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

Reviewer role: PhD physicist (Step 2). Tags: [read] from the project documents (patent text, deep read), [web] with URL, [memory] with confidence. Source base: one patent (fully readable, no source gap) plus the Investigator's web findings, which are mostly the inventor's own later statements and those of his circle (interested parties). The patent publishes no measured data.

Corrections to Step 1

1. facts.json says the content of the cited prior art (Hall, US 2,115,578) was not checked. Checked: it is a 1938 "Radiant energy receiver", a thermal-pneumatic detector (blackened absorber heats gas, a diaphragm acts as a capacitor plate, an amplifier reads it), meant for infrared/radiant heat from aircraft, icebergs, stars [web] https://patents.google.com/patent/US2115578A/en . It has nothing to do with element identification. It shows the examiner's only citation was a conventional radiation detector, so the patent office did not test the central physical claim (patent examination does not require demonstration of operability in the way a laboratory test would; [memory] medium).
2. The deep read says the 79.2 vs. calcium example is "internally inconsistent". Agreed, but note the example is hedged ("If atomic weight 79.2 is calcium") and the chart line 45 at 30 degrees does give about 80 [read: deep read, page image p.1]. The slip is in the patent's illustration, not a proven numerical contradiction in the chart. Calcium is about 40.08 [memory: high].
3. The Step 1 deep read treats "detector 12" and "coil 12" as separate objects by correct inference from Fig. 1 (detector is 72). I confirm; the Investigator's GB 663,978 abstract [web] https://patents.google.com/patent/GB663978A/en agrees. No change.
4. Claims.md claim 6 says the plants "stayed green" and the control "lacked chlorophyll". The patent says "the control plant" (singular) and "relatively healthy" for the treated plants [read: p.3]. Claim 6 should be read as a qualitative, unquantified report. Tompkins & Bird add aluminium-lined boxes, a water-pipe ground and "seed grain" [web, secondary] ; these are details from the inventor's later telling and are not in the patent.
5. The patent says the "radiations" were ground to a "zone from violet up into ultra-violet" but also that they travel as "probably electrons" on wires [read: p.2-3]. Claims.md claim 1/2 are correct as written; I flag that the patent itself hedges the nature of the radiation.
6. Hite's device was a Pathoclast (medical rate instrument), not the prism analyzer (Investigator already says so). I keep Hite as relevant only to the detector and amplifier (claim 5), not as a test of claims 1-3.

Mainstream view, stated fairly

Atoms and molecules do emit and absorb electromagnetic radiation, and spectroscopy identifies elements by it. That is real, established, and exactly the kind of idea the patent borrows. But the established mechanism has features this patent contradicts:

• Line emission needs excitation (flame, arc, spark, laser, X-ray beam). A cold sample at 40-90 F (278-305 K) emits thermal radiation that follows the Planck law, peaking in the infrared near 10 micrometres. Number check (worked): the Planck factor at 400 nm and 300 K is exp(-hc/(lambda k T)) = exp(-120) = 8e-53; at 300 nm exp(-160) = 3e-70; at 200 nm exp(-240) = 7e-105 [calculation, constants from memory: high]. A violet/UV photon is 3.1 to 6.2 eV; room-temperature thermal energy is about 0.026 eV. Thermal emission in the claimed band is therefore zero for any practical purpose. Atoms in the ground state do not emit; they only absorb.
• Where elements' characteristic emission is observed unprompted, it is X-ray or gamma emission from radioactive isotopes, not a universal property of all elements.
• A prism separates wavelengths by dispersion, n(lambda). Characteristic lines do not follow atomic weight monotonically (hydrogen is a lone Balmer/Lyman series; sodium D lines sit at 589 nm, calcium's strong line at 422.7 nm and K, H lines at 393/397 nm, iron has thousands of lines). No known spectral property increases with atomic weight "in the same order" across 90 elements.
• Chemistry is governed by the electrons (atomic number Z); atomic weight (nuclear mass) shows up only in tiny isotope shifts (parts per 10^5 or smaller in wavelength). The patent's ordering by atomic weight, with isotopes spread by weight, corresponds to no atomic-spectroscopy quantity [memory: high].
• Electrons conducted in a wire follow circuit theory, not optics; "radiation" that is both an electrical current in copper and a ray refracted in glass is not a single known thing. A wire carries a light-frequency EM field poorly (skin depth in copper at 1 PHz is about 2 nm, [memory: medium]) and a metal is opaque to it.
• Feeling "drag" on a coated plate is a tactile judgement. Known confounders are the ideomotor effect and unconscious cues (the same family as dowsing and the Ouija board, which Campbell himself used as a simile [web: Investigator, Astounding June 1956]), and electro-tactile or electrostatic stickiness from an oscillating amplifier (Hite's explanation; he is a critic, see below).

