Ives: Physicist review (Step 2, redone after QE round 2)

Reviewer: PhD physicist. Redo after QE called MAJOR gaps (qe.md round 2). Tags: [read] from a document in the folder (deep/ reads, _text, digests, research/); [web] with URL; [memory] with confidence; [calc] my own arithmetic (SymPy/NumPy, project venv, 2026-10-07).

Source status

Now in hand and weighed: the 1941 Ives-Stilwell Part II (JOSA 31:369); the gravitation papers (JOSA 29:183, 1939; Phys. Rev. 72:229, 1947; JOSA 38:413, 1948); the conservation-law derivation (Phil. Mag. 36:392, 1945); the 1949 rod-and-clock paper (JOSA 39:757) with the digests of JOSA 38:879, 40:185 and Proc. APS 95:125; the 1942 radiation-pressure paper; the 1939 rotating-disk paper; the 1941-42 ripple-tank papers; the 1951 Rumford lecture photon passage; Ruderfer's appendix in Moray's Sea of Energy, 5th ed. pp.271-274 [read: review/Morray-TH/_text, 5th-ed. file, lines 16545-16660]. Depth: the six deep reads (1939, 1945, 1947, 1948, 1949, 1941) used page images for equations. The 1942 radiation-pressure paper, the 1939 disk passage, the 1951 photon passage and Ruderfer's appendix I read myself in the extracted text; the ripple papers and the 1937 proposal papers I know from digests plus QE's quoted lines. Still not read: JOSA 42:540 (1952) beyond the earlier reading (in copyright, cited by reference), the 1953 JOSA 43 notes, Nature 168:246, Proc. Roy. Dublin Soc. 26:9, Jones JOSA 29:337, the Pobedonostsev papers beyond the 6-page SPIE paper. Anything on those rests on the earlier files or [memory].

Bottom line for the reader

Ives was a first-rate experimenter and a careful, honest theorist who built a complete Lorentz-ether version of relativity and, remarkably, of the Schwarzschild solution of general relativity (GR). The new papers change my earlier judgement in one important way: his gravitation work is not a one-number fit. It is the exact Schwarzschild solution for a single static mass, re-derived in a fixed-ether language, and he says himself that it adds nothing measurable. Everywhere I could check, his mathematics holds and his theory agrees with mainstream predictions. Where he makes claims that go beyond "same predictions, different explanation" (SR is internally inconsistent; the photon will vanish; earlier radiation-pressure results are wrong; disks dish), either the claim fails on analysis, or it was never tested, or it is a difference of definition. The Ives-Stilwell experiment itself is Established and, with Part II, now answers the caveat the 1938 paper left open.

Mainstream view, stated fairly

SR follows from the relativity principle plus constant light speed (Einstein), or from the relativity principle inside Lorentz's ether framework (Poincare 1904-05). Robertson (1949) and Mansouri-Sexl (1977) showed that Michelson-Morley (MM), Kennedy-Thorndike (KT) and an Ives-Stilwell (IS) type time-dilation test fix the transformation up to the convention for distant simultaneity [memory: high]. Ives said the same in 1937-38 and 1947 [read: 1938 p.226; JOSA 37:810-811]. A Lorentz-Poincare ether with an undetectable preferred frame is empirically equivalent to SR. The mainstream drops it as superfluous (Occam, and Bridgman's operationalism), not as refuted. For gravity, the mainstream view is GR; its weak-field limit is also reproduced by any theory with the right static potential and post-Newtonian parameters beta = gamma = 1, and many "flat-space" or ether-style theories have been constructed that match some of GR's predictions (Birkhoff, Whitehead, Poincare-style). Those that survive have to match the full set of tests, including gravitational radiation and strong-field systems, which is where most fail [memory: high].

Two things the reader needs before the claims

The Schwarzschild metric in plain language. Near a mass M, at distance R, GR says that a clock at rest runs slow by the factor y^(1/2), where y = 1 - 2GM/(R c^2) (G is Newton's constant, c the speed of light), and that a ruler laid along the radius is longer than the coordinate distance by 1/y^(1/2). Ives's "interval" ds^2 = c^2 y dT^2 - R^2 dtheta^2 - dR^2/y is just a compact way of recording those two facts (clock slowing and radial stretching) [read: deep/1939]. For the Sun's surface 2GM/(Rc^2) = 4.25e-6 [read: deep/1939]. Mercury's orbit and starlight at the Sun's limb feel this tiny effect.

What "gamma" means in Ives. Ives's gamma (I write y) is 1 - V^2/c^2 with V the speed acquired falling from infinity, V^2 = 2GM/R. It is not the SR Lorentz factor [read: deep/1948]. His postulate: a body at rest in the field has the length, clock rate and mass of a body moving at that fall speed. That is the "escape-velocity trick", a classic heuristic that gives the right clock-slowing factor.

