StClairJohn: Science Contrarian review (Step 2)

Reviewer role: judges by experimental evidence, not consensus; builds the strongest honest case; examines how the claims were rejected; applies the same rigour to the inventor. Tags: [read] document, [web] URL, [memory] with confidence.

Reading done: facts.json, claims.md, all five deep reads (US 2003/0197093 in full; 0164824 key sections and numeric checks; others via facts.json summaries and spot passages), the Investigator's sites.md and all research/ notes (research/files/ is empty, so no downloaded documents). Primary text checked: _text/US20040164824-*.txt (the single experimental sentence, FIG. 1 legend), _text/US20030209637-*.txt (8971 m/s, "trillion"). My own computations are in section "Checks I ran". Web budget: the search tool's session quota was already exhausted; I made one fetch (Google Patents status). Several Investigator leads (Horsburgh thesis, USPTO file wrappers, copyright deposit copies, Pais comparison) remain unread; I say so where it matters.

Evidence base in one paragraph

Five unexamined-to-grant US applications by one self-filing inventor (John Quincy St.Clair, "Hyperspace Research Institute", San Juan PR; no assignee, "federally sponsored research: Not Applicable"). Nobody else is quoted as having tested anything. The five documents contain no measured quantity except one sentence and one photograph in 0164824 [read]. The Investigator found three sister applications (not in the folder) with the inventor's own experiment descriptions (0200925, 0102810, 0285325/2007) [web, per research/2026-10-04-freepatentsonline.md; I did not re-fetch them]. So the honest summary is: "a set of claims never tested by anyone but the inventor, with crude self-reports", not "claims tested and refuted". Both halves matter below.

How the claims were treated (the "rejection" question)

• Google Patents shows 0197093 "Abandoned", and 0145019 abandoned [web: https://patents.google.com/patent/US20030197093A1/en]. A summary from the fetch tool (a small model's reading of the page, one source) says 0197093 lapsed on 2003-11-06 for failure to respond to an office action. The examiner cited six references (Minovitch US5052638, Willis US5305974, Brotz US6318666 and others) [web, Investigator], so it was searched, but I have not read any office action. Abandonment for non-response is not a scientific rejection; there is no recorded negative experiment, no reanalysis, no blocked publication, no funding denial. The inventor never sought peer review (Google Scholar: no paper by him [web, Investigator]).
• Category for the whole corpus: NOT "tested and failed"; NOT "dismissed after unsound tests". Closest: never properly tested, with no one on record having looked (the neglect is nearly total, and partly self-inflicted: no quantities, no apparatus description, no data). Where the claims are internally broken or violate conservation laws, no experiment is needed (Refuted by calculation).
• Absence of evidence (record goes quiet after 2008; last filing 2006-06-07, last copyright TXu001618682 2008-03-17). Competing explanations, from the Investigator's log: (a) one-person private effort that stopped (self-filed, no assignee, no funding; a forum remark that the "institute" was his bedroom is unsourced); (b) inventor moved to Madison WI (LinkedIn profile, EverybodyWiki, both unsourced or weak) and continued privately; (c) death or incapacity (no obituary found; a Reddit snippet about his father's death is about the father). (d) Suppression/classification: nothing in the record supports it and nothing excludes it; a one-person patent applicant with abandoned applications leaves a typical thin record. What would decide: USPTO file wrappers (application nos. 10/127,098; 10/142,582; 10/170,847; 10/372,556; 11/017,093), Puerto Rico corporate registry, Copyright Office name search for post-2008 entries.

Claim 1: Oppositely wound coils give a negative "spacetime spring constant", negative mass, and a wormhole to "hyperspace" (0197093)

