advanced Sciences R&D Corp. from 1967 to 1996, Sunbiz no. 712451). Decider: Missouri Secretary of State corporate filings, Kansas City newspapers, creditor and bank records.
2. Regulatory pressure. For: the inventor's reason (b), AMA-driven FDA pressure on users [web, autobiography p.90]; this is his account only, in one published source; other radionics devices were prosecuted in the same era (general background, [memory: medium], not checked). Against: an FDA/court query found nothing naming Hieronymus [web, Investigator]; Hite says he "remained uncensored" [web, Hite p.13]. Decider: FDA Notices of Judgment 1948-1955, Post Office fraud-order records, user prosecutions naming Radiation Laboratories instruments.
3. Absorption into another firm under a new name. For: Hite names Pathometric Laboratories and "Radionics Incorporated" as makers of related instruments [web, via Investigator]; the inventor rebuilt a Pathoclast in the late 1930s. Against: no ownership or acquisition document has been seen. Decider: USPTO assignment records for US 2,482,773 (two assignments to Radiation Laboratories, Inc. are reported) and Missouri filings.
4. Trade secrecy among customers or practice-based circulation (instruments leased or sold to practitioners rather than published). For: the 1972 notes and the autobiography treat calibration and operator skill as personal; the instruments were "Hieronymus-type" and still sold later (Kelly Research Technology, 2018) [web, via Investigator]. Against: none found. Decider: the Data Bulletin and customer records.
5. Suppression of a successful test. Evidence for: only the inventor's FDA/AMA narrative, one source. Evidence against: Hite's "uncensored"; no recorded seizure; no recorded test to suppress. The record does not show a test being blocked; it shows no test being run. The Investigator's FDA query was negative, but that is one query; silence there is not contradiction.

Verdict on the anomaly: not resolved. Explanation 1 and 4 are best supported by the available (inventor-side) evidence; 2 rests on the inventor's account alone and has not been checked against neutral records; 3 and 5 have no positive evidence. A genuinely independent record (court, FDA, state filings, newspapers) is the decider, and it has not been searched past a first query; the USPTO, Espacenet, Missouri and Georgia registers, newspapers, IAPSOP/Borderland and WorldCat are queued in sites.md as "not reached (time limit)".

Investigator findings retained in the record (whatever the verdict)

• Patent family GB 663,978 (filed 18 Jul 1949, published 2 Jan 1951 per Google Patents) and CA 486,008 (26 Aug 1952); a grant is not evidence of operability.
• Campbell (Astounding, June and Aug 1956): one documented independent build from the patent alone; positive; operator-dependent (best with young girls, worst with technically trained men); works unplugged; fails with a defective tube. Operator dependence and unplugged operation are the observations an ideomotor explanation predicts; they are also what the patent's electrical model does not predict. Both are retained.
• Autobiography numbers (1600-division quadrant; Be about 186, Bi about 1097; about 35 elements; Analyzer No. 508 peaks); test plan with half-graduation agreement; whether performed is not stated in the passages read.
• 1972 "Random Notes" (Hite appendix): 60 Hz AC on plates by design; odd number of stages; Bakelite insulates; Plexiglas transmits; light "shorts"; opaque black Plexiglas prism refracts; works with power disconnected; calibration drift.
• Huntsville duplication ("Rho Sigma"/Rolf Schaffranke, about 20 engineers, 300-copy report to von Braun), Apollo 8 monitoring (signal lost in the Moon's shadow while radio worked), submarine test off Key West: single-circle accounts with no data. Documented in two published sources from the same circle (1973 book, 1988 preface); report not located. Checkable in principle against mission timelines; no log seen.
• Simon Martin's cress trial (baseline only).
• 13 citing patents (e.g. US 2,930,977; 2,972,251; 4,260,933; WO 2013/155580), unread.
• Biography: KCP&L engineer 1919 to about 1946; WWI signal service; Hite could not confirm an engineering degree. A practising electrical engineer would know how an oscillating amplifier behaves, which neither supports nor refutes the effect.
• Lineage: Abrams, Pathoclast/Pathometric, Campbell's symbolic machine, modern radionics: context, not support. Mainstream critics (Gardner 1957, Sladek 1973, Williams 2000, Sampson & Vaughn 2000) class it as pseudoscience; their characterization, not experiments on the prism analyzer.