Worked check: could a prism of the kind described give the claimed element-by-angle scale?

Use the patent's geometry as read: "angle of incidence ... of the order of 5.5 deg" to the face [read: p.3]. Take that as a ray striking the face at 5.5 deg from the surface (84.5 deg from the normal), the most generous reading (largest angular sensitivity). Fused silica dispersion (Malitson Sellmeier formula, [memory: high]), refraction angle from sin(theta_t) = sin(84.5 deg)/n:

Result: the entire band from visible down to the silica cutoff spans about 4 degrees; the violet-to-UV zone the patent claims (about 400 to 200 nm) spans 2.7 deg. The patent's scale runs from roughly 0 to 80 degrees and Fig. 3's line runs about 10 to 40 deg over atomic weight 10 to 120 [read: page image p.1, approximate]. The dispersion of any transparent material cannot spread 90 elements over tens of degrees within the claimed band; ordinary glass is also opaque below about 320 nm [memory: high]. The mismatch is a factor of order 10. So, if the radiation really were violet/UV light refracted by a glass/quartz prism, the angle-vs-element scale is impossible by optics alone. The only way out is to say the "radiation" is not ordinary light, which then removes the "optical" claim and leaves nothing measurable to test with a prism except the operator's reaction.

Cross-check on the inventor's later data: on his 1600-division quadrant, beryllium isotope about 186 and bismuth about 1097 [web, inventor's autobiography pp.237-238, via Investigator]. In degrees: 186/1600 x 90 = 10.5 deg, 1097/1600 x 90 = 61.7 deg. The patent's chart line 45 rises about 0.27 deg per unit of atomic weight from about 10 deg at A = 10; extrapolated to Bi (A = 209) it gives about 64 deg [calculation]. So the inventor's two sets of statements (1949 chart, 1988 narrative) are mutually consistent to a few degrees. This is a point in his favour for internal consistency over four decades, but it is not independent evidence: both come from the same person and both are linear in atomic weight (a straight ruler), which is what a defined scale would give whatever the underlying cause.

Claim-by-claim review

Claim 1. Every element emits room-temperature "radiations" (electrical and optical, violet to UV)

a. Mathematics: the patent has no equations. Verbally it is inconsistent with itself in one respect: "probably electrons" [read: p.3] versus "optical" radiation in the violet-UV zone [read: p.2]. These differ by orders of magnitude in physical properties. b. Experiments: contradicts the Planck law and atomic spectroscopy as shown above. No known experiment shows universal cold-emission of violet/UV photons. Related real phenomena are exoelectron emission, chemiluminescence, radioactive decay, and ultra-weak photon emission by living tissue (the latter measured with photomultipliers at tens of photons per cm2 per second, [memory: medium]). None is universal, element-specific and room-temperature violet/UV from inert samples. c. Tested: the patent reports no test. Campbell (1956) reports "positive results" building from the patent [web, Investigator: Astounding June 1956]; he also reported that it works equally unplugged and depends on the operator [web, Aug 1956]. Hite tested a different (Pathoclast) instrument and found 60 Hz-pulsed RF and no correlation with samples [web: https://leehite.org/documents/The%20Hieronymus%20Hoax.pdf]. No controlled, blinded test of the analyzer itself has been found. The inventor describes Huntsville engineers who "duplicated" it with success and a 300-copy report [web, autobiography preface; single published source from within the circle, report not located]. That evidence is stated here in full and is unverified rather than refuted. d. If true: a new, universal, element-specific radiation field at 300 K would require violating thermal radiation theory or adding a source of energy (the energy source is unstated), which also threatens conservation of energy unless a new reservoir exists. Quantum mechanics would need ground-state atoms to emit continuously. Verdict: Tested, not supported in the strong form (as a physical emission, calculation shows thermal excitation is zero); Never properly tested in the sense of a neutral, controlled test of the actual analyzer (the sceptic's test, Hite, was of a related device). Split verdict: (i) "radiation of violet/UV photons from cold elements": Refuted by calculation (Planck factor 1e-52 or smaller); (ii) "some unknown non-electromagnetic influence that the operator senses": Never properly tested.

