Ives · Source catalogue

National Academy of Sciences memoir (47 pp.) of Ives, a Bell Telephone Laboratories physicist who led facsimile and television development, held about 100 US patents, was OSA president 1924-25, NAS member from 1933, and collected coins. Its entire treatment of relativity is one framing paragraph by the memoirists plus a long quotation from H. P. Robertson (Caltech), whom they call 'the leading authority in this field'. Robertson says Ives's Ives-Stilwell work is 'one of the most valuable supports' for special relativity and one of three crucial optical experiments (with Michelson-Morley and Kennedy-Thorndike) leading to the Lorentz transformations; that k(v)=1 within error; that he examined Ives's theoretical papers expecting errors, found the deductions valid but 'completely equivalent in substance to the special theory of relativity', with the apparent difference due to Ives's 'insistence on maintaining an aether framework and mode of expression'; and that he never convinced Ives. The memoirists add that, so far as they know, Ives remained faithful to his own interpretation. The memoir gives no prose on the gravitation theory, the E=mc2 critique or interferometers in gravitational fields; those appear only as bibliography titles in the memoir (e.g. Behavior of an Interferometer in a Gravitational Field, JOSA 29:183 and 38:413; Physical Significance of Birkhoff's Gravitational Equations, Phys. Rev. 72:229; Revisions of the Lorentz Transformations, Proc. Am. Philos. Soc. 95:125, 1951). The bibliography lists no E=mc2 paper, but it is incomplete after 1951: it omits JOSA 40:185 (1950), the 1952 mass-energy paper, Proc. Roy. Dublin Soc. 26:9 (1952) and the 1953 'Genesis' and 'Note' papers [read: extraction pp.43-44], so the gap is not evidence about the E=mc2 paper; whether the omission is deliberate (Robertson wrote the relativity assessment) is not recorded. Honors: Rumford Medal 1951, Fitzgerald Centennial lecture Dublin 1951, Medal for Merit. No direct quotation of Ives on relativity appears; his position is known here only through Robertson and the memoirists. Death: the memoir says he died on 13 November 1953 'within the walls' of the Century Club of New York, where he was stricken with a heart attack [read: extraction pp.28-29]; Wikipedia says Upper Montclair, NJ; the digest (part 02, p.29) wrongly says 'at the National Academy of Sciences' and wrongly says 'approximately 170 patents' (the memoir says 'no fewer than one hundred'; about 85-100 legible entries). Longstreth Medal years in the memoir are 1906, 1914, 1918 (Wikipedia 1907, 1915, 1919). STANDING-WAVE AND PHOTOELECTRIC WORK (QE deep read, mainstream and uncontested, relevant to follow-up Q10 on Moray's interests): Ives answered a questionnaire 'I regard my work on the photoelectric effect as probably my most important contribution to science' (1917, then 1922-1938, nearly all at Bell Labs) [read: p.19]. The puzzle was the 'selective' photoelectric effect of alkali metals: with plane-polarized light at oblique incidence, emission can be 20 or more times larger when the electric vector lies in the plane of incidence [read: p.19-20]. Ives guessed (c.1930) that this is a standing-wave pattern formed in the thin alkali film by interference between the incident beam and the beam reflected from the metal substrate, an idea the memoir traces to his 1907-1908 Johns Hopkins thesis on Lippmann colour photography (a film pressed against a mercury mirror records standing light waves with period half the wavelength) [read: p.7-8, 21]. The memoirists: 'This hunch was right ... photoelectric emission, in a wide variety of cases, was proportional to the intensity of the standing waves in the alkali-metal film. Thus was resolved one of the major problems of photoelectricity' [read: p.21]; one of the two memoirists (unnamed) believes this is what Ives meant by 'most important contribution'. Also: film thickness moves the long-wave limit out to the free-atom resonance wavelength and back (all five alkalis, 1929-30 papers); 'Standing Light Waves; Repetition of an Experiment by Wiener, Using a Photoelectric Probe Surface' (with T. C. Fry, JOSA 23, 1933, volume/page garbled by OCR); and the 1951 Rumford Medal lecture 'Adventures with Standing Light Waves', Proc. Am. Acad. Arts Sci. 81:1-32. The memoir gives no numbers, no precision and no energy-conversion efficiency; all evaluative statements are the memoirists' (admirers writing for the NAS), not independently checked here. Other relativity titles in the bibliography not obtained: Interference Phenomena with a Moving Medium I and II with Stilwell (JOSA 31, 1941; 32:25, 1942); Clock Paradox in Relativity Theory (Nature 168:246, 1951); Unsymmetrical Doppler Effect in Hydrogen Canal Rays (Rev. Sci. Instr. 9:283, 1938; see the Faraj entry). [Update 2026-10-07: the papers named by those bibliography titles (JOSA 29:183, JOSA 38:413, Phys. Rev. 72:229, Proc. Am. Philos. Soc. 95:125, JOSA 31 and 32 moving-medium papers) are now in the manifest under Ives-papers/ and have their own entries; Nature 168:246 and Rev. Sci. Instr. 9:283 are still not obtained.]

Key claims

• Ives's 1938-41 canal-ray experiment with Stilwell is one of three crucial optical experiments that together lead to the Lorentz transformations (Robertson).
• The Ives-Stilwell result shows the parameter k(v) equals unity within observational error, a direct confirmation of Einstein's special principle of relativity (Robertson).
• Ives's theoretical deductions are valid but his theory is completely equivalent in predictions to special relativity; the difference is only the aether framework (Robertson).
• Ives considered his work disproved special relativity and to the end remained faithful to his own interpretation (Robertson; Buckley and Darrow 'so far as we know').

Relevance: Contemporary, peer-level account of Ives's relativity work by an eminent relativist, and the fullest bibliography of his 1937-1951 papers in the folder. Robertson's equivalence finding is central to weighing Ives's claims. Also the source of Ives's standing-wave photoelectric work, which the memoirists and Ives rank among his most important contributions.

Mainstream conflict: None in the memoir itself; it states the mainstream (Robertson) view that Ives's experiment supports relativity and his theory is equivalent to it.

SECONDARY source: a history-of-physics essay. Section 5.1 describes the Ives-Stilwell 1938 experiment (realising Einstein's 1907 canal-ray proposal; financed by Bell Telephone Laboratories). Ives and Stilwell viewed the H2+/H3+ canal-ray beam along its axis, with a mirror recording light toward and away from the spectrograph simultaneously, and measured the shift of the 'center of gravity' of the two Doppler-displaced H-beta lines against the undisplaced line: lambda_Q = lambda0(1 + B^2/2), with first-order term 2 lambda_D = 2 lambda0 B cos(theta) (eqs. 21-24, equal to the Mandelberg-Witten notation). They obtained the atom velocity from the first-order splitting and, alternatively, from eV = Mv^2/2, and reported agreement 'within measurements precision'. The essay gives no numbers (voltages, shifts, error bars); the numbers come from Ives and Stilwell's own 1938 paper (Tables I-III, entry below). Ives and Stilwell concluded that the result fixes the clock rate nu = nu0(1 - V^2/c^2)^(1/2) in the ether and, with Kennedy-Thorndike, the Lorentz-FitzGerald contraction; they argued a positive result gives 'more decisive evidence for the Larmor-Lorentz theory' than null experiments. Ives's own account (letter of 24 Oct 1950 to H. Butterfield, quoted from Hazelett and Turner 1979, p.90): at Princeton in 1938 he presented the work in terms of FitzGerald, Lorentz and Larmor, 'It was certainly not diplomatic', and Pauli told him in Europe that Princeton colleagues wrote that his work 'didn't amount to anything' (single published source, Ives a party). R. C. Jones (1939) re-derived the predictions from special relativity, saying they were 'identical with those obtained by Ives from electron theory', and noted the undisplaced line had not been measured; Ives and Stilwell accepted this and answered in Part II (JOSA 31, 1941). Otting (1939) repeated it with no interpretation; Mandelberg and Witten (1962) with better precision. The authors' own thesis: all readings assume atoms are clocks, untenable after quantum mechanics; the experiments corroborate relativistic dynamics and only indirectly the Lorentz transformations. Footnote 20 summarises Ives's theory: an ether exists plus the relativity principle (credited to Poincare), three light speeds, clocks set by transport using 'rod-to-clock quotients', transformations that approach Lorentz's (when the transport speed q goes to 0) but never coincide. QE deep read of Ives's 1937 papers: the phrase 'rod-to-clock quotient' does not occur in the 1937 papers (Ives says 'observed velocity'; it appears in his 1953 'Genesis' paper); and 'never coincide' is accurate as a statement about formulas but not about predictions, since for the same transport-synchronisation operations special relativity gives Ives's numbers exactly. Arthur Miller (cited fn.35) holds Ives-Stilwell is 'the only positive proof' of time dilation; the authors disagree. Pages 1-15 and 22-44 digest only.