Claim 1: Moving clocks run slow (second-order Doppler shift of hydrogen canal rays)

a. Mathematics. Consistent. Centre of the Doppler pair shifts by lambda0 beta^2/2; for H2+ at 18,350 V, beta^2 = eV/(m c^2) with m the ion mass, giving 0.0475 A [calc, earlier]. New check on Part II: lambda0 beta^2/2 at 42,260 V for H2+ (mass 2 u) gives about 0.110 A, in line with the paper's computed 0.1098 A [read: deep/1941 Table I; qe.md check 4 gives 0.1114 A at 43,000 V]. Printed slips in the originals (an exponent sign on JOSA 28 p.216; a likely misprint 0.0012 for about 0.0118 in Part II's H3O entry) do not change the result.

b. Consistency with experiment, and the old caveat. In 1938 every shift (at most 0.047 A) was below the 0.09 A H-beta doublet separation, so a change of intensity ratio between the two components could fake part of it. The authors set a goal of doubling the shift. Part II reached about 43,000 V with observed 0.113 and 0.1145 A (computed 0.1098 and 0.1073) at 42,260 and 42,280 V, about 1.2 times the doublet spacing, with "the relationship previously observed ... maintained" [read: deep/1941 p.1, Table I]. A pure intensity-transfer effect cannot give a shift larger than the doublet spacing, so this answers the caveat on the authors' own numbers. Qualifications, kept: the new points are on coarser plates; the ten Table I entries (9 H2, 1 H3) mostly sit above the computed curve (6 above, 1 equal, 3 below by 0.0007-0.0054 A) which the authors attribute to a mirror under 100% reflecting [read: deep/1941]; and the spacing 0.09 A is inferred from "about half the separation", not printed. Jones's 1939 objection (stationary particles dragged along) was tested: with the mirror covered an apparent velocity of about 750 m/s at 14,000 V, with it in no shift within error; a double-ended tube gave 0.0744 A against 0.0717 A at 28,300 V, one point without an error bar [read: deep/1941]. Authors with a stake (Ives and Stilwell) judging their own control: fair but not independent. Later repeats: Mandelberg and Witten 1962 (exponent 0.498 +/- 0.025 against 0.5, via Giuliani); Saathoff et al. 2003 (about 1e-9) [web: https://www.rlp-forschung.de/public/facilities/84/publications/45197]; Reinhardt et al. 2007 Nature Physics (2e-10) [web: https://link.springer.com/article/10.1038/nphys778]; Botermann et al. 2014 PRL 113:120405 (below 4e-9, Li+ at 0.338 c) [read: arXiv 1409.7951 abstract]. Independent of atoms: CERN muon ring, gamma = 29.3, agreement to 2e-3 [web: https://inspirehep.net/literature/124922]. Dissent, kept: Pobedonostsev and Parshin (Khlopin Institute; SPIE 1989), repetition at 150-300 keV and 1.5-2.0 MeV, report the second-order shift 40-82% smaller than SR at 150-300 keV and within about 10% at MeV energies; deep-read of 6 pages, no systematic-error budget, separate optical arms, and their classical model does not explain the deficit [read: deep/Pobedonostsev]. One published paper by authors with a rival explanation; no replication found. It conflicts with storage-ring tests at 1e-9, which use independent methods and much larger boosts. I cannot assess it without a systematic budget; it stays an open item, not refuted by silence.

c. Tested and replicated: yes. Interested party: Bell Laboratories (employer) wrote that the result "furnished the first direct proof" of the compensation embodied in Larmor, Lorentz and Einstein [read: Bell Laboratories Record Aug 1947]; Einstein called it "the most direct proof yet" [read: NYT 27 Apr 1938, headline and lead only].

d. If false: Lorentz transformations, accelerator lifetimes and GPS timing would fail.

Verdict: Established (time dilation to about 1e-9 in modern repeats; Ives and Stilwell's own accuracy about 5% statistical, 10-15% by later reassessment; Part II removes the doublet caveat). Pobedonostsev-Parshin stays on record as unreplicated dissent.

Claim 2: The result shows a real slowing in a fixed ether, favouring Larmor-Lorentz over Einstein

2A. IS plus KT fix clock rate and contraction (l = 1). Valid and Ives's best contribution; the 1947 note shows MM alone allows Voigt, equal-volume and no-clock-change variants and the canal-ray points exclude them [read: JOSA 37:810-811]. His 1937 papers 27:177 and 27:389 show KT "cannot by itself test the Larmor-Lorentz hypothesis" [read: qe.md 3.12]. Verdict: Established.

2B. More decisive for Larmor-Lorentz than null experiments because null results could be entrainment or ballistic light. Not supported. IS gives the same prediction in SR and Lorentz-Poincare theory, as Jones (1939) and Robertson showed. Entrained ether is excluded by stellar aberration and Fizeau drag; ballistic theories by double-star data (de Sitter 1913, with later objections) and fast-pion photon-speed measurements (Alvager et al. 1964) [memory: medium; not web-checked]. Those exclusions do not depend on IS. Part II's discussion credits the prediction to Larmor and Lorentz and uses neither "ether" nor "relativity" in its pages [read: deep/1941]. Verdict: Tested, not supported (the claim of special support over SR); the comparison against entrainment and emission was settled elsewhere.