a. Mathematics. Arithmetic is correct: 1/(r+a)^2 - 1/r^2 = -a(2r+a)/[(r+a)^2 r^2] = -8/9 for r=1, a=2 [read p.9; deep read]. But the argument is not internally consistent: (i) the inner coil's minus sign is assigned, not derived; (ii) "curvature = 1/r^2" is non-standard (a circle's curvature is 1/r), and the FIG. 7 unit check to N/m works only with it; (iii) the FIG. 8 expression keeps only the curvature factors, so it is not in N/m; (iv) negative K with positive M gives omega^2 < 0, i.e. exponential growth (instability), not "therefore M is negative"; (v) the FIG. 9 metric contains none of the coil quantities and puts tesla directly into a metric slot with no G/c^4. The 8/9 = cos^2(19.47 deg) "coincidence" is guaranteed by choosing the radius ratio 1/3 (deep read; I re-derived: k = (r/R)^2 - 1). Verification of the GR step: I recomputed the Einstein tensor of the FIG. 9 metric with SymPy. Result: G_tt = -B^2(1 - sin2θ)/[r^2 + B^2(1 - sin2θ)]^2 (matches the deep read), always <= 0. Negative "mass" is a property of the chosen ansatz everywhere, and the sharp spikes are where the metric degenerates as r -> 0. b. Experiments. No measured quantity (no current, field, mass, frequency). Bucking/bifilar coils are ordinary (they cancel far field and inductance) and no experiment has shown negative mass from them [memory, high]. Negative energy density is known only in quantum settings (Casimir effect, squeezed states), tiny and bounded by quantum inequalities [memory, high]. The inventor's own sister application 0102810 reports that a flat bifilar bucking coil produces "hyperspace energy... felt as it rises through the feet" and body rocking for nine seconds [web, FPO via Investigator]: a subjective report, no instruments, no blinding. c. Tested? By the inventor only, and only subjectively (above). Not replicated. Lightning anecdotes (tire rim merged into tree; wheat straw in a pole) are offered without place, date or witness; they are reports, not data, and no mainstream explanation by the examiner is on record. d. If true: GR would need laboratory-scale sources of negative energy density with no quantum inequality bound, and the weak energy condition would be violated by bulk coil fields whose measured stress-energy is positive (B^2/2mu0). Morris-Thorne 1988 show a wormhole throat needs exotic matter [memory, high]; that is a requirement, not a recipe. Also a new "hyperspace" where c is 8971 m/s would have shown up in precision tests of Lorentz invariance.

Verdict (split): the derivation of negative spring constant to negative mass: Internally inconsistent (assigned sign; wrong inference from instability). The physical claim (a coil makes a wormhole/negative mass): Never properly tested (no neutral test on record, and none was attempted; not dismissed after testing). Strongest honest case: the bucking-field geometry is real and constructible, and a numerical prediction could be made if the inventor had given a current and size; he gave none, so nothing is falsifiable as stated.

Claim 2: A micron-sized braided gold coax generates "hyperspace energy" (white mist, FIG. 1); radii follow from tetrahedron geometry (0164824)

a. Mathematics. The radii a = 8.342461828 um, b = 7.559058141 um come from four "constraints"; the deep read could reproduce them to 1.5e-4 only by solving three constraints with c free (c = 0.3531 um, not the printed 0.3596 um); the nested-radical 80.1104395 gave 80.30; the "unity" sum gave 1.0036. I confirmed independently: sqrt(4/3) x (-95.91546344) = -110.7536373 and (sqrt(4/3) - 1) x base = -14.83817383 as printed; exp(-14.83817383) = 0.3596 um; and -ln(reduced Planck length) = 80.1104 [see checks]. So the easy arithmetic is right, the hard steps are not reproducible, the constraints are SI-unit choices (logarithm of a dimensional length depends on the metre), and ten significant figures are not supported by the inputs. Numerology of this kind can be fitted to anything; it carries no predictive content unless it predicts something not already used. b. Experiments. The only evidence: "Electrical experiments with micron-sized braided gold wire coaxial cable show that it is capable of generating substantial amounts of hyperspace energy" plus FIG. 1, "electron microscope photograph" [read p.17, para 0002 and legend]. The deep read of the image: small low-resolution halftone, bright cloud-like patches on black, no visible braid, no scale, magnification, voltage or instrument label. Diffuse bright halos around charging, poorly conducting or contaminated samples are common in electron microscopy [memory, high; a competing explanation, not a finding about this picture]. The image neither supports nor excludes anything because the object is not identifiable. A gold cable of 16 um outer diameter is physically feasible; no drive signal, power or measurement is stated. c. Tested: by the inventor, one sentence. No controls, no second observer, no replication. This is the nearest thing to an experiment in the five documents and it is non-diagnostic, not negative. Nobody else tried to build it, and the instructions are not enough to build it (no dielectric, no frequency, no power). d. If true: a medium with permittivity so high that c drops from 3e8 to 1 to 9e3 m/s that emits light-like "mist" would have to be a new physics sector coupling to ordinary EM, not a refractive-index effect (a high-permittivity dielectric slows light in the dielectric, which is ordinary [memory, high], and does not "flow in from a co-dimension").