Key claims

• Ives and Stilwell measured a second-order Doppler shift of hydrogen canal-ray light that agreed with lambda0(1+B^2/2) within their precision.
• Ives and Stilwell read the result as showing a real slowing of clocks moving through a fixed ether, supporting the Larmor-Lorentz theory.
• The same predicted shift follows from special relativity (Jones 1939), so the 1938 result does not by itself discriminate between the ether theory and special relativity.
• Atoms are not clocks; the experiment corroborates relativistic dynamics, and the Lorentz transformations only indirectly (Bonizzoni and Giuliani).
• Ives held that moving clocks and rods really change, with an ether and a relativity principle, in a theory whose transformations approach but never coincide with Lorentz's.

Relevance: Best source in the folder for what the experiment was and how it was read; secondary, with no numerical data. The exponent slip in the quoted formula is in Ives and Stilwell's printed 1938 paper itself (p.216), not introduced by the secondary authors.

Mainstream conflict: Ives's ether reading conflicts with special relativity's interpretation; the authors also dissent from the mainstream 'atoms as clocks' reading.

SECONDARY source on Ives. Ohanian reports that in 1952 Ives gave a lengthy analysis of Einstein's 1905 derivation, concluding it was logically circular and ending 'The relation E = mc2 was not derived by Einstein' (Ives, 1952), and that Jammer (1961), Arzelies (1966) and A. I. Miller (1981) supported this. The Ives-Jammer-Arzelies argument as Ohanian gives it: substituting the particle kinetic-energy formula K = mc^2(1/sqrt(1-v^2/c^2) - 1) gives K2-K1 = (m2-m1)c^2(...), whose comparison with Einstein's K2-K1 = -E(...) forces E/((m1-m2)c^2)=1, i.e. Einstein 'unwittingly assumed' what he set out to prove. Stachel and Torretti (1982) answered that Einstein's eq. (2) is independent of eq. (3), so there is no vicious circle; Ohanian agrees ('Ives, Jammer and Arzelies are wrong') but also rejects Stachel-Torretti's acceptance of Einstein: the flaw is a non sequitur, an unwarranted extrapolation from a nonrelativistic approximation of internal motions, valid only for slow internal motions. Ohanian accepts E=mc2 as true and credits the first general proofs to Laue (1911) and Klein (1918). His counterexample is a one-dimensional gas in a massless box whose energy is not particle-like for motion along the axis unless wall stresses are counted. Ohanian cites Ives's paper as Am. J. Phys. 42, 540-543; this is Ohanian's slip. The paper is Ives, 'Derivation of the Mass-Energy Relation', JOSA 42(8):540-543, August 1952, confirmed from the saved copy (research/files/, OCR) read by QE. Ives's own abstract credits Poincare (1900, 1904), says Einstein's reasoning, 'questioned by Planck, is defective', and ends 'The relation E = mc2 was not derived by Einstein'; the printed relation is E = m_M c^2 with m_M the mass of matter. Ives accepted E = mc2 and gave his own derivation; Planck (1907) raised the objection first; 'circular' is the Stachel-Torretti / Ohanian label, and Ives narrowed the charge in 1953. Ohanian does not mention the Ives-Stilwell experiment or the ether.

Key claims

• Einstein's 1905 derivation of E = mc2 is circular (Ives 1952, as reported by Ohanian).
• Einstein's 1905 argument is not circular (Stachel and Torretti 1982; Ohanian agrees) but does not establish E = mc2 for bodies with relativistic internal motion (Ohanian).
• The first general proofs of mass-energy equivalence are Laue (1911) and Klein (1918) (Ohanian).
• The relation E = mc2 itself is valid (Ohanian).

Relevance: Only treatment in the folder of Ives's 1952 mass-energy critique; shows a mainstream physicist agreeing in part with its conclusion while rejecting its reasoning.

Mainstream conflict: Ives's circularity charge was rebutted in the mainstream literature (Stachel and Torretti); Ohanian's non sequitur finding is itself a criticism of Einstein's 1905 proof, not of E=mc2.

SECONDARY source. Giuliani retraces the Doppler formula for photons from Einstein (1907, atom as moving clock, nu/nu0 = sqrt(1-v^2/c^2)) through Schroedinger (1922, energy-momentum conservation in photon emission, largely forgotten; Davisson 1938 rediscovered it). Section 2.3 treats Ives-Stilwell 1938: financed by Bell Telephone Laboratories; Ives is 'a staunch opponent of relativity' with an ether-based theory in papers collected by Hazelett and Turner (1979), whose 'grossly anti-relativist preface' Giuliani says shadows the papers (Giuliani's characterisation); Miller's judgement that Ives's Larmor-Lorentz theory 'was never developed to the point where it could be seriously considered as an alternative' (Miller's view). Setup, formulas and conclusion as in Bonizzoni and Giuliani 2000 (eqs. 19-22, same agreement 'within measurements precision'). Jones 1939 (Bell Laboratories) and Otting 1939 are covered. Accuracy: Mandelberg and Witten (1962) assess that Ives-Stilwell and Otting points fit to about 2-3% but real uncertainty is about 10-15%; their own result is an exponent 0.498 +/- 0.025 (prediction 0.5), about 5% overall. Later: Olin 1973, Hasselkamp 1979 (direct 90-degree H-alpha), laser tests (Mansouri-Sexl test theory, preferred-frame framework), Chou 2010 optical clock. Giuliani's thesis: atoms emit photons with energy and momentum, not waves; the wave and clock readings, the ether reading and the photon-recoil reading predict the same sqrt(1-B^2) factor at this precision; the experiments corroborate relativistic dynamics plus quantum hypotheses, and time dilation only indirectly. He gives no numerical plate data for 1938 and does not claim the experiment decides between ether and special relativity.

Key claims

• The 1938 Ives-Stilwell fit was 2-3% but the real uncertainty about 10-15% (Mandelberg and Witten's reassessment).
• The 1962 repeat found the Doppler exponent 0.498 +/- 0.025 against the predicted 0.5.
• The ether theory, special relativity and Schroedinger's photon-recoil treatment all give the same second-order factor sqrt(1-v^2/c^2) at the precision of these experiments.
• Atoms are not clocks; the experiments corroborate relativistic dynamics plus quantum hypotheses, only indirectly time dilation (Giuliani).
• Ives's coordinate transformations converge to Lorentz's only as the transport velocities of the synchronising clocks tend to zero.

Relevance: Gives the accuracy numbers and the later experimental history (including the preferred-frame test theory) that bear on whether Ives's result discriminates his ether reading.

Mainstream conflict: Giuliani's 'atoms are not clocks' dissents from the textbook time-dilation reading; he does not endorse Ives's ether.