2C. Ether versus Einstein's interpretation. The difference is an undetectable preferred frame and absolute simultaneity. Ives conceded the elusiveness of the ether by "any measurements of the sort we have been considering" (1937) [read: claims.md]. Bridgman (sympathetic) judged a real but indeterminable absolute simultaneity "a complete negation of the operational attitude" [read: Lalli 2014 quoting the 1959 memo; memo unseen]. Verdict: Not testable as stated.

2D. Rotation. Ives derives the first-order Sagnac difference himself in 1938 and objects to the interpretation; the 1953 text claims SR applies only to uniform motion [read: JOSA 28:296; 1953 Genesis]. Von Laue (1920) and Langevin (1921) derived Sagnac in SR; relativity and a fixed ether predict the same first-order result [web: https://en.wikipedia.org/wiki/Sagnac_effect]. Verdict: Refuted by calculation (the statement that SR cannot handle rotation); the 1938 derivation itself is fine.

a-d summary: (a) mathematics consistent; (b) consistent with experiment, identical to SR; (c) not separable by any test; (d) if true, the mainstream would add an undetectable entity but change no prediction.

Claim 3: Ives's alternative theory from classical ether physics (rod-clock quotients, transported clocks)

Three routes. All are checked against the text.

(a) 1937 transported rods and clocks. Reduces exactly to Lorentz as transport speed goes to zero and equals SR for the same transport procedure; difference is a synchronisation convention, first order in transport speed: about 1.7 ps over 1 km at 300 m/s [read: claims.md, 40-digit check]. Robertson checked and found valid, "completely equivalent in substance" [read: NAS memoir p.24; stake: Robertson refereed Ives]. Misprints in eqs. 14, 25, 33, 34, 38-40 do not alter conclusions.

(b) 1943-45 conservation-law route. Mass-velocity law M = M0/(1 - v^2/c^2)^(1/2) from classical radiation pressure on a moving mirror (the classical result P = 2E(c-f)/(c+f), [read: deep/1945 step A]), radiation momentum E/c and mass equivalent E/c^2 (Maxwell/Poynting/Poincare 1900, pre-SR [memory: high]), and conservation of mass, momentum and energy. Clock factor D1, length factors D2, D3 then fixed by D3^4 D1^3/D2^2 = (1 - V^2/c^2)^(1/2) plus two light-clock relations D1 D2 = 1 - V^2/c^2, D1 D3 = (1 - V^2/c^2)^(1/2), giving D1 = D2 = (1 - V^2/c^2)^(1/2), D3 = 1 [read: deep/1945 eqs. 30-33; SymPy reproduces]. The deep reader identified three load-bearing premises that conservation laws alone do not supply: the inertia of energy; an unexplained step that the extra term "must drop out" (equivalent to a relativity principle for the process, and it yields the relativistic velocity addition w = (V+v)/(1+vV/c^2)); and universality and isotropy of clock behaviour (equivalent to MM and KT) [read: deep/1945 step F]. My assessment: a legitimate and elegant derivation of the same group structure; "superfluous hypothesis" overstates it, because the relativity principle re-enters as the "must drop out" step. No measurable difference from SR is named. This is the Lorentz-Poincare viewpoint of 1900-05 that Ives re-derived with radiation as the probe.

(c) 1948-51 operational route and the "inconsistency" charge. He reserves "velocity" for ether quantities, defines the "rod-clock-quotient" Q (first in JOSA 38:879, 1948), and asserts Einstein's c was "firmly established by experiment" wrongly, since only out-and-back speed had been measured; the constant ratio in his revised transformations "is not equal to c and should not be called the velocity of light" [read: digests 1948, 1951]. The 1949 charge: SR "leads to an internal inconsistency if clocks are admitted for distant time setting" and the Lorentz transformations are "an approximation" [read: JOSA 39:757]. Deep-read result, which I accept: his eqs. (19)-(20) are term for term the Lorentz transformation written with clock readings lagging Einstein time by k r' (SymPy identity); SR predicts exactly his one-way result c/(1 - ck) for a clock carried at speed q, about c + q/2, tending to c as q goes to 0; offsets are tiny (a clock walked at 1 m/s over 10 m: 5.6e-17 s; fractional light-speed change 1.7e-9); Ives names no experiment, and the 1950 and 1951 papers cite only MM and IS [read: deep/1949]. The identification of q with proper velocity (gamma u) is the deep reader's [memory: high]. So there is no contradiction in SR: it defines synchronisation by light signals and also predicts what transported clocks read. What survives is a true standard point: one-way light speed and simultaneity are conventional to the extent that clock transport and light-signal synchronisation are different procedures that agree in the q to 0 limit (Reichenbach, Gruenbaum [memory: high]). This also answers Ives's position that q = 0 "could not be performed": SR does not require it, since the finite-q correction is calculable.