Verdict: Never properly tested (not tested neutrally; not dismissed after testing; the sole datum is non-diagnostic). The derivation of the dimensions: Not testable as stated beyond numerology (several printed results not reproduced).

Claim 3: The tetrahedron diagram (base constant ln(h/c) = -95.91546344) derives constants, shows 10 non-redundant dimensions, and shows the electron and proton are the same particle (0164824, 0197093; Investigator: copyright VAu000372532, 1996)

a. Mathematics. Mass x wavelength = h/c for any particle by definition of the Compton wavelength, so the "base constant" is the same for the electron, proton and Planck mass and distinguishes nothing. Plotting ln m against ln λ puts every particle on one line of slope -1; "the electron and proton are one" then follows only by assertion. I checked ln(h/c) = -95.9155 (-95.91548 with CODATA 2018 h; printed value differs by 2e-5, consistent with older constants). The "24 dimensions, 10 non-redundant, attributed to Witten" is not matched by any statement I know of (string theory uses 10, 26) [memory, high; deep read made the same point; I did not find a citation in the document]. b. Experiments. Electron and proton differ in lepton versus baryon number, in mass (ratio about 1836, not derived by the diagram), and the proton has measured quark substructure (deep inelastic scattering, 1960s-70s) while the electron does not [memory, high]. Those are direct experimental distinctions. Spin 1/2 and charge magnitude coincide, which the inventor can cite, but that is not evidence of identity. c. Tested: the identity claim is tested in the sense that existing scattering experiments contradict the strong form; the "derivation of constants" is not a testable prediction. No replication question arises. d. If true: lepton and baryon number conservation, the quark model and the measured proton structure would all need re-explanation. Investigator lead weighed: the diagram predates the patents by six years (copyright registrations 1996 onward, deposit copies unpublished; I did not read them). That establishes priority and persistence, not correctness. The inventor's one quantitative test lead is in 0285325 [web, Investigator]: a hand-held pendulum beside a 8.5 inch clay dome swung six inches at 2.78 MHz, "compared with e^14.838168 Hz = 2.7806 MHz". My check: exp(14.838168) = 2.78057e6, and 1/(0.3596358547e-6 m) = 2.78059e6. So the "frequency" is simply the inverse of the cable length scale c = 0.3596 um read in 1/m and relabelled Hz: a unit coincidence inside his own system, not an independent prediction from the diagram. A 2.78 MHz wave has a free-space wavelength of 108 m, far larger than a dome of 8.5 inches, so a resonant antenna reading is not obvious either. This is the one claim in the corpus that is cheap to test by anyone (pendulum shielded from air and electrostatic force, blind frequency settings, sweep for the response at other frequencies). Its experimental data (alternating 3/4 inch swings every 20 kHz) are the inventor's own with no control stated.

Verdict (split): "m x lambda = h/c" Established but trivial (identity). "Constants derived geometrically / 10 dimensions": Not testable as stated (numerology; irreproducible steps). "Electron = proton": Tested, not supported (by standard scattering and quantum-number evidence, not by any experiment aimed at the inventor's idea).

Claim 4: EM stress-energy (crossed electrostatic fields; static field plus radio wave; field angular momentum in a metric) gives net lift or propulsion without reaction mass (0209637, 0230675, 0145019)