Born Philadelphia 1882; BA Penn 1905, PhD Johns Hopkins 1908 (colour photography). Worked for his father Frederic Ives; US Army Signal Corps aerial research head 1918. At AT&T from after World War I: telephotography (demonstrated 1924, commercial 1925), television proposed January 1925; 10 March 1926 demonstration (50 lines, 16 frames/s); 27 April 1927 (as Optica dates it; NAS memoir and Wikipedia give 7 April 1927) Hoover speech Washington to New York over 200 miles, with over 200 engineers; colour television 1929, two-way picture phone 1930, coaxial-cable high-resolution television 1937. WWII night-vision work; Medal of Merit; Franklin Institute Longstreth Medal; Rumford Prize. OSA charter member, over 120 papers (67 in JOSA, chiefly television or relativity); endowed the OSA Frederic Ives Medal and received it in 1937. Died 15 November 1953 (other sources in the folder give 13 November). The page says nothing of his relativity views beyond counting papers on relativity. Published by Ives's own society, a party with institutional pride in him. Discrepancies with other sources: Optica dates the Hoover demonstration 27 April 1927, NAS and Wikipedia 7 April 1927; Optica gives 50 lines (1926 demonstration), the NAS memoir 50 scanning holes, Wikipedia '185-line'; death 15 November 1953 here, 13 November in the NAS memoir and Wikipedia.

Key claims

• Ives headed Bell/AT&T development of facsimile and television, including the 1927 long-distance demonstration.
• Ives published over 120 papers, 67 of them in JOSA on television or relativity.

Relevance: Biographical background for the Ives entry; death date differs from Wikipedia and the NAS memoir (15 vs 13 November 1953).

Mainstream conflict: none

Tertiary source. Herbert Eugene Ives (31 July 1882 - 13 November 1953), American scientist and engineer who headed facsimile and television development at AT&T/Bell Labs (Director of Electro-Optical Research). Born Philadelphia to Frederic Eugene Ives and Mary Olmstead; Penn and Johns Hopkins (1908); married Mabel Lorenz 1908, three children; 1920 aerial photography book; OSA president 1924-25. First colour fax 1924 (Rudolph Valentino); 7 April 1927 (Optica says 27 April) '185-line' (the NAS memoir and Optica give 50 holes or lines; unreconciled) long-distance television of Hoover (station 3XN Whippany); ikonophone two-way television-telephone by 1930; parallax panoramagrams (3D); numerous patents. Best known for the 1938 Ives-Stilwell experiment, 'direct confirmation of special relativity's time dilation', although Ives rejected special relativity and 'has been described as the most authoritative opponent of relativity in United States between the late 1930s and the early 1950s'. Following Lorentz he sought to show the physical reality of relativistic effects; he regarded the experiment as proof of the ether; he was discouraged by the scientific community's reaction. Wikipedia quotes Robertson that Ives's deductions were valid but equivalent to relativity. Numismatist, President of the American Numismatic Society 1942-46. Honors: Longstreth Medal three times (1907, 1915, 1919 per Wikipedia; 1906, 1914, 1918 per the NAS memoir), OSA Frederic Ives Medal 1937, Truman's Medal for Merit 1948 (blackout lighting, optical communication), APS and NAS member. Wikipedia says Ives 'mistakenly thought' his theory would disprove relativity (a judgement not Ives's). Wikipedia places his death in Upper Montclair, NJ; the NAS memoir says the Century Club, New York.

Key claims

• The Ives-Stilwell experiment is regarded by the scientific community as among the best supports for special relativity, while Ives took it as proof of an ether.
• Ives headed the AT&T team (over 200 people) that produced long-distance television in 1927.

Relevance: Overview and biography; death date 13 November 1953 agrees with the NAS memoir (Optica gives 15).

Mainstream conflict: Presents the mainstream view that Ives's conclusions were mistaken.

Primary report of the experiment. H-beta (4861 A) light from hydrogen canal rays (H2 and H3 particles, 6,788-18,356 V; no H1 found) is observed along the beam directly and, through a small concave mirror about 7 degrees off axis, in the receding direction, on a Wood plane-grating spectrograph (10.87 A/mm). The observed quantity is the shift of the centre of gravity of the two displaced lines against the undisplaced line, expected about 0.01-0.05 A, measured with a Zeiss microscope; probable error about 0.0025 A. Table III (read on the page image): eight plates, e.g. 18,350 V, H2: computed 0.0478 (from voltage) / 0.0469 (from line spacing), observed 0.047 A; 6,788 V, H3: 0.0116 / 0.0109 / 0.011 A; observed/computed-from-voltage ratios 0.948, 0.911, 0.945, 0.982, 1.035, 0.980, 0.923, 0.983 (mean 0.964, a 3-4% mean deficit that is the reviewers' and QE's calculation; the authors do not discuss it). The authors say the points 'lie along the theoretical straight lines to within the probable errors' and conclude 'the change of frequency of a moving light source predicted by the Larmor-Lorentz theory is verified'. They call it 'the necessary optical complement to the Michelson-Morley experiment' and the positive result 'more decisive evidence for the Larmor-Lorentz theory' than null results, which could equally be explained by an entrained ether or a ballistic theory. The sign of the centre shift is never stated in pp.217-226 (it follows from nu = nu0(1 - V^2/c^2)^(1/2): towards the red; QE inference, memory high); the paper does not use 'red' or 'blue' for it. Systematic checks the authors report: mirror asymmetry under 1/100 of 1% (tested with an imaging-concave and a plane mirror), spectrograph linearity (10.870-10.877 A/mm), the Earth-motion orientation test N, S, E, W at about 18,000 V (Table II: 0.036-0.052 A, mean 0.0468 vs 0.0472 A computed), magnetic field varied tenfold (no effect), faint molecular H2 lines that gave 'very erratic' shifts until voltages were chosen to avoid them (only seven voltages usable). Admitted 'least satisfactory feature': every shift is smaller than the 0.09 A H-beta doublet separation, so a varying intensity ratio of the doublet components could fake part of it; the remedy proposed is voltages at least twice and preferably several times higher. No separate direct and reflected values are tabulated (only their mean); the authors report no asymmetry between them.

Key claims

• The centre of gravity of the Doppler-displaced H-beta lines from H2 and H3 canal-ray particles is shifted from the undisplaced line by lambda0 V^2/(2c^2), within a probable error of about 0.0025 A.
• The frequency of a moving source falls as (1 - V^2/c^2)^(1/2) relative to the ether, so with Kennedy-Thorndike the moving apparatus contracts along the motion and not across it.
• The result gives more decisive evidence for the Larmor-Lorentz theory than the null experiments, which entrained-ether or ballistic theories could also explain.

Relevance: The central experiment of the whole project; supplies the data (Tables I-III) that the secondary sources lack. Interested-party note: both authors championed the Larmor-Lorentz interpretation they conclude for.

Mainstream conflict: The measurement is accepted; the ether reading is not (special relativity gives the same predictions; Jones 1939). The authors' own caveat on the H-beta doublet is not discussed by the secondary sources.

The only experimental dissent on claim 1 found in the record. A Soviet repetition with H2+ beams at 150-300 keV and 1.5-2.0 MeV on two Khlopin Institute accelerators, H-alpha 6563 A observed at about 77 and 257 degrees on a grating spectrograph with photographic plates (exposures 6-8 h, 48 h at MeV energies); the 257-degree arm by lens, the 77-degree arm through a flexible optical fibre. Table (13 energies): the observed second-order shift is 60, 69, 82, 40, 49, 56, 53, 69, 53 and 42 per cent smaller than the special-relativity value at 150-300 keV (observed/SR from 0.17 to 0.60); 9 per cent smaller at 1500 keV, 3 per cent larger at 1750 keV and 10 per cent larger at 2000 keV. Stated uncertainty 0.04 A per line position (0.08 A on the shift); the low-energy deficit is 0.31-0.52 A. The authors conclude 'The results received by us do not confirm the relativistic theory. Especially considerable deviations (more than twice) are available within the range of 150-300 keV', and also say 'it is possible that observation results are substantially effected by other factors unknown before'. Their classical alternative is Marinov's 1972 formula lambda = lambda0/[(1 - beta^2 sin^2 theta)^(1/2) + beta cos theta], which needs a photon 'rest mass' 2h nu0/c^2; for their geometry it predicts about 5% MORE than special relativity, so it does not explain the deficit, and they propose experiments near 90 degrees and on the angle dependence. Reviewer-side calculations [calc, QE deep reader]: no systematic-error budget is given; the two viewing arms are separate optical paths, and a misalignment of about 0.5 degrees would shift the shift by about 0.4 A, the size of the low-energy deficit (this shows what is possible, not that it happened); the paper's own SR columns imply viewing angles of 78.2 degrees at 150-300 keV and 77.2 degrees at MeV energies, with the red arm 0.11 degrees off antiparallel at high energy; the MeV 'SR' values are 0.25-0.28 A below the pure SR value, against which the observed shifts are 13% low, 1% low and 6% high; the Fig. 7 minimum (2.325 MV) lies beyond the highest measured energy (2.0 MV), so it is an extrapolation. Fuller accounts exist (Pobedonostsev et al., Khlopin preprint PU-199/L, year printed 1966, probably 1986; Zh. Tekh. Fiz. 1989, issue 3, pp.84-99) and were not obtained. No later work or replication by these authors, nor any reply, has been searched for.