a. Mathematics: valid and, in the three routes, equivalent to SR (b: same experimental content; c: no separating test; d: nothing in mainstream predictions would change). Reception [read: Lalli 2014, Lalli 2016]: Robertson's referee report on the 1947 Phys. Rev. paper called the Lorentzian framing "a bit (ca. 40 yrs) out of date", yet it was published; Miller: never "developed to the point where it could be seriously considered as an alternative". Ives's 1951 letter to Butterfield ("the classical English school ... against the paradox of the Semitic school") is, in Lalli's reading, implicitly anti-Semitic; one sentence quoted, letter unseen, Lalli 2013 unread. It is documented attitude of the inventor in a single historian's account, relevant to motive and reception, not to the equations. Verdict, split: Ives's three derivations: Historically superseded but instructive (valid mathematics, empirically equivalent to SR; the 1945 route is the most instructive because it shows how little of SR is extra). The charge that SR is internally inconsistent over clock setting: Refuted by calculation (the finite-q results are SR's own; shown by the SymPy identity). The charge that Einstein's "c is established by experiment" overreached for one-way light: partly right, as a statement about conventionality; Not testable as stated.

Claim 4: Ives's gravitation theory is the Schwarzschild solution in ether language (verdict changed)

My earlier file said "Never properly tested; a fit to one number; extra terms looked chosen". That was wrong. It rested on an abstract. The papers show no constant is fitted (see 4a). I correct it here and in "Corrections to earlier documents".

a. Mathematics. Consistent, and I reproduced the key numbers. - 1939: from three assumptions (equivalence to fall from infinity; null interferometer reading preserved; Fitzgerald contraction relative to local light speed) he gets light speeds c_R = c y (radial) and c_T = c y^(1/2) (transverse) and the exact interval ds^2 = c^2 y dT^2 - R^2 dtheta^2 - dR^2/y, which he calls "the Schwarzschild formula of general relativity" [read: deep/1939 eqs. 1-2, SymPy]. It is exact in Schwarzschild coordinates, not a 1/c^2 truncation. Light bending is stated in words only; his light-speed field gives the full GR limb deflection, 1.7512 arcsec by my calculation (4GM/(c^2 R) at R = 6.957e8 m) [calc], not the "half" value 0.88 arcsec. He rejects an isotropic-light-speed ether, which gives no star displacement, as "in conflict with observation" [read: deep/1939]. - 1947: rewriting Birkhoff's flat-space equations in ordinary time and force, "A particle in motion in a gravitational field experiences an additional force, normal to the direction of motion, of value v^2/c^2 times the component of the gravitational force normal to the direction of motion" [read: deep/1947 theorem, p.3]. For Mercury v is about 48 km/s, so v^2/c^2 = 2.5e-8 [calc]. He does not endorse Birkhoff as physics ("An independent physical derivation of these transverse forces would be welcome"), and notes an infinite number of energy-neutral extra terms could be written [read: deep/1947]. The Gauss analogy is a resemblance only [read: deep/1947]. - 1948: one postulate gives mass m_g = m0/S with S = (y - rdot^2/(c^2 y) - r^2 thetadot^2/c^2)^(1/2), force equations (22)-(23), orbit equation d^2u/dtheta^2 + u(1 - 6 G^2 M^2 b^2/(c^2 h^2)) = K-bar (eq. 25), where u = 1/r, h the angular momentum per unit mass and b = 1 to 1 part in 1e8. Perihelion advance per orbit 6 pi GM/(c^2 a (1 - e^2)): I get 42.98 arcsec per century for Mercury [calc: a = 5.791e10 m, e = 0.2056]. Mass variation alone (Sommerfeld's equation, coefficient G^2M^2/(c^2 h^2)) gives exactly one sixth, 7.16 arcsec per century [calc; read: Ives 1948 p.416]. Nothing is fitted. Two steps are Ives's own choices (F_theta = 0; the split into a Birkhoff-like pair); footnote 9 admits GR gives "no ground for favoring" his pair over Birkhoff's symmetrical one, and the deep reader showed both give identical orbits [read: deep/1948]. - Ives is explicit: "The force equations are actually implicit in the general relativity solution, which is commonly claimed to negate the concept of force"; and r^2 thetadot/S = h "is a relation also obtained in the general relativity treatment" (Eddington 1923; Phillips 1922) [read: deep/1948 p.416]. The printed force and orbit equations are first order in 1/c^2; the deep reader's circular-orbit check shows a difference of about 2.5 parts in 1e8 for Mercury, absorbed in the solar GM [read: deep/1948 [calc]].

b. Consistency with experiment. Because the interval is Schwarzschild's, it inherits GR's results for a single static central mass: perihelion advance (43 arcsec/century), gravitational red shift (Pound-Rebka, 1960 [memory: high]), deflection 1.75 arcsec, and Shapiro delay (same light-speed field; Ives did not discuss it, and it follows [calc by inspection: radial speed c y]). Not addressed by Ives: Shapiro delay, binary pulsars, gravitational waves, frame dragging, strong-field and cosmological tests. Those are outside a theory that has only the one-body solution. GR passes all of them (Hulse-Taylor period decay, LIGO) [memory: high].