a. Mathematics. 0209637: mixes Gaussian 4pi with SI eps0; "imaginary newtons" and a time-origin-dependent "lever arm"; curvature never enters the force formula (deep read); the author himself admits a continuous radial field gives zero net force. 0230675: the negative curvature spike exists only because the angular momentum was made imaginary (real value gives G_zz > 0); the two rotors' L_z = qrA cancel; with the patent's own n = 1/h the sign of dL/dr reverses (deep read; I did not re-run these). 0145019: the printed Faraday tensor omits the wave's magnetic field and so is not a Maxwell solution; with B included, div T = 0 in vacuum and the static-wave cross term averages to zero, leaving radiation recoil P/c = 3.3 uN per kW (deep read; consistent with linearity of Maxwell's equations in vacuum [memory, high, Jackson]). The wedge electrostatics (sigma ~ rho^(-2/5) for a 300 deg opening) is correct textbook work, which is worth saying: the inventor can do a classical calculation properly when he stays inside it. b. Experiments. The inventor reports none for these devices. Related published work: asymmetric-capacitor "lifters" produce thrust by ion wind in air and not in vacuum [memory, medium-high]; EmDrive-type claims did not survive controlled tests [memory, medium]; I did not web-verify these in this run (search quota exhausted), so they are background, not load-bearing. Conservation of momentum for closed systems is among the best-tested laws. c. Tested: never built or tested in the record (no dimensions, voltages, thrust). Not dismissed on test grounds. d. If true: either momentum non-conservation or a coupling of EM stress-energy to a "hyperspace" reaction medium. Gravitational effects of lab fields are of order G S / c^3 ~ 5e-43 m^2 per the deep read; I did not recompute. Strongest honest case: electromagnetic field momentum is real (Feynman's disk, field angular momentum qA), and gravito-electromagnetism (Jefimenko; weak-field GR) is a legitimate framework; frame dragging was measured by Gravity Probe B [memory, high]. The claim fails on size: any coupling at lab fields is some 40 orders of magnitude too small, and the document supplies no mechanism to beat that except claim 5.

Verdict (split): static field plus wave giving net force (0145019), and closed-system electrostatic thrust (0209637): Refuted by calculation (superposition and momentum conservation; only P/c remains). Imaginary angular momentum in a metric (0230675): Internally inconsistent. Overall device performance: Never properly tested.

Claim 5: Thrust is amplified enormously by hyperspace permittivity (stated about a trillion) or by inductance gradient on grooved rotors acting with negative curvature (0209637, 0230675)

a. Mathematics. Stated amplification ~10^12 [read, para 0027ff of 0209637]. From the inventor's own numbers, c = 299792458 and 8971 m/s, (c/8971)^2 = 1.117e9 (my check), three orders short; and the "enclosed reference calculation" is not in the document. 0164824 gives "as low as 1 m/s", a different slow-light value (inconsistent with 8971). Whether raising permittivity increases force depends on fixed V or fixed Q (not specified). The rotor-groove force: F = (dL/dr) K J^2 A^2 with the energy "postulated by taking a hint", the radial gradient yields a radial force, not lift (deep read). b. Experiments. None. The only third-party datum is Horsburgh's bubble "beaconing" at about 3 Hz, which the inventor matches to 2.83 Hz from a modified Lense-Thirring formula with a fine-structure factor [web, Investigator via FPO 0200925]. I could not read the thesis. The match is within the tolerance of "a few hertz"; the modification factor is chosen by the inventor. Frame-dragging for a shell is (4/3)(G/c^2)(M/R) omega [memory, high] but G M/(R c^2) for lab masses is ~1e-27 or less, so the term is vanishingly small. That a real bubble observation exists is evidence of an interesting hydrodynamic effect; assigning it to a rotating wormhole is the inventor's reading. c. Tested: no. d. If true: Lorentz invariance and the constancy of c, alpha and the vacuum permittivity would be violated in an accessible sector.

Verdict: the numerical amplification: Internally inconsistent (10^12 versus the inventor's own 10^9; two incompatible slow-light speeds). Existence of hyperspace permittivity: Never properly tested (no experiment aimed at it). Rotor force law: Internally inconsistent (direction, postulated energy, sign).