Claim 2. The radiations travel along wires, respond to L, C, R and can be refracted, focused, diffracted

a. Internal consistency: a quantity that behaves as current in copper and as a ray in glass needs a single theory; none is offered. The 1972 notes add that Bakelite insulates, Plexiglas transmits, light "shorts" the energy, and an opaque black Plexiglas prism refracts like glass [web, Hite appendix pp.33-34]. Refraction through an opaque body is not a property of light. These rules are ad hoc; each is a separate, nonexclusive claim. b. Experiments: ordinary EM signals can be sent down wires and refracted, but at very different frequencies and by different structures (coax vs. lenses). Nothing in measured EM propagation unites them. Not consistent with the skin-depth argument above. c. Tested: an electronic version is testable with a spectrum analyser or photomultiplier at the claimed electrodes. No record of such a test. The patent's conductor is "magnet wire" or a coil near a sample [read: p.3]. d. If true: a new carrier able to cross conductors and dielectrics selectively would need to be added to Maxwell's equations, with its own coupling constants. The inventor's later description (stored, amplified, shielded, destroyed by X-rays and UV, not inverse-square) [web, autobiography pp.173, 227-228] also gives a signal that does not fall with distance as radiation does, which contradicts energy conservation for a radiating source (no inverse-square falloff means more power than the source supplies at large range), unless the signal carries information only and not energy. Verdict: Never properly tested (as stated), with the "refracted through opaque black Plexiglas" form being Internally inconsistent with a light-like nature. I record the inventor's claim as stated; it has not been put to a neutral test.

Claim 3. Each element leaves a prism at a characteristic angle increasing with atomic weight; mixtures give composite signatures

a. Mathematics: no formula, only an empirical chart made by the maker per instrument [read: p.4]. The worked example (scale 30 deg gives A = 79.2, "if atomic weight 79.2 is calcium") has the wrong element for that weight, though the conditional wording hides it. The quadrant geometry, "30 to 60 degrees" prism angles and the 5.5 degree incidence are ambiguous. The angle-versus-weight claim is internally consistent with the 1988 data as shown above. b. Experiments: contradicts the atomic spectroscopy picture, where line wavelengths depend on Z and shell structure, not on atomic weight; and the dispersion table above shows the angular range is impossible by prism physics. The inventor's later nuclear-weight story (protons x 1.00758 + neutrons x 1.00894, separating Ti-50 from Cr-50 by a difference of 0.003 mass units out of 50) [web, autobiography pp.266-267] would need a resolution of 6e-5, far beyond what a hand-operated prism can do and below the thermal stability of any tuned circuit built from 1940s condensers (hand-estimate; [memory: medium]). Real chemical identification by weight is done with mass spectrometry, which needs ionisation and vacuum. c. Tested: the inventor says he ran about 35 elements and planned inter-operator and inter-instrument tests (agreement within half a division) [web, autobiography p.233]; whether these tests were performed or what they showed is not stated in the passages the Investigator read. Campbell's independent build gives operator-dependent results [web]. No blinded sample-identification trial found. The test that would settle it is simple and cheap (see below), which sharpens the point that none is on record. d. If true: atomic emission would have to be sorted by nuclear mass rather than by electronic structure, contradicting the Bohr/Schrödinger picture and isotope-shift data. Verdict: Split. Dispersion-based mechanism: Refuted by calculation. Operator-based identification (what the device actually does if it works): Never properly tested (no neutral blind trial on record; no evidence it was dismissed without testing by an agency, though the FDA/AMA climate the inventor cites is documented only in his own account).

Claim 4. Intensity or quantity is measured by the largest series resistance that still keeps the detector reaction

a. Mathematics: a resistance read-out in ohms with a decade box (1 Mohm in 100 kohm steps etc.) is a well-defined measurement of a resistance, but the patent gives no calibration of ohms to quantity ("a chart must be prepared" [read: p.5]) and no units. Claim 6 wording differs from the spec (maximum resistance keeping the reaction vs. "unaffected"). The inventor's 1972 notes say the calibration drifted over the decades and the operator "adds a factor" [web, Hite appendix]. b. Experiments: no known quantitative relation between a sample's mass and a shunt resistance threshold exists. A threshold set by the operator's perception depends on the amplifier gain, which in an oscillating amplifier is highly nonlinear. c. Tested: not tested in the patent. The 1972 notes are the only description of calibration, and they describe drift. d. If true: a conserved quantity would need to be conveyed from sample to detector without power; no mechanism. Verdict: Not testable as stated (no units, calibration or objective endpoint), with a testable stronger form: Never properly tested.