Open questions

1. Does the "stick" appear with the power disconnected, and does it correlate with the sample, when operators are blinded? (The pivotal untested observation; sources conflict: Campbell and the inventor report yes, Hite's mechanism needs power.)
2. What do the inventor's 1988 data tables actually show, and who operated the instruments? (A deep read of autobiography pp.228-267 is the best next step.)
3. Is the 1949 chart's linearity in atomic weight a property of the operator's responses, of the instrument geometry, or simply of the scale as defined? The Be/Bi match to the chart (0.256 vs 0.25 deg per unit) is consistent but not independent.
4. What was the 1931 plant setup, and does any wired-plant replication exist?
5. Did Radiation Laboratories, Inc. end for commercial or regulatory reasons, and where is the Experimental Special Data Bulletin?
6. Did the Huntsville duplication report exist, and what did it contain?
7. Does the "nuclear weight factor" curve in the autobiography track free-nucleon masses or real isobar masses (the two order Ti-50 and Cr-50 oppositely)?
8. Sources the project still lacks: the autobiography, JBR May-June 1988, the Hite PDF, Astounding June and Aug 1956 (all identified by the Investigator, none downloaded).

Source honesty and limits

• Verdicts on claims 1 and 3 (mechanism) rest on [calculation] using physical constants and the fused-silica Sellmeier formula, which three analysts reproduced independently; the Sellmeier coefficients themselves are [memory: high] and could not be fetched from refractiveindex.info.
• All inventor, Campbell, Hite, Tompkins & Bird and Huntsville material is [web] via the Investigator, not re-read by reviewers or me. Where a statement is only that, the verdict does not depend on it except the "operator dependence / works unplugged" observations in claims 2 and 5, which carry weight as reported observations, not as established facts.
• The Fig. 3 slope and offset are read from a small drawing by Step 1 and the mathematician; about 10 percent uncertain.
• No reviewer or I read any replication of the plant, prism or friction claims; none is known to me. I do not state that any replication or test occurred or failed beyond what the Investigator's sources report (Campbell positive and non-blind; Hite negative on a Pathoclast).

QE revision

QE (qe.md, 2026-10-04, Opus) returned MINOR: no headline verdict changes. I read qe.md, the deep read addendum (a)-(d) on the patent (deep/US2482773-f82c2d.md, new), the updated claims.md and facts.json, and checked the retention items against the Investigator's notes in research/. The deep reads QE asked for on the autobiography, Astounding 1956, the 1957 issue and the Hite PDF have not been done: none is in research/files/ (owner approval to download was never given). Everything below about those sources is still [web, via Investigator], read by keyword search only. Only the patent deep read (task 1) is new.