Investigator findings kept in the record whatever the verdict

• Sister applications with the inventor's own experiments (0200925 "Chakra Vortex Accelerator": arm stays raised after spinning 99 times a day for months; sonoluminescence rig; 0102810 bucking-coil foot sensation and 9-second body rocking; 0285325 pendulum data) [web, FPO]. All self-reports, uncontrolled; they are the only experimental narrative on record and should be read as such, not discounted as hearsay.
• Value of Omega = c^2/G = 1.346812891e27 kg/m [web, 0285325]. My check: c^2/G = 1.3466e27 with G = 6.6743e-11 (his 1.3468e27 corresponds to G = 6.6732e-11, as he states); resolves the facts.json gap.
• Copyright registrations 1996-2008 (tetrahedron diagram first in 1996) [web, publicrecords.copyright.gov]; the Library-of-Congress "tet0565" reference.
• 16+ related applications, US only, no foreign family (Google Patents, Justia) [web]; 0197093 had six examiner citations.
• Related work: Jefimenko's cogravitational "K" field and gravito-electromagnetism (St. Clair uses Jefimenko's symbol) [web: https://en.wikipedia.org/wiki/Oleg_D._Jefimenko]; Lense-Thirring; bifilar/Tesla coil lineage; Pais Navy patents (not compared yet).
• Commentary (IPWatchdog 2013, unmakeme 2021, Reddit/UFO threads) are commentators' characterisations, not tests. The record found shows no replication, endorsement or independent measurement.
• Open leads to close: Horsburgh thesis; USPTO file wrappers; deposit copies of the tetrahedron diagram; Puerto Rico registry; Díaz 2016 and Egely 2024 mentions.

Where the inventor fails the same standard I apply to critics

• Every claimed effect lacks a quantity: no current, field, voltage, thrust, mass. A claim with no number cannot fail a test, which is not a virtue.
• Unit mixing: logs of dimensional quantities (depends on the metre), Gaussian and SI together, tesla in a metric, a length-derived number relabelled as Hz.
• Citation of "enclosed" references that are not in the document (Thorne, reference calculation); Witten attribution without citation.
• Hundreds of thousands of g, teleportation over light-years, medical "surgery-less" extraction, with no supporting datum. Self-reports of body sensations are the only evidence for physical effects.

Corrections to Step 1

1. facts.json for 0197093 says G_tt "spikes at theta = pi/4, 5pi/4 where the metric degenerates". Exactly at those angles (r > 0) G_tt = 0; the spikes are at B^2(1 - sin2θ) = r^2, depth -1/(4r^2), adjacent to those angles, sharpening to the degenerate point as r -> 0. SymPy (my run) gives G_tt = B^2(sin2θ - 1)/[B^2(1 - sin2θ) + r^2]^2. The conclusion (the spike belongs to the ansatz, not the coils) stands.
2. claims.md says 0164824 holds "the one experimental statement in the corpus". True for the five folder documents only. The Investigator found more experimental narrative in sister applications; that should be recorded in facts.json/claims.md as outside-folder material.
3. facts.json says the Ω value is missing for 0209637; it is supplied by the Investigator (c^2/G = 1.3468e27 kg/m). Needs adding.
4. The 2.78 MHz "prediction" in 0285325 is not independent: it equals 1/(0.3596 um) read in 1/m as Hz (above). Not in Step 1 (outside folder) but relevant to claim 3.
5. The ln(h/c) check: printed -95.91546344; with CODATA 2018 h I get -95.915482 (2e-5 off, older constants). Minor, explain in the primer.
6. 0145019's "textbook-correct" electrostatics: stands (wedge exponent -2/5 is right: for opening angle 300 deg the exponent is pi/beta - 1 = 180/300 - 1 = -0.4). I checked.
7. Items not re-verified by me and carried on trust from deep reads: 0230675 sign reversal with n = 1/h, 0209637 force-direction claim, 0145019 T^mu nu cross-term average. Plausible from linearity but not recomputed here.

Checks I ran (SymPy/Python)

• (c/8971)^2 = 1.117e9; c/8971 = 33,418.
• ln(h/c) = -95.91548 (CODATA 2018); sqrt(4/3) x base = -110.75364; (sqrt(4/3)-1) x base = -14.83817; ln(1.616255e-35) = -80.1104.
• exp(14.838168) = 2.780574e6 vs 1/(0.3596358547e-6) = 2.780590e6.
• c^2/G = 1.34659e27 kg/m.
• Einstein tensor G_tt of the FIG. 9 metric (above).
• Wedge exponent pi/beta - 1 at beta = 300 deg: -0.4.

Summary of verdicts