c. Tested. Nobody tested Ives's formulation separately because it predicts the same numbers as GR for the cases it covers. Reception: published in JOSA and Phys. Rev., so not suppressed; Ruderfer later used it (below). Ives himself noted where the framework is underdetermined: his 1939 chronotopic-interval paper lists several combinations of contraction and clock rate in a field that all give null interferometer results and says choosing needs experiments "where the gravitational parameter varies", such as near the Sun [read: digest 1939 JOSA 29:294, p.7-8].

d. If true: nothing in GR's predictions changes. What would change is interpretation: gravity as forces in flat space with an anisotropic-ether light speed rather than curved space-time. Ives himself says so: "no non-Euclidean geometry, nor new law of gravitation", but "anisotropy of the ether" [read: deep/1939].

Physicist's assessment of the scope. The theory is the Schwarzschild solution without field equations. GR determines the metric of any matter distribution from Einstein's field equations; Ives's postulate fixes it only for a static spherically symmetric source (and a test particle in it). It says nothing about two bodies, radiation, or how the field responds to its source. Ruderfer's statement that "any Lorentz-ether theory of gravitation which provides the static Newtonian gravitational force is necessarily equivalent observationally to general relativity" [read: Moray 5th ed. p.272] is therefore too strong. Observational equivalence at post-Newtonian order needs the extra input that the theory gives beta = gamma = 1, which Ives builds in through his postulate, not from a static Newtonian force alone; for example, scalar flat-space theories with a Newtonian static limit give the wrong light deflection [memory: medium; not web-checked]. Ives's own hedging ("purely optical significance", gravitational equivalence assumed) is more accurate than Ruderfer's gloss. On Birkhoff's flat-space theory: it reproduces classic tests but admits as sources only a perfect fluid whose sound speed equals c, a severe constraint, and has no derivation from more general principles [web: search summary of Birkhoff's critics; the arXiv paper "Field Theory of Gravitation: Desire and Reality", https://arxiv.org/pdf/gr-qc/9912003, and the St Andrews MacTutor page on Graef, https://mathshistory.st-andrews.ac.uk/Extras/Graef_Einstein/ turned up in the search; I did not open them]. Ives only rewrote Birkhoff's equations; he did not adopt the sound-speed constraint.

Verdict, split: (i) For a single static mass, the predictions (perihelion advance, red shift, light bending): Established (they are GR's; Ives's contribution is the derivation). (ii) Ives's derivation shows the postulate is internally consistent and not fitted: no verdict change needed beyond (i). (iii) As an independent ether theory of gravitation: Never properly tested, and incomplete (no field equations, no radiation, no multi-body treatment; it was never tested because it makes no different prediction). (iv) The ether reading ("anisotropy of the ether", forces in flat space) versus GR's geometry: Not testable as stated. (v) The 1939 conjecture that a large interferometer might detect motion in a field: Never properly tested (Ives expected invariance; no experiment on record).

Claim 5: Einstein did not derive E = mc^2 in 1905

Unchanged from my earlier reading, which was from the 1952 paper itself [read: JOSA 42:540-541]. Ives accepts E = mc^2 and gives his own derivation (Poincare 1900 radiation momentum S/c^2, relativity principle; Pauli 1920). Algebra manipulates correctly. Whether it shows circularity depends on whether H - E = K + C is independent: Stachel and Torretti (1982) say yes; Ohanian (2008) finds no vicious circle but a non sequitur valid only for slow internal motions [read: deep/Ohanian]; Planck (1907) raised the first-approximation worry [read: Ives 1952]. Jammer 1961, Arzelies 1966 and A. I. Miller 1981 sided with Ives (via Ohanian). E = mc^2 itself holds to 4e-7 in Rainville et al. 2005 [web: https://news.mit.edu/2005/emc2]. Note that Ives's own 1942-45 derivations use m = E/c^2 for radiation as input (Poincare 1900, pre-SR); the mathematician checks whether that input is already relativistic; as radiation momentum E/c is Maxwell's, I find it classical [memory: high]. a. Mathematics: Ives's algebra correct; (b) E = mc^2 not in doubt; (c) answered in print (Riseman and Young, 1953); (d) nothing in physics, only attribution. Verdict, split: "Einstein's 1905 derivation is circular": Refuted (Stachel-Torretti; the premise is independent). "The 1905 proof is complete only for slow internal motions": Historically superseded but instructive (Planck, Ohanian agree; Laue 1911 and Klein 1918 completed it). E = mc^2 itself: Established.