Key claims

• A 1989 repetition at 150-300 keV found the second-order Doppler shift 40-82% smaller than special relativity predicts.
• At 1.5-2.0 MeV the same experiment agrees with special relativity within about 10% either way.
• Special relativity and a classical (Marinov) model give nearly the same result for their geometry, so a measurement near 90 degrees is needed.

Relevance: Weighed under claim 1: it is the one experimental dissent, but it has no systematic-error budget, uses separate optical arms, and its own classical model does not account for the deficit. It does not touch the storage-ring tests at the 10^-9 level, which carry the verdict 'Established'.

Mainstream conflict: Disagrees with special relativity at 150-300 keV; the mainstream and later storage-ring results disagree with the dissent.

Ives's primary derivation behind claim 3. He assumes a fixed ether, rods that contract and clocks that slow by (1 - v^2/c^2)^(1/2), and ideal 'unvarying' rods and clocks; clocks are synchronised by transport at a finite 'observed velocity' (W0 for clocks, Y0 for rods: known distance divided by time on the moving clock; this is the modern proper velocity, which can exceed c). The phrase 'rod-to-clock quotient' is NOT in these papers (it is in his 1953 'Genesis' paper). Results: measured lengths and times depend only on W0 and Y0, not on the ether speed v (eqs. 13, 19, 'independent of v'); the measured one-way light speed on every moving body is the same function c_m of W0 and Y0, different from c, approaching c as W0 goes to 0; for infinitesimal W0 and Y0 the expressions are exactly the Lorentz transformations. Footnote 6 is his central criticism: deriving the Lorentz transformations from constant c carries a concealed restriction that all clocks and rods move at infinitesimal speeds. The sequel finds a 'mimicry' of the ether contractions by observations between platforms, states reciprocity ('the observations do not tell us that the observed clock is fast or slow'), and calls special relativity 'a structure built on the special case'. QE checks: against Lorentz formulas used with Einstein-synchronised clocks the finite-q results differ at first order (a set-back of about L W0/(2c^2): 1.7 ps over 1 km at 300 m/s; one-way c changes by about W0/2c, 1.7e-9 at 1 m/s), but against special relativity applied to the same transport operations the numbers agree with Ives's to all 40 digits tested, so the difference is a synchronisation convention, not a physical prediction (agreeing with Robertson). Ives names no experiment that separates his theory from special relativity: the transverse Doppler experiment is 'decisive' for the clock-rate exponent (fn 2), experiments with fast-moved clocks and rods would force qualification of the constant-c assumption (fn 4), his evidence for the ether is stellar aberration, which he calls 'not metric, but by sheer weight of enormous probability' (p.271), and he concedes the ether's 'elusiveness by any measurements of the sort we have been considering'. Misprints (none changes his conclusions): eq. (14) sign, (25) stray exponent, (33) ratio inverted, (34) missing square and root, (38)-(40) missing 1/W0. T. C. Fry did the algebra (fn 5). Only the direction of motion is treated; the general finite-q transformation is left as 'probable'. [Update 2026-10-07: the sequel JOSA 27:310-313 is a separate manifest document with its own entry; this entry's discussion of the 27:310 material remains valid.]

Key claims

• Clocks synchronised by transport at finite observed speed, with FitzGerald-Larmor-Lorentz contractions, give measured lengths and times independent of motion through the ether.
• The expressions reduce exactly to the Lorentz transformations as the transport speeds of rods and clocks go to zero.
• Special relativity is the special case of infinitesimal-speed operations and offers no explanation for the choice of operations.

Relevance: The primary text for claim 3 (Ives's alternative theory); shows it is equivalent to special relativity in predictions and differs in interpretation and synchronisation convention.

Mainstream conflict: Interpretation only (real contraction in a fixed ether versus relativity of simultaneity); no predicted measurable difference.

Ives's own analysis of the Sagnac experiment and the clock paradox. 1938: he reduces Sagnac to a polygon of mirrors with a moving source and detector, so rotation is not essential, and derives t1 - t2 = 4 pi r^2 w/(c^2 - r^2 w^2) (eq. 5), the standard first-order Sagnac result (4 A w/c^2), saying 'This last formula agrees with the result of the Sagnac experiment'. He also reproduces Langevin's result that transported ('local time') clocks give velocity exactly c (eq. 11). He objects to the interpretation: with transported clocks two clocks at one place read different times, 'more than one time at one place is a physical absurdity', and 'the physical fact cannot be evaded by juggling the measuring instruments'. His conclusion: the result leaves 'the argument of Sagnac ... as strong as it ever was'; the effect depends on rotation with respect to the 'pattern of radiant energy from the stars', not on relative rotation of matter. He states relativists have dismissed Sagnac 'as involving motion of rotation' (his characterisation of others). He does NOT say special relativity cannot compute the effect or that it gives a different first-order number; the claim that 'Einstein's scheme applies only to uniform motion' is from the 1953 'Genesis' paper, not these pages. He bounds the rotating-ring case by argument (Ehrenfest paradox cited), saying it differs from his polygon model only at second order. 1937 (clock paradox): the setting change of a transported clock is independent of the platform's ether velocity; the leading term is T W0^2/(2c^2), the standard twin-paradox lag; relative velocity alone does not predict which clock is behind (it depends on motion relative to the ether, shown by stellar aberration), so 'the clock paradox originates in a fallacious conclusion' that only relative motion matters. He also proposes an 'aberration of clocks' test (transported clocks over six months would need stability better than 1e-8 s over a 1e-4 s transit). The usual SR resolution (the travelling clock changes inertial frame) is not discussed in these pages. [Update 2026-10-07: JOSA 27:305-309 is a separate manifest document with its own entry.]

Key claims

• The Sagnac effect is not essentially about rotation and is given by 4 A w/c^2 to first order; it tracks rotation relative to the stars.
• Transported 'local time' clocks give the speed c but only by using a separate clock for each beam.
• The clock paradox arises from assuming only relative motion matters; the ether frame resolves it.

Relevance: Primary source for verdict 2D (Sagnac and rotation) and for Ives's position on the clock paradox, which no earlier file covered.

Mainstream conflict: Interpretation only; his first-order Sagnac number is the textbook one.