Claim 5. A coated plate whose surface "drag" under the fingers changes with the radiation (mechanism unknown)

a. The patent admits "What actually happens ... is not known" [read: p.4]. That is honest and not an inconsistency. b. Experiments: surface friction does change with electric field and current: electrovibration/electroadhesion is a known effect in modern touch-screen haptics, where an applied AC voltage of order 100 V at tens to hundreds of Hz changes the perceived friction of a finger on an insulating-coated plate [memory: high]. A 60 Hz mains-driven amplifier whose output goes straight to the plate (Hite's tracing, [web, Hite pp.6-14]) is a plausible source of such a "stick", and an oscillating tube amplifier (odd number of stages, feedback) would produce settings where the stick appears, affected by the operator's posture as he reports. This is a real mainstream explanation, from a critic, and it fits Hite's device. It does not by itself explain Campbell's and the inventor's observation that the stick works with the power disconnected [web: Astounding Aug 1956; Hite appendix], or Campbell's observation that it fails when a tube is defective whether plugged in or not. These latter points are recorded and unexplained; unplugged operation fits the ideomotor effect (operator expectation changing finger pressure and skin moisture) better than any electrical mechanism, but this is an inference, not a demonstrated result. c. Tested: Hite (critic, no stated commercial stake) tested a Pathoclast and found the stick settings shifted with his posture and bore no relation to samples. Campbell, a sympathetic builder, found operator dependence (best with young girls, worst with trained men). Neither is a controlled blinded trial of the analyzer. d. If true: that an element-specific signal modulates surface friction would require a coupling of an unknown field to the skin or coating. Minimal change to physics if the effect is electrotactile; major if element-specific. Verdict: Plausible, untested for "an electrical effect changes finger drag"; Never properly tested for "the drag encodes the sample's element".

Claim 6. Dark-room plants wired to sunlit outdoor electrodes stay green; the unwired control lacks chlorophyll

a. No mathematics. The reasoning (green plants therefore element radiations conducted) is a non sequitur even on the patent's own premise: it needs the alternatives excluded (leaks of light into the room, a different substrate, different seed lots, watering). b. Established botany: in angiosperms (flowering plants, including grains and cress), the last step of chlorophyll synthesis, protochlorophyllide to chlorophyllide, requires light (the enzyme is light-dependent protochlorophyllide oxidoreductase). Seedlings in true darkness are etiolated (yellow-white) [memory: high; Wikipedia's Etiolation page confirms that chloroplasts not exposed to light are called etioplasts, https://en.wikipedia.org/wiki/Etiolation , but the fetched text did not itself explain the POR step]. Gymnosperm (conifer) seedlings and some algae and ferns do green in the dark [memory: high], so species matters, and the patent gives no species. Greening needs only small light doses: a flash of a fraction of a second at about lux levels initiates the conversion [memory: medium]. A "darkened basement" with no photometer and no species leaves many ordinary ways for the treated boxes to receive light or differ from the control. c. Tested: one experiment, by the inventor, undated in the patent (1931 in later accounts [web, autobiography p.88]), no counts, no measurement, one control box, no replication, no blinding [read: p.3]. Tompkins & Bird (1973) retell it with more detail (aluminium-lined boxes, ground to a water pipe, separate copper wires, seed grain; [web, secondary]). Simon Martin's cress trial found dark-grown seedlings yellow and light-grown green [web, JBR 1988], consistent with textbook botany but with no wires, so not a replication. I found no replication with wires. d. If true: plants would have to receive light-borne nutrition or signalling through copper wires, which would amount to a new photosynthetic channel; it would also require explaining chlorophyll synthesis without photons at the protochlorophyllide step. Verdict: Never properly tested. The result is plausible for several mundane reasons but the report is too thin to evaluate; the apparatus is simple and the test is easy to run with a proper design (see below). Not dismissed by any published failed replication, so "tested, not supported" does not apply.