Claim 6: Atoms are not clocks (Bonizzoni, Giuliani; not Ives's claim)

Their point: the IS formula follows from energy-momentum conservation in photon emission (Schroedinger 1922), not from clock rates. Correct, and it does not weaken the physics: the lab frequency is the Lorentz transform of the rest-frame emission frequency, one formula in two descriptions. Time dilation of genuine clocks is tested independently by muons at gamma = 29.3 [web: https://inspirehep.net/literature/124922]. Giuliani's real contribution is historical. Verdict: Plausible, untested as an interpretive dissent (identical numbers); time dilation itself Established.

Claim 7: Ives's separate physical claims (new section)

7a. The photon (1951 Rumford lecture). Text, checked: "I do not minimize the difficulty of arriving at an explanation of all optical phenomena, such for instance as the Compton effect ... nevertheless, on the basis of a long preoccupation with standing waves, I venture to predict that this will be done, and that the photon will go the way of the 'caloric' that Rumford demolished" [read: _text 1951, lines 2185-2191]. His grounds: standing waves (a photon would have to exist and not exist alternately at loops and nodes); Lorentz's 1910 remark "von Lichtquanten, keine Rede sein kann"; Beck-Sommerfeld bipartition-angle data said to favour the wave picture; the quantum placed in the atom as a "vestibule" [read: same]. a. Mathematics: the standing-wave point is sound physics (the photoelectric response to a standing wave tracks the intensity, which Ives showed and the NAS memoir praises, uncontested). b. Experiment: quantum optics is built on the light field being quantised, and the decisive result postdates his lecture: Grangier, Roger and Aspect (1986) measured fewer coincidences at the two outputs of a beam splitter than any classical wave model allows (13 standard deviations), consistent with a single-photon state [web: https://courses.physics.illinois.edu/phys513/sp2019/reading/week1/GrangierSinglePhoton1986.pdf, via search summary]. In fairness to Ives: the photoelectric effect and Compton scattering can be given semiclassical treatments with a classical field and quantised matter (Lamb and Scully 1969; Wentzel/Beck for Compton) [memory: medium; not web-checked], so his 1951 evidence does not by itself require photons; antibunching and single-photon interference do. Standing waves in quantum optics are modes of the field, with no "photon at a loop" localisation implied. c. Not tested as a prediction by Ives or his contemporaries; tested later and not supported. d. If true, quantum optics, lasers and photon-counting technology would need reformulating; there is no sign of that. Verdict: Tested, not supported (the strong prediction that the photon would be abandoned; the later antibunching results contradict a purely classical field). His partial point, that wave descriptions explain more optical phenomena than the particle picture suggests, is mainstream-compatible. Kept in the record; pertinent to Moray's interest in radiant energy.

7b. Radiation pressure on a moving absorber (JOSA 32:32, 1942). Ives finds P = E_i(c+v)/(c-v) on a perfect absorber moving toward a point source, half the value on a perfect mirror, and says Lorentz's P = E_i and Poynting-Abraham's P = E_i(1+v/c) are wrong [read: _text 1942, eq. 7 and text after it]. His mirror result P = 2E_i(c+v)/(c-v) is the standard Doppler result and I reproduced the mirror momentum balance. For the absorber, my own derivation: let u be the energy density in the source frame and b = v/c. The agent holding the absorber at constant speed against the radiation must supply F, while the absorber's internal energy grows at rate E_a = u(c+v) + F v (incident energy plus work done by the agent), and its mass grows by E_a/c^2, carrying momentum (E_a/c^2) v. Momentum balance F = u(1+b) + E_a v/c^2. Solving: F = u(1+b)/(1-b), identical to Ives's P [calc: SymPy confirms F - u(1+b)/(1-b) = 0]. Poynting's u(1+b) is the incident momentum rate alone, which omits the momentum carried by the absorber's growing mass. So Ives's formula is correct for the quantity he defines (the force needed to hold constant velocity), agrees with SR, and the earlier values are not "wrong" so much as a different quantity. This is a different explanation, not a different prediction. Physical realisation of first-order v/c terms: Poynting-Robertson drag on dust grains [memory: high]. Verdict: Established (a consequence of Maxwell's stress plus energy-momentum conservation; I derived it); the claim that earlier published values are wrong: Not testable as stated (they answer a different question).

7c. The rotating disk "dishes" (JOSA 29:472, 1939). Ives's argument is that the Wood-Hackett "twist" in the double toothed wheel follows from the Fitzgerald contraction alone, and that for a spinning disk (Ehrenfest paradox: rim contracts, radius does not) one should drop the tacit assumption that the disk stays flat; the disk "dishes", like an inverted saw blade, with a paraboloidal profile [read: _text 1939 pp.4-5, Fig. 4]. The mainstream treats the Ehrenfest paradox through elasticity: a body cannot be spun up Born-rigidly, spin-up induces stresses which deform the disk and change its geometry [web: https://arxiv.org/pdf/2305.07953, summary via search]. The kinematic dish Ives draws assumes no elastic restoring forces. Size check [calc]: a 10 cm radius disk at 100 rev/s has beta = omega R/c = 2.1e-7; the dish depth from his geometry is about (sqrt3/2) beta R = 18 nm (my small-beta expansion of his construction); the contraction strain beta^2/2 is about 2e-14, while the ordinary centrifugal strain of a steel disk is about 6e-5, some 3e9 times larger [calc: rho 7800 kg/m^3, E 200 GPa, nu 0.3]. The relativistic effect is lost entirely under the classical elastic response, which any real test would have to subtract. Verdict: Never properly tested (no experiment on record; the effect is far below the classical stress background); and the detail of whether SR plus elasticity predicts dishing or in-plane strain is Not settled in the documents. Kept.