Faraj, an independent author with a stake in the emission (ballistic) theory (his related papers are on Michelson 1913, Sagnac and Majorana 1918), says Ives and Stilwell measured 'only a slight asymmetry in the observed longitudinal Doppler effect for equal positive and negative velocities along the line of sight', not the transverse Doppler effect, and that this is 'an emerging consensus'. He does NOT derive or measure this: he attributes it to D. Y. Gezari (2009, 'Experimental Basis for Special Relativity in the Photon Sector') and cites W. Kantor (1971) calling the experiment inconclusive. He does not cite Ives's Rev. Sci. Instr. 9:283 (1938) 'Unsymmetrical Doppler Effect in Hydrogen Canal Rays', so the claim cannot be traced to that paper through this document; that paper and Gezari 2009 have not been read. The asymmetry has a simple meaning: the mean of the blue and red lines differs from the rest wavelength by 0.02-0.11 A, which is the Ives-Stilwell centre-of-gravity shift read as an asymmetry; Ives and Stilwell read the same shift as the transverse Doppler effect. Faraj's argument that the emission theory and Larmor-Lorentz predictions are identical over 850-2100 km/s rests on his eq. 3.5, which QE's reader found invalid: from his own eq. 3.3 the emission-theory shift is lambda(gamma^2 - 1) = about lambda beta^2, twice the Larmor-Lorentz value (1/2) lambda beta^2 (1100 km/s: 0.0653 A against 0.0326 A; for Ives-Stilwell's 0.0202 A prediction the emission value is about 0.0404 A against 0.0185 A observed) [calc, medium confidence pending an outside source]. Table 3.2's observed column (10 values, 0.0185-0.1145 A) is unverified: no per-row source, velocity or ion species. His circularity worry (velocity from eV = m_H c^2 (gamma - 1) shares the factor with the effect) is small at beta of about 0.003. The paper also states that a classical wave theory gives a symmetric null in wavelength but an asymmetric frequency result. Faraj's own Table 3.2 residuals (observed minus predicted) run from -0.0017 A at low velocity to +0.0091 A at high, which he does not comment on.

Key claims

• Ives and Stilwell measured a slight asymmetry in the longitudinal Doppler effect, not the transverse Doppler effect (attributed to Gezari 2009).
• In the 850-2100 km/s range the predictions of the emission theory and the Larmor-Lorentz theory are identical, so the experiment is inconclusive (Faraj; his algebra does not support this).
• Ives and Stilwell's method of finding velocity from eV = m_H c^2 (gamma - 1) is circular (Faraj).

Relevance: The only source in the folder that frames the central experiment as an anomaly in the longitudinal effect; kept in the record and weighed under claim 1 and claim 6, with the attribution and algebra caveats.

Mainstream conflict: Disputes that the experiment confirms time dilation; mainstream reading is that the centre shift is the second-order Doppler effect.

Secondary source by a historian of science with no stake. Lalli says Ives pursued two strategies (a direct criticism of the second postulate; an alternative ether theory) and that 'Larmor-Lorentz theory' meant the Lorentz transformations read as a real contraction of lengths and real slowing of clocks in a stationary ether with absolute simultaneity; Ives later dropped the name but kept the concepts. He compares Ives's ether ('the seat of the pattern of radiant energy received from the fixed stars') with Lorentz's and Larmor's, finds Ives used a simplified optical-only version, and relates it to Gruenbaum's 'doubly amended ether theory', which Lalli says is empirically equivalent to special relativity for optical experiments. Ives saw himself as heir of an 'English school' (Newton, Maxwell, FitzGerald, Larmor, Lorentz) and contrasted 'natural philosophy' with the view that 'the sole object of physical theory is to predict the results of experiment' (Ives 1940, Science 91:83). Ives claimed a truly operational method against Einstein's 'pseudo operational procedure' (Ives 1951, Proc. Am. Phil. Soc. 95:126); Bridgman, in an unpublished 18 August 1959 memo (Harvard archives), called Ives's real-but-indeterminable absolute simultaneity 'a complete negation of the operational attitude which [Ives] correctly adopt[ed] in most other situations' and said Ives handled the one-way velocity of light 'in exactly the old uncritical absolute Newtonian fashion' (Lalli's quotations; the memo is not seen). The 5 October 1951 letter to Elmore Butterfield (Library of Congress, Box 13), as Lalli quotes it: Ives would like to think he was championing 'the classical English school of physical thought ... against the paradox of the Semitic school'; Lalli adds 'Elsewhere, I discuss the implicit anti-Semitism that this sentence seems to imply' (Lalli 2013, HSNS 43:41-104, not read). Only this one sentence of the letter is quoted; its surrounding context is unknown. 'Progressive marginalization' and 'hidden network of allies' are from the abstract of Lalli 2013, not this chapter. Context: Ives's father Frederic Ives; PhD under R. W. Wood; Bell Labs electro-optical director from 1925; NAS 1933; D. C. Miller and Michelson as other ether holders.

Key claims

• Ives's 'Larmor-Lorentz theory' treats the Lorentz transformations as real contraction of lengths and slowing of clocks in a stationary ether.
• For optical experiments Ives's theory is empirically equivalent to special relativity (Lalli, via Gruenbaum and Janssen).
• Bridgman judged Ives's real-but-indeterminable absolute simultaneity incompatible with operationalism (unpublished 1959 memo, as quoted by Lalli).

Relevance: Gives Ives's self-understanding, the Bridgman objection and the 1951 letter in context; direct reading of the chapter replaces the earlier second-hand use through the Investigator's note.

Mainstream conflict: none (a historical account); it records that mainstream physics regarded the ether reading as superseded.

The experimenter's own thesis for a rotor-Moessbauer aether-drift experiment (57Co source and absorber on a magnetically suspended ultracentrifuge rotor), whose limit on a preferred-frame velocity is V < (2.0 +/- 5.2) cm/s, about six orders of magnitude below the Earth's orbital speed (the author calls it 'the most sensitive test of the theory'; his own characterisation). Ives appears in the text on pp.49, 53-54 and 158 and in the references on p.166 (JOSA 28:215; JOSA 38:879): the whole word 'Ives' occurs 8 times, not 27 (the earlier figure came from a case-insensitive count that also matched 'gives'). Ives (1948, JOSA 38:879) showed that a one-way measurement of light speed depends on the clock-synchronisation method but not on the velocity V through the cosmological frame, 'because the rods and clocks ... are equally affected by gamma', so special relativity and the Fitzgerald-Lorentz contraction theory are equivalent in a local frame (Preikschat's short summary; Ives 1948 JOSA 38:879 is now in the manifest as Ives-papers/, digested only (see its entry); QE round 2 section 7). Robertson (1949) and Builder (1958, 1960) extended this; Ruderfer (1960) proposed the rotor experiment, and Ruderfer's 1961 analysis showed both theories predict the same null at first order, so the experiment does not separate special relativity from the contraction theory. Within Robertson's framework Preikschat writes the Ives-Stilwell experiment as the measurement of g0 (time dilation), the Michelson-Morley as g1/g2 and Kennedy-Thorndike as g0/g1; accuracies for the Earth's orbital velocity: Joos g1/g2 1 part in 400, Kennedy-Thorndike g0/g1 1 part in 4 (Jaseja 1 in 10^3), Mandelberg-Witten g0 1 in 20 (Moessbauer experiments improve it by 4.5); this experiment g0 to 3 parts in 10^12 and g0/g1 to 3 parts in 10^4 (quoted, not re-derived). The author concedes the north-direction counting fluctuations exceed the instrumental ones but are 'well within statistics' and that the experiment 'is not sensitive enough to distinguish between real and instrumental effects'; he notes the limit is on velocity relative to the assumed (orbital) frame, and that under Sciama's theory of inertia the relevant velocities would be much larger. He frames the preferred frame as cosmological or Machian, with the classical aether as the historical form.

Key claims

• A rotor Moessbauer experiment limits a preferred-frame velocity of the Earth to (2.0 +/- 5.2) cm/s.
• Ives (1948) showed one-way light-speed measurements are independent of the velocity V through the preferred frame, making special relativity and the contraction theory locally equivalent.
• In Robertson's framework the Ives-Stilwell experiment measures the time-dilation coefficient g0.

Relevance: Links Ives to the Ruderfer line of aether-drift tests (the Ruderfer connection to the Moray archive) and to the modern test frameworks; an experimenter-authored source, to be weighed as such.

Mainstream conflict: Concludes in favour of Einstein's postulates at the stated sensitivity; the neo-Lorentzian view is reported as still discussed (Prokhovnik 1967).

JOSA 42(8):540-543, August 1952. The abstract credits Poincare (1900, 1904) and says Einstein's reasoning, 'questioned by Planck, is defective'; the paper ends 'The relation E = mc2 was not derived by Einstein'. The relation as printed is E = m_M c^2 with m_M the mass of matter. Ives accepts the relation and gives his own derivation. Ohanian's 'Am. J. Phys. 42, 540' citation is Ohanian's slip.