Investigator's findings I weighed (retained whatever the verdict)

• Patent family GB 663,978, CA 486,008: confirms the figure numerals; supports that the claim set survived in other jurisdictions (the grant is not evidence of operability) [web].
• Autobiography data (1600-division quadrant; Be about 186, Bi about 1097; 35 elements run; Analyzer No. 508 peak counts matching isotope counts): the most specific quantitative claim by the inventor. My check shows it is linear and consistent with the 1949 chart; it is unreplicated and unblinded as far as read, and the passages do not say who operated. Kept as the best available inventor-side numbers.
• 1972 "Random Notes": operator intent essential; works unpowered; odd number of stages; calibration drift. These are the inventor's characterization as noted by another person, whose name is not given on the pages read.
• Campbell 1956: one documented independent build from the patent with positive operator-dependent results, and the unplugged result. Campbell was an editor with a strong interest in the topic (an interested party in the sense of promoting the idea) but with no commercial stake in the device found.
• Huntsville duplication, Apollo 8 monitoring, submarine test: single-circle accounts, no data. Not evaluated beyond: the Apollo claim of loss of signal when the capsule was in the Moon's shadow while radio worked is, if the log exists, checkable against mission time lines; no record seen.
• Hite 2014/2020: independent engineering critique of a related device; his claim that Hieronymus admitted wording the patent so that it could not be used to build a working device is unsourced in the text [web, Investigator note], and I do not weigh it. It conflicts with Campbell building from the patent alone.
• Absence of evidence: Radiation Laboratories, Inc. disappears around 1950. The inventor's reasons (banking rules for out-of-state notes; AMA/FDA pressure) are his own. Competing explanations: commercial failure, regulatory pressure, absorption into Pathometric/other firms, loss of customers' confidence. Hite says Hieronymus "remained uncensored" [web]. The Investigator found no FDA notice of judgment naming him. Records that would decide: Missouri corporate filings, FDA notices of judgment, Kansas City newspapers, the Experimental Special Data Bulletin (Jan 1948 to at least July 1949).
• Related lines (Abrams, Pathoclast, Campbell's symbolic machine, modern radionics): lineage and context, not support.
• Hall's 1938 patent: thermal detector, no bearing on element emission (see Corrections).

Summary of verdicts

What would settle it (to leave the final weighing to the reader)

1. Blind trial of identification: a helper prepares coded, sealed samples (e.g. 10 elements and blanks, random order); the operator, who cannot see them, records scale readings for each; success is compared with chance. Ten samples with five correct would be 1 in 1000 under random guessing, readily detected. This could be done with Campbell's or any replica built from the patent; it needs no new physics and has not been recorded.
2. Instrument check: a photomultiplier or spectrometer at the pickup coil with a sample present and absent; a spectrum analyser on the conductors. If the "radiation" exists in the claimed band it shows.
3. Unplugged and shielded controls for the detector: same operators, mains disconnected, a metal-screened enclosure; compare with the powered case (tests Hite's and the ideomotor explanations against the inventor's).
4. Plants: a species and protocol fixed in advance, light-tight verified with a photometer, many boxes randomly assigned to wired/unwired, chlorophyll assayed by absorbance, scored blind.

Source honesty and limits

• Primary text read in full by the deep reader and cross-read by me in the deep notes; the figures were read from a rendered page by Step 1 (I relied on their reading, since I did not re-render p.1). My statement about Fig. 3 numbers is approximate for that reason.
• Everything about the inventor's later work, Campbell, Hite, Tompkins & Bird comes from the Investigator's notes of web sources; I did not re-read those sources myself, except the Hall patent. They are retained as they stand and tagged [web, Investigator].
• [memory] items at medium confidence: skin depth in copper at optical frequencies; electrovibration voltage and frequency range; ultra-weak photon emission rates; flash-greening light levels; the claim that patent examination does not require demonstrated operation. None of the verdicts rests on those; the verdicts rest on the Planck-factor and dispersion calculations (physical constants and the fused-silica formula, high confidence) and on the absence of any recorded controlled test.
• Web checks (2 of about 6 used): Hall patent content verified [web] https://patents.google.com/patent/US2115578A/en ; etiolation page [web] https://en.wikipedia.org/wiki/Etiolation (partial support only; the POR enzyme step is retained at [memory: high]).