7d. Ripple-tank analogues with Stilwell (JOSA 31:14, 1941; 32:25, 1942). Air-jet sources on mercury, medium speed v = 11.28 cm/s against a wave speed c = 19.20 cm/s, v/c = 0.585. Interference patterns of the moving case equal the stationary one when lengths are scaled by (1 - v^2/c^2)^(1/2) and periods by its inverse; Fresnel drag and aberration emerge. One measured ratio, 0.852, against the ideal (1 - v^2/c^2) = 0.658 [calc: 1 - 0.585^2], attributed to capillary dispersion [read: digest, qe.md check 7]. This is a correct and classical result: waves in a moving medium obey the wave equation with the medium's wave speed in the role of c, and that equation is covariant under the corresponding "Lorentz" transformation (the acoustic-metric idea) [memory: high]. It shows that observers whose rulers and clocks were themselves made of the wave could not detect their motion through the medium, if rulers and clocks contract and slow as Larmor-Lorentz say. It does not test light, and the ripple medium's "contractions" are not the ruler's physical contractions; they are the pattern's scaling. Verdict: Established (as a classical-wave analogy, with the dispersion caveat); not a test of the ether.

7e. Measurement statements (1948-51). Different explanation, not a different prediction (claim 3). Verdict: Not testable as stated.

7f. Interferometer in a gravitational field (1939/1948). Ives's one open door, a large interferometer, is a conjecture he expects to close in GR's favour (claim 4(v)).

Why Ives interests Moray, and whether a Lorentz ether adds forces (the owner's questions)

Why Moray. The link is Ruderfer's appendix "Neutrino Structure of the Ether" (Lett. Nuovo Cimento 13:1, 1975) reprinted in Moray's Sea of Energy 5th ed. pp.271-274 [read: lines 16545-16660]. Ruderfer argues in three steps: 1. The common rejection of an ether applies "to a Galilean ether, but not to a Lorentz-invariant ether since its description of observation is necessarily identical to that of relativity". 2. "The most direct verification of the validity of a Lorentz ether, and one which is independent of relativity, is found in the theoretical works of IVES": a classical derivation of the Lorentz transformations (1945), consistency with experiment (JOSA 1937-52), and a "classical derivation of the Schwarzschild metric, hence ... an identical explanation of the three famous tests of relativity"; and, he adds, "any Lorentz-ether theory of gravitation which provides the static Newtonian gravitational force is necessarily equivalent observationally to general relativity", so "the continued neglect of Ives' gravitational theory is unwarranted". 3. An admissible ether "focuses attention on its properties"; he defines it as "the totality of all the presently unobservable forms of energy that may exist in the microcosm", proposes a sea of low-energy neutrinos whose "total unobservable energy ... may far exceed the total energy contained in matter, i.e. the vacuum is then a plenum rather than a void". Ives matters because a real Bell Labs physicist of reputation (Optical Society medallist, NAS member [read: NAS memoir]) gave Moray's ether a respectable experimental and theoretical footing: an ether that survives every test. Stake: Ruderfer is an ether advocate with his own neutrino-ether model; Moray's book reprints his note. The energy content is Ruderfer's and Moray's addition; none of the Ives papers I read assigns the ether an energy that could be tapped, and no Ives paper mentions Moray [read: claims.md; deep reads]. The step from "a Lorentz ether is not ruled out" to "a plenum of unobservable energy exists" is not supported by Ives's work. It is also a different idea from Ives's: a Lorentz ether is defined by symmetry (it is undetectable by any rod, clock or light test), whereas Ruderfer's neutrino sea is a proposed physical population of particles, which would have to be consistent with neutrino-detection limits and with Lorentz invariance [memory: medium].