Key claims

• Einstein's 1905 derivation of mass-energy equivalence is defective; the relation was derived by others (Poincare).
• Ives gives his own derivation of E = m_M c^2.

Relevance: Primary source for claim 5, replacing the earlier reliance on Ohanian's summary.

Mainstream conflict: Stachel and Torretti (1982) and Ohanian (2008) reject the circularity charge.

Ives applies Larmor-Lorentz clock-rate change to clocks carried at different velocities. When a platform changes velocity from v1 to v2 a transported clock acquires a setting difference proportional to (v2-v1)T/c, T being the light transit time; he calls this the 'aberration of clocks' and compares it with stellar aberration, both indicating motion through the ether [read: digest p.1-2]. On the clock paradox (two clocks each appear slow to the other, yet both cannot be) he proposes that the ether frame resolves it: the setting change of clock B depends on its motion relative to the ether, not on relative motion with A alone; three special cases give the same maximum change but different actual outcomes, so relative velocity alone cannot predict which clock lags. Whose clock lags can be predicted from stellar-aberration observations of each clock's ether motion [read: digest p.3-5]. Practical detection of clock aberration is called impossible at the precision needed (about 1e-8 s) [read: digest p.2-3]. The usual special-relativity resolution (the travelling clock changes inertial frame) is not discussed in these pages (per round-1 deep read).

Key claims

• Clock setting after transport depends on motion relative to the ether, so relative motion alone does not say which of two clocks lags (the clock paradox).
• A 'clock aberration' proportional to (v2-v1)T/c parallels stellar aberration and indicates motion through the ether.

Relevance: Source for Ives's position on the clock paradox and for claim 2 (ether as reference frame).

Mainstream conflict: Interpretation: special relativity resolves the paradox by the travelling clock's change of inertial frame, a point not discussed in this paper; no predicted measurable difference is claimed.

Sequel to JOSA 27:263. Takes a platform, an observed body and an observed clock in relative motion, with lengths measured by a rod passed at a defined velocity and times by clocks moved apart at a defined velocity, all subject to the FitzGerald-Larmor-Lorentz contraction (1-v^2/c^2)^(1/2), v and c referred to the ether [read: digest p.1]. He derives the transit time of the observed clock between two platform points (eqs. 1-9) and finds three cases: (I) setting-clock velocity W0 near 0, the clock appears slow by (1+Vo^2/c^2)^(1/2)/Vo-type ratio; (II) W0 = Vo, transit times equal; (III) W0 > Vo, measured time less. These describe the observation sequence, not absolute clock rates [read: digest p.2-3]. He then treats apparent lengths from traces made at simultaneous clock readings; the full solution depends on rod and clock velocities and he restricts to a simpler case [read: digest p.3]. The 1937 papers carry misprints (round-1 deep read: eqs. 14, 25, 33, 34, 38-40) that do not change the conclusions.

Key claims

• Apparent lengths and times of a moving system depend on the velocities at which the measuring rods and clocks are transported, and approach the Lorentz-transformation values only as those velocities go to zero.

Relevance: Part of the primary derivation behind claim 3 (Ives's classical-ether route to Lorentz-type transformations).

Mainstream conflict: None in prediction (round-1 deep read: equals special relativity for the same transport procedure; difference is a synchronisation convention).

Huygens-construction chart (example v/c = 0.4) showing why a moving Michelson interferometer gives a path difference from the 'error of reflection' at a moving oblique mirror (aberration angle about v/c). The FitzGerald-Lorentz contraction can be seen as tilting the mirror just enough to remove it [read: digest p.1-2]. With contraction ratio (1-v^2/c^2)^alpha : 1 the equal-arm Michelson-Morley result is satisfied by either the usual form (alpha = 1/2 along the motion only) or a second form (1-v^2/c^2 parallel, (1-v^2/c^2)^(1/2) perpendicular); the Kennedy-Thorndike unequal-arm experiment requires both paths to stay constant and is satisfied only by the second form (Table I) unless clock frequency also changes [read: digest p.2-4]. Kennedy and Thorndike set out to test the Larmor-Lorentz clock-frequency reduction; Ives notes the same result is equally explained by unequal contraction in two directions with no change in time reckoning [read: digest p.3].

Key claims

• Michelson-Morley and Kennedy-Thorndike results can be explained by several combinations of length contraction and clock-rate change; they do not by themselves fix the clock-rate factor.

Relevance: Origin of Ives's argument that a separate optical experiment (the canal-ray Doppler shift) is needed to fix the clock rate; background to claim 1.

Mainstream conflict: None (a re-analysis of null experiments; special relativity gives the standard combination).

Ives writes the contraction and clock-frequency changes with an unknown exponent beta (clock frequency scaled by (1-v^2/c^2)^beta) and notes that Kennedy-Thorndike alone cannot fix beta. He proposes observing the Doppler effect in hydrogen canal rays (displaced lines, end-on and at right angles) to determine it. For beta = 1 the end-on lines show first-order symmetric displacements plus asymmetric second-order terms, from which ether motion could be deduced (Fig. 2, v_s/c = 0.1, v_0/c = 0.2); for beta = 0 (Larmor-Lorentz factor) the modified formula is v' = v(1-v_s^2/c^2)^(1/2)/(1-v_0^2/c^2)^(1/2) x [1-(v_0/c)cos psi]/[1-(v_s/c)cos psi] [read: digest p.1-3]. Per QE round 2 he credits 'Einstein and Ritz' with seeing the crucial role of the canal-ray experiment 30 years earlier [read: qe.md round 2, section 3.12].

Key claims

• The Kennedy-Thorndike result leaves the clock-rate exponent undetermined; the canal-ray Doppler effect can determine it.

Relevance: The proposal that led to the Ives-Stilwell experiment, in Ives's own words (claim 1); the files previously quoted it only via the 1938 paper.

Mainstream conflict: None.

Starting from FitzGerald-Larmor-Lorentz contractions of length and frequency, Ives argues that one invariant can be determined from measurements on a uniformly moving particle using rods and clocks moved at observed finite speeds; he identifies this 'chronotopic interval' (Minkowski's s) with the undistorted dimension of an interferometer that always gives a null result [read: digest p.1]. Three postulates: velocities in an 'invariant framework' are found with unaffected rods and clocks (equivalent to Lorentz's 'stationary' rods and clocks and 'essentially identical with the luminiferous ether, serving as a theoretical tool'); light speed in it is c; moving material rods and clocks are reduced by (1-v^2/c^2)^(1/2) [read: digest p.1-2]. When the clock-transport velocity goes to zero the expressions become the Lorentz transformations, 'a limiting case applying only where clock manipulation velocities are infinitesimal' [read: digest p.4]. Per QE round 2 he tabulates combinations of contraction and clock rate in a gravitational field that all give null interferometer results and says choosing among them needs experiments 'where the gravitational parameter varies', such as near the Sun [read: qe.md round 2, section 3.3 after digest pp.5-8]: an explicit statement by Ives of where his framework is underdetermined.

Key claims

• The chronotopic interval equals the undistorted size of a null-result interferometer.
• The Lorentz transformations are the limiting case of clock-transport velocity going to zero.

Relevance: Links claim 3 (derivation of Lorentz-type transformations) to claim 4 (gravitation): the same interval, with a gravitational parameter, is the Schwarzschild form of JOSA 29:183.

Mainstream conflict: None in prediction; the ether framework is an interpretation.

In Wood's two-toothed-wheel version of Fizeau's experiment (wheels at the ends of a long axle, light sent one way, apparatus rotating and translating), R. W. Wood and F. E. Hackett concluded a 'twist' of the axle beyond simple FitzGerald contraction was needed (Hackett compared it to the Wiedemann effect). Ives argues the twist is not supplementary but follows from the contraction itself when a moving body is constrained to a fixed orientation: a rectangular element of the cylinder becomes a rhomboid (Figs. 3, formulas for B', H', H'' and T), with the contraction plus Larmor-Lorentz clock rate sufficing for all rotating-disk problems [read: digest p.1-3]. For the Ehrenfest paradox (a rotating disk's rim contracts but its radius does not) he proposes that the disk 'dishes' like a saw blade, with a paraboloidal profile for small angular velocity [read: qe.md round 2, section 3.6, lines 197, 234-242].