What the new patent deep read adds, and how I weigh it

1. Converse statement [read: p.3]. Verbatim: "Conversely, the flow of electrons along a conductor produces a radiation having characteristics of the radiation from each respective element." The patent gives no experiment, frequency or element-selection rule, and the apparatus only receives. It also does a hidden job: electrode 32, fed by a wire, is treated as a re-radiator into the prism, so the prism sees a re-radiation and not the original emission from the sample. I add it to claim 2. Verdict on the converse alone: Not testable as stated, because no criterion says which element a given current represents. Without such a criterion any current in any wire is "a radiation of some element". This does not change the claim 2 verdict (Internally inconsistent; the operator question Never properly tested). It sharpens the inconsistency: the identification then rests on the operator's tuning and touch.
2. Plants and "from the light" [read: p.3]. The patent's own interpretation is that sunlight carries element radiations and the wires deliver them. This matters for how I stated claims 1 and 6. - The Boltzmann calculation in claim 1 excludes thermal violet/UV emission at 40-90 F. It does not address a claim that sunlight already contains such radiation. Sunlight does contain violet and UV photons; no one disputes that. The disputed step is that those photons (or an "element radiation" riding on them) are conducted by a wire into a dark room and feed the plants with elements. That step is untested, and nothing in the claim says what the elements are or how they reach the plants. - The sentence "Such of the known elements as are required to feed growing plants have been transmitted to the plants through metallic conductors" is a claim of transport of matter's character without matter. The mathematician's mass-balance point (a radiation cannot supply atoms) applies to it as written. - What the control lacks: there is one control and it had no apparatus. A wired-but-unexposed box is absent, so a wire, plate or leak-of-light effect cannot be separated from the claimed one. The patentee himself writes "apparently". Verdict stays Never properly tested. - A tension I now record, from the Investigator's notes on the autobiography, not from the patent: the inventor later says strong light dissipates eloptic energy and that X-rays and UV destroy it [web, autobiography pp.227-228, via Investigator, keyword read]. The patent places the radiation in the violet-to-UV zone and the plant claim has sunlight as the source. The two statements are not reconciled in the sources I have. I keep both and do not resolve it.
3. "Zone not yet fully explored" [read: p.2]. This is the patentee's stated reason why no spectrum of these emissions exists. He is an interested party and the patent supplies no wavelength, spectrum or instrument for the zone. For context, I note from [memory: high] that UV spectroscopy of the violet and near-UV was well developed by 1946; this is background and does not decide any claim. It also means the claim of a "new" optical zone is a claim about a zone that was already instrumented, and in that zone the thermal tail argument (claim 1) applies fully.
4. Lens variant (Fig. 4; patent claims 10-12) [read: pp.4, 6; page image p.1]. The spec spends about ten lines on it, says it works "substantially" as the prism does, and reports no lens experiment, no focal lengths, no lens material and no units on the 0-80 axial scale. Dispersion check [calculation, same Sellmeier formula as in the table above]: the focal length of a thin lens scales as 1/(n-1). For fused silica, n-1 = 0.4701 at 400 nm and 0.5505 at 200 nm, so the focal length falls by 0.4701/0.5505 = 0.854, about 15 percent, across the whole violet-to-UV band. A scale with a distinct focal point for every element from hydrogen to uranium would therefore have to fit within about 15 percent of one focal length. In ordinary glass the band is cut off near 320 nm [memory: high], so the available span is smaller. This is the same conclusion as for the prism: the band gives a small spread against what an element scale needs. I fold it into the claim 3 dispersion verdict (Refuted by calculation if the entity is violet/UV light in a glass or quartz lens). The lens does not give independent evidence; it is a second route to the same assertion. No ordering law (such as atomic weight) is stated for the lens. The 15 percent figure is mine and does not depend on QE's earlier approximate "about 15%" except in agreeing with it.
5. Smaller items. Attorney C. Earl Hovey; Fig. 3 stops at about A = 120, with the range H to A = 115 needing about 29-32 deg against 3-6 deg available [read: page image p.1; calculation, consistent with QE section 4]. The extrapolation to Bi (about 61.7 deg) comes from the inventor's later numbers and the same linear ruler, as already noted in the recomputation table. Where I wrote "the chart needs about 10 to 65 degrees", the drawn part supports only the lower range and the rest is extrapolated. The factor of 5 to 10 remains from the drawn range alone, and 10 to 20 with the extrapolation.

Verdict changes

Why 3 (ii)(b) is conditional. QE is right that I put two numbers together from different passages. The 0.028 deg target comes from the half-graduation test plan on the 1600-division quadrant (autobiography pp.233, 237-238) and the Ti/Cr factors from Analyzer No. 508's "prism dial setting" plot (pp.266-267) [web, via Investigator]. I assumed the dial uses the quadrant's scale. I have not seen that this is so, and I have not seen whether the inventor says the instrument separates the pair by angle or only that the computed factor does. The Investigator's own note says "plotted factor against prism dial setting, reporting a smooth curve". So: the resolution comparison shows that if the quadrant's stated resolution applies, then a 0.0007 deg separation cannot be read. It shows nothing more. The factor-ordering result does not depend on the book. The inventor's smooth curve might track the factor or the real masses; the passages read do not say.

What would settle 3 (ii)(b): the autobiography pp.227-242 and 266-267 deep read (QE task 2), specifically which scale No. 508 used and whether he claims angular separation of Ti-50 and Cr-50.