"Even if Ives is correct then Aether follows Lorentz and Einstein ... a Lorentz Aether contributes no new forces beyond those specified by Lorentz/Einstein so it's only a difference of explanation of an observed phenomenon." Your reading is right for the core theory, and it is Ives's own: Robertson, "completely equivalent in substance"; Ives 1937, the ether's "elusiveness"; Ives 1939, "no ... new law of gravitation"; Ives 1948, the force equations "are actually implicit in the general relativity solution". Think of the ether as a bookkeeping frame: in it, rods shrink and clocks slow because the motion through it changes the electromagnetic forces that hold matter together. The result, as measured by those rods and clocks, is exactly Lorentz's transformation, and Einstein's version gives the same numbers. Then the ether adds no force. Qualifications, so your summary is exact: (i) In gravity, Ives writes the same physics as forces rather than curvature. The transverse force of size v^2/c^2 times the normal part of gravity (1947) is not a new effect: it is the same thing as GR's position-dependent mass and clock factor viewed another way (his 1948 paper shows the two are equivalent) [read: deep/1947, deep/1948]. Relative to SR alone, gravity does need position-dependent mass, clock rate and length, not only speed-dependent ones, which is why plain SR mass increase gives one sixth of Mercury's advance. (ii) The ether does carry one extra thing: a preferred frame, an unobservable velocity relative to it. It affects no prediction. (iii) Ives added some physical claims that do go beyond (photon, disk, radiation pressure), covered in claim 7. Of these, the photon prediction is the only one that differs from mainstream prediction, and it fails. (iv) The ether in Ruderfer's and Moray's use adds an energy reservoir. That is not part of the Lorentz ether as Ives, Lorentz or Einstein specified it.

The Investigator's findings, weighed

• Ives's papers (now read in the depth listed above): the primary voice, careful and honest about the limits of his results (he flags his own caveats in 1938, 1939, 1947 and 1948).
• Einstein's April 1938 comment (NYT) and Bell Labs' sentence (employer, stake): both camps claimed the result at once, consistent with claim 2C.
• Robertson's roles (memoir author, referee): stake stated. His technical judgement stands on its own; Bridgman and Mansouri-Sexl reach the same.
• Lalli's work: Ives as "the most important American opponent of Einstein" of his era, marginalised, with a "hidden network" of allies. Historical facts about reception and motive; the physics verdict does not depend on them. Keep.
• Pobedonostsev and Parshin: see claim 1.
• Botermann 2014: IS to 4e-9.
• Preikschat/Ruderfer: Moessbauer rotor, 2.0 +/- 5.2 cm/s limit on preferred-frame effects [read: Preikschat thesis]; null for lab motion through the ether.
• Faraj (General Science Journal): an emission-theory advocate arguing IS measured only a longitudinal asymmetry; the supporting algebra is invalid (deep/gsjournal).
• Absence of evidence: Ives's Bell Labs files and the Library of Congress MSS27508 (5,000 items) are not examined; nothing disappeared from the record; the Princeton 1938 hostility account is single-source (Ives himself via Hazelett and Turner; he is a party).
• Ives-Fry 1933 and the photoelectric papers: mainstream and uncontested; standing-wave selective photoemission (ratios 15-25 against about 4 from absorption, ApJ 60:209; peaking 5-10 times above room temperature near 650 K, ApJ 64:128) [read: qe.md 3.10, digests]. Bears on Moray's interest in standing waves; no energy-conversion efficiency is given.

Corrections to earlier documents (for the orchestrator)

1. physicist.md (my own earlier version), claim 4. Said "Never properly tested ... a one-orbit fit ... extra terms look chosen". Wrong. The 1939, 1947 and 1948 papers show an exact Schwarzschild derivation with no fitted constant; evidence: deep/1939, deep/1948 (43 arcsec reproduced, one sixth reproduced [calc]). Verdict replaced in the table below. Same fix needed in arbiter.md claim 4 and primer section 4.4 ("one matched number does not test a force law with that much freedom" and "Light bending, Shapiro delay, redshift ... not addressed in anything read").
2. claims.md claim 4 / my statement of Ruderfer's "necessarily equivalent" claim. claims.md quotes it neutrally; it should carry the physicist's qualification that the static Newtonian force alone does not fix GR's post-Newtonian parameters (see claim 4).
3. physicist.md claim 3 (earlier). "Primary derivations unread / Refuted by SR treatment" now rests on the 1949 deep read. The earlier verdict stands, with the better evidence.
4. claims.md claim 7b says earlier published radiation-pressure results are "wrong" and "not checked against textbook results". I checked: Ives's value is correct and equals SR's for the force needed to hold constant velocity; Poynting-Abraham's is the incident momentum rate only. This is a definition difference. Correct arbiter and primer if they repeat "wrong".
5. claims.md claim 7d says the ripple medium "reproduce[s] the stationary interference pattern when lengths are scaled by the FitzGerald factor". Accurate, but the primer should add that this is the wave equation's own covariance, not a test of light.
6. qe.md section 4 digest errors (H-alpha for H-beta; "He" for H2; 1941 factor; 1942 exponent): confirmed by my reading of the 1941 deep read and 1942 text. Additionally, the digest of JOSA 29:294 says c_R = c y^(1/2) = c_T "along and across"; the 1939 deep read has c_R = c y, c_T = c y^(1/2). Do not use that digest sentence.
7. claims.md claim 1: "nine H2 points ... (six above, three at or below)". From Table I as transcribed in the 1941 deep read: 9 H2 entries, of which 6 above, 1 equal, 2 below; including the H3 entry, 6 above, 1 equal, 3 below. Check the wording against the table.

Verdict table (changes marked)