Key claims

• The Wood-Hackett 'twist' of a rotating, translating axle follows from the FitzGerald contraction alone.
• A rotating disk dishes (paraboloidal profile at small angular velocity) under the FitzGerald contraction.

Relevance: A stated physical deformation of a rotating disk, offered as a consequence of real contraction; one of the claims in claims.md item 7 that goes beyond reinterpretation.

Mainstream conflict: Whether the dishing differs from what special relativity with real elastic stresses predicts is not settled in the documents read; left to the Physicist (QE round 2 section 3.6). The Ehrenfest paradox itself is a long-standing mainstream discussion [memory: high].

The 'artificial Doppler effect' of Belopolski and Galitzin (a rapidly moving mirror with multiple reflections) cannot be expressed as relative motion of source and observer. Ives derives the reflected frequency from wave theory (single and multiple reflections; system moving perpendicular to the mirror at V by double application), asks whether it could detect absolute motion through the ether, and finds that lengths and clock frequencies must contract in the ratio (1-V^2/c^2)^(1/2) for a null result, the same factors as for Michelson-Morley although entering by a different route. Concludes the experiment would give a null result and cites Bateman: the Lorentz transformations can be derived from reflection from a moving mirror [read: digest p.1-3].

Key claims

• A moving-mirror Doppler experiment would give a null result for absolute motion, with the same contraction factors as Michelson-Morley.

Relevance: Another independent route to the same contraction factors (support for claim 3).

Mainstream conflict: None.

Larmor, Levi-Civita, Poynting, Bucherer, Lorentz, Abraham and others treated radiation pressure on moving receivers with disagreeing results, none establishing that irradiation and pressure are both unaffected by motion of the whole source-and-receiver system. Ives takes a moving perfect reflector: reflected energy density E_r = E_i[(c+v)/(c-v)]^2 (eq. 1), energy balance E_i(c-v) - E_r(c+v) = Pv (eq. 2), pressure P = 2E_i(c+v)/(c-v) (eq. 3). From momentum he gets mass equivalent of radiant energy m = E/c^2 (eq. 4), then for a perfect absorber moving toward a point source P = E_a(c+v)/(c-v) [read: digest p.1-2; qe.md round 2 section 3.8]. He states that Lorentz's P = E_i and Poynting-Abraham's P = E_i(1+v/c) imply first-order effects for a co-moving source-absorber system, which his result removes [read: qe.md round 2 section 3.8].

Key claims

• Earlier published formulas for radiation pressure on a moving absorber (Lorentz; Poynting and Abraham) are wrong; the correct pressure on an absorber moving toward a point source is E_a(c+v)/(c-v).
• The mass equivalent of radiant energy is E/c^2, derived classically from radiation pressure.

Relevance: A physical claim that earlier results were wrong, not a reinterpretation; belongs with the 1943-45 conservation-law derivations (claims.md item 7).

Mainstream conflict: Not checked against textbook radiation-pressure results here; left to the Physicist. The mass-energy equivalence for radiation m = E/c^2 is also a 1900 Poincare result [memory: medium; QE round 2 section 3.4].

Extends Compton-type scattering to a general study of a wave packet striking a reflecting particle using classical wave theory and only the Newtonian conservation of mass, momentum and energy. A packet of energy density E emitted in time Delta t has energy E(c Delta t). For a mirror at average velocity f(v) during impact the reflected frequency is nu = nu0(c-v)/(c+v) and the reflected energy E0((c-f(v))/(c+f(v)))^2. Equating pressure to work, he derives a mass equivalent for the packet, then requires conserved (mass + mass equivalent) and momentum, giving f(v) = c^2(g-1)/(g v) with g the mass-increase factor; by successive approximation g approaches 1/(1-v^2/c^2)^(1/2), the Lorentz mass-velocity law. Kinetic energy Mc^2(g-1) equals the net radiant energy flow, so energy is conserved. Ives notes the derivation needs no relativistic assumptions about moving observers or Lorentz contractions [read: digest p.163-166]. Cited as 'the detailed version of Section 2' of the 1945 paper (deep read of Phil. Mag. 36:392).

Key claims

• The Lorentz mass-velocity law follows from classical wave theory and Newtonian conservation of mass (with m = E/c^2 for radiation), momentum and energy.

Relevance: First of the three conservation-law derivations of mass, length and clock-rate changes (claim 3, item 7).

Mainstream conflict: Whether the premise 'mass equivalent E/c^2' already imports relativistic content is an open question for the Mathematician (QE round 2 section 3.4; deep read of the 1945 paper lists three premises that do work).

The absorbing-particle case is harder than the reflecting one because the radiation becomes heat and kinetic energy of the particle's atoms. Ives gives the constituent atoms random thermal velocities in addition to the bulk motion (grouped as forward and backward sets, velocities v+w1 and v-w2, with w0 the common thermal velocity when the particle is brought to rest without heat escaping), requires forward and backward momenta to cancel, and derives total momentum as a function of v, heat energy H and rest mass M0; the kinetic energy contains the factor [1-(v^2/c^2)]^(-1/2) plus a heat term, and conservation of energy equates the packet energy to kinetic energy plus absorbed heat [read: digest p.1-3].

Key claims

• Momentum and kinetic energy of an absorbing particle, including its heat content, follow from Newtonian conservation laws and give the relativistic mass factor.

Relevance: Completes the 1943-45 conservation-law derivation set (claim 3, item 7).

Mainstream conflict: Same open question as the 1943 paper; no experiment named.

Voigt (1887) first suggested a natural clock would change rate on motion, proposing T' = T/(1-v/c), differing from the Larmor-Lorentz-Einstein factor 1/(1-v^2/c^2)^(1/2) later confirmed. Lorentz's transformation carried a velocity-dependent coefficient l; several physicists (Cunningham, Silberstein) held that l is unimportant, which Ives says is true for Michelson-Morley but not for clock rate. Using his graphical interferometer he shows that Voigt's choice, Lorentz's final l = 1, a constant-volume spheroid (period proportional to (1-v^2/c^2)^(3/8)) and a no-change pattern give different clock periods, and Fig. 3 compares these with the canal-ray results, 'which decide unequivocally for l = 1' [read: digest p.1-2]. Lorentz chose l = 1 from mass variation (electromagnetic speculation); Einstein, taking Lorentz's conclusion as a new principle, deduced the contraction factor and predicted the second-order Doppler effect in canal rays as an experimentum crucis sought for about thirty years. Per QE round 2 Ives writes that Voigt offered no physical justification, Lorentz claimed no objective reality for local time, only Einstein 'believed the equations' predictions', and that the clock-rate test is the most conclusive because Michelson-Morley can be explained by entrained ether and mass variation by charge variation [read: digest p.3; qe.md round 2 section 3.11].

Key claims

• The canal-ray experiment decides among Voigt, Lorentz and other clock-rate coefficients in favour of the Larmor-Lorentz-Einstein factor.
• Einstein alone treated the clock-rate prediction as a crucial test.

Relevance: Ives's own fairness to Einstein on the history; relevant to claim 2 and to the reader's picture of Ives (QE round 2 section 3.11).

Mainstream conflict: None.

Argues that the velocity of light should be measured by one-way signals, feasible now with crystal clocks, and distinguishes 'velocity' (distance in the medium over time, both measured with rods and clocks stationary in the medium) from the 'rod-clock-quotient' Q obtained with FitzGerald-Larmor-Lorentz-contracted rods and clocks; the confusion between the two is said to cause obscurity in special-relativity presentations [read: digest p.1-3]. For out-and-back transit on a platform with medium velocity W past it, t1 = D'(1-W^2/c^2)^(1/2)/(c-W) and t2 = D'(1-W^2/c^2)^(1/2)/(c+W); the contractions eliminate W so the rod-clock quotient gives 2D'/t' = c. Q1 for a single stationary clock equals c; Q2 (two stationary clocks) and q (moving clock) are defined separately [read: digest p.2-3]. This is where the term 'rod-clock-quotient' is first defined (the 1953 term used in earlier files was already in use in 1948; QE round 2 section 7 item 5). Preikschat (1964) summarises it as showing one-way light speed depends on the clock-synchronisation method but not on velocity through the cosmological frame, so special relativity and the FitzGerald-Lorentz theory are equivalent in a local frame [read: Preikschat entry].