Investigator findings retained, added by this revision

These were in sites.md and research/ notes but missing from this file or the primer. They stay in the record whatever the verdict (all [web, via Investigator], with the source named in the research note; none re-read by me):

• Campbell, Astounding Aug 1956, specifics: success varies by person, highest for girls of high-school age and younger, lowest for technically trained men "sure what can happen"; unreliable for Campbell, reliable for his daughters and friends; works as well unplugged; and will not work if a tube is burned out or defective, plugged in or not [research/2026-10-04-archive-org.md, section 3]. The defective-tube observation is the one item that cuts against a purely ideomotor reading, since an expectation effect should not care about a tube. It also sits oddly beside "works unplugged", since a tube needs power to act as an amplifier. I record that these two statements are not reconciled in the source I have. They are the second most important discriminator after the power-off test in the open question below, and a controlled build would test both.
• Campbell's "symbolic" machine: an India-ink circuit diagram with a prism symbol on a vernier dial, described in Campbell's letters reproduced in the autobiography from about p.123 [web, via Investigator]. Hieronymus's comment (pp.123-124, also quoted on Wikipedia) is that Campbell's emphasis on the "magic" side set his work back "at least a hundred years". The date of the symbolic version and whether Campbell reported that it worked are not established in the research notes; QE is right that the primer's attribution to "Campbell (1956)" is not supported by the June and August 1956 notes. The inventor's objection shows his own position: that the device is physical, not symbolic. It is the inventor's characterization.
• Production mix [web, Hite, via Investigator]: about 200 desktop medical units at $1,500, about 10 optical substance-analysis (prism) units, 12 or more consoles at $2,000. The patented prism analyzer was a minor product line. Most instruments sold were medical rate instruments, which bears on how far Hite's Pathoclast result reaches (already limited in correction 5 above). This is Hite's figure, and he is a critic.
• Early field use [web, autobiography p.88; Tompkins & Bird p.333, via Investigator]: Abrams instruments, mineral-finding with Stanley Rogers, and a claim that the energy of buried silver, gold and other minerals leaves the ground at an angle that varies with the moon cycle. A testable claim, no data in the passages read.
• Origin: first "energy Analyzer" built 21 Nov 1941 with Dr. Grace Edwards, D.C., of Omaha present [web, autobiography p.89, via Investigator]; this dates the analyzer three years before the 1944 parent application.
• Numerical claims of the inventor [web, autobiography pp.173, 227-228, 267, via Investigator]: analyzers erratic for hours after atomic-bomb tests; radioactive isotopes "radiate" from inches to two feet or more; per-isotope "radiation distances" (vanadium 18.5, 5.5, 15.5 in.; bismuth 13, 24.5, 22, 26 in.); the listed properties of eloptic energy, including that X-rays and UV destroy it and strong light dissipates it. Testable in principle; no tables, names or blinding in the passages read.
• Apollo claims beyond what I listed earlier [web, autobiography pp.338-340 per the research note, via Investigator]: a "lethal" belt around the Moon; the analyzer went dead in the Moon's shadow while radio worked, and kept working behind the Moon during the radio blackout when sunlit. Checkable against mission timelines. No log seen.
• Records leads [web, via Investigator]: an AIEE membership list naming "Hieronymus, T. G., pres., Radiation Laboratories, Inc., Kansas City, Mo." (IEEE Xplore PDF; fetch blocked, list date unknown); The Truth about Radionics (1947, anonymous; the attribution to Hieronymus is the editor's "highly probable" guess, unverified); The Eloptic Directory (Sept 1976); Goodavage's Analog interview (Jan 1977, not found). Hite's remark that Hieronymus was a senior IRE member or AIEE fellow, and that the "Dr." was probably honorary, is unverified and is Hite's statement.

Other QE points accepted

• The physicist's claim 1 heading "Tested, not supported in the strong form" is superseded. No test of the emission is on record; the physicist's own split says Refuted by calculation. That is the label used above.
• QE's process notes on the primer (symbolic-machine attribution, the "patent's detector fed from 60 Hz mains" wording, biography tags) concern primer.md, not this file. One correction applies here too: the 60 Hz plate supply is from Hite's Pathoclast tracing and the inventor's 1972 notes, not the patent. The patent does not say how the detector is powered. My claim 5 text said Electrovibration's 1954 parameters "closely resemble the patent's detector"; the resemblance is to Hite's Pathoclast and the 1972 notes. Read that sentence as limited to the coated plate, the finger and the insulating coating, which the patent does describe.
• The Sellmeier coefficients are still [memory: high] and unfetched. I did not use a web call on them. Budget used: 5 of 8 before; none added now.

Open questions, updated

Add to the list above: 9. Does a wired-but-unexposed plant control behave like the wired exposed ones? 10. Is the defective-tube observation reproducible, and does it hold with the supply disconnected? 11. Which scale does Analyzer No. 508 use, and does the inventor claim angular separation of isobars? 12. When did Campbell report the symbolic machine, and what result did he claim?