Key claims

• Only out-and-back light speed has been measured; a one-way speed depends on how distant clocks are set.
• 'Velocity' (in the medium) and the 'rod-clock-quotient' (what contracting rods and clocks measure) must be distinguished.

Relevance: Primary source for Ives's later one-way light and simultaneity argument (claim 3).

Mainstream conflict: Mainstream: one-way light speed is conventional (Reichenbach 1928, Gruenbaum 1963) [memory: high]; the 1949 deep read finds no prediction that differs from special relativity.

Ives argues the invariance seen in two-way light-signal measurement in Michelson-Morley can be extrapolated to one-way measurements, while accepting (with Robb's before-and-after characterisation) that arrival time at a distant point is indeterminate. This leads to Lorentz-type transformations in terms of observable rod and clock readings, including the self-observed velocities of the moved clocks that set distant clocks. He differs from special relativity by holding that equal out-and-back transit times, as a definition of simultaneity, is required by neither the relativity principle nor experiment, since only two-way transit has been measured [read: digest p.1-2]. A hypothetical experiment with moving sources, mirrors and a radiometer is expected to give simultaneous arrival from all sources. The derivation yields F = [q(1+q')^-1 - 1]/q and an isotropic coordinate system in which the contractions are [1-(W^2/c^2)]^(1/2) in every direction; in that system the equations return to the original Lorentz transformations when clock velocities go to zero 'though this limiting case does not correspond to a performable operation'. He chooses the sign from the experimental fact that a slowly transported chronometer shows time close to D'/c. Section VIII: 'No precision laboratory measurements on one-way light signal transit times ... have thus far been performed ... Direct and decisive confirmation is furnished by ... the Ives-Stilwell experiment' [read: digest p.4-7; deep read of JOSA 39:757, section 5].

Key claims

• Two-way invariance extrapolates to one-way signals with the arrival time left indeterminate, giving Lorentz-type transformations in terms of performable operations.
• The Ives-Stilwell experiment is direct confirmation of the extrapolation.

Relevance: Stage in Ives's argument that special relativity's light-signal synchronisation is a convention (claim 3).

Mainstream conflict: The 1949 deep read finds the equations are special relativity in carried-clock coordinates, with no prediction that differs [read: deep read of JOSA 39:757].

Ives's Rumford Medal lecture. Standing waves of light have been his recurring preoccupation since his Johns Hopkins dissertation on Lippmann colour photography (Lippmann photograph of the spectrum April 1907) [read: digest p.11-14]. He reviews Wiener (1890), Rayleigh (1887), Lippmann, Ramon y Cajal's swollen-film microscopy, the Crova photometric method, WWI aerial photography at Langley Field, and Bell Labs picture transmission and television (by 1927 people in New York could see and speak with Washington) [read: digest p.12-19]. The photoelectric part: the selective effect of thin alkali films with polarised light, where at high incidence emission with the electric vector perpendicular to the surface is ten or more times greater than absorption predicts (absorption predicts a factor of two to three), is explained by the standing-wave pattern at the film; the theory 'confirmed to fine detail' for rubidium on glass; the spectral progression of maxima toward the red with atomic weight follows from the optical constants N and K; computed against observed emission for rubidium and cesium (slides 18, 19) agree closely; and the 1933 repeat of Wiener's experiment with a photoelectric film on a quartz wedge confirms multiple maxima and polarisation reversals never seen before [read: digest p.20-30]. The final section (digest part 02, p.32-35) argues against the photon: Lorentz's 1910 coherence objection, Poincare's 'vestibule' (the quantum in the atom, not the radiation), standing waves requiring photons to exist and not exist alternately in loops and nodes, wave mechanics placing h in the atom, Beck and Sommerfeld's bipartition-angle data said to favour the wave picture, and Sommerfeld preferring the corpuscle against that evidence. Ives writes: 'I venture to predict that this will be done, and that the photon will go the way of the "caloric" that Rumford demolished' [read: qe.md round 2 section 3.9, from extracted text lines 2185-2191]. Per QE he concedes 'the difficulty of arriving at an explanation of all optical phenomena, such for instance as the Compton effect' [read: qe.md round 2 section 4, check 6].

Key claims

• Photoelectric emission from thin alkali films is proportional to the energy density of the standing-wave system at the film and follows the wave theory in fine detail.
• The photon concept will be abandoned, as Rumford's work ended the caloric theory (Ives's prediction, 1951).

Relevance: Supplies Ives's own words on the standing-wave photoelectric work (context for Q10 and Moray's interest in standing waves and radiant energy) and a stated position against the photon that must stay in the record (QE round 2 section 5).

Mainstream conflict: Mainstream: the photon is established (photon counting, antibunching, Compton scattering) [memory: high; not verified on the web here]. Ives himself concedes Compton scattering is an open difficulty. The photoelectric standing-wave work is mainstream optics and uncontested in the round-1 reviews.

Lorentz found his transformations by trial; Poincare (1904) formulated the principle of relativity and revised them, and Einstein (1905) elevated the constancy of the light velocity to a principle with an operational procedure assigning c to synchronise distant clocks [read: digest p.1-3]. Ives criticises this as a logical contradiction with 'mystical appeal' lacking support from objective matters of fact, since precise one-way light velocities have never been measured [read: digest p.3]. He proposes a second revision based strictly on performed operations (Bridgman's operational principle): from the independence of light velocity from matter and Poincare's principle, the out-and-back transit time must be invariant, which fixes equal factors for rod lengths and clock frequencies. The revised transformations use only performable measurements; the constant ratio in them 'is not equal to c and should not be called the velocity of light'; and 'abandoning the "principle" of constant light velocity removes the basis for geometric theories claiming fusion of space and time' [read: digest p.4-7]. Footnote 18 (cited in the 1949 deep read): experimental confirmation is Michelson-Morley at one end and Ives-Stilwell at the other.

Key claims

• The constancy of the light velocity is not established experimentally (only out-and-back speed has been measured) and should not be a principle.
• Performable-operation transformations replace the Lorentz transformations; their constant ratio is not c.

Relevance: Late statement of Ives's critique of special relativity's postulates (claim 3).

Mainstream conflict: The 1949 deep read shows equivalence to special relativity in different clock-setting coordinates; no differing prediction named.

Repeats Wiener's 1890 experiment with a thin photoelectric film in place of the photographic film, to test the idea that photoelectric emission from thin alkali films tracks the energy absorbed, i.e. the energy density in the standing-wave system above a reflecting surface. Geometry: light at 60 degrees incidence, polarised parallel (electric vector in the plane of incidence) or perpendicular; the film cannot hang in space, so a quartz wedge is evaporated onto platinum (from a tungsten filament sheathed in quartz beads) and a very thin cesium film is formed on it; the cell is made very long to limit cesium leakage; the wedge shows faint interference bands 4-5 mm apart [read: digest p.1-3]. The authors expect 'only positional and qualitative agreement' and call the result 'very satisfactory' (QE round 2 section 3.10: Ives's own wording in 1933 is more guarded than the 1951 lecture's 'confirmed to fine detail') [read: qe.md round 2 section 3.10].

Key claims

• Photoelectric emission from a thin film lifted above a mirror follows the standing-wave energy density, with the parallel and perpendicular components differing in position.

Relevance: Experimental basis for the standing-wave photoelectric claim in claims.md; formerly 'known only from the bibliography title'.

Mainstream conflict: None stated (optics of thin films).

Photoelectric currents from molten Na, K, Rb, Cs and their alloys at various angles and polarisations. The selective effect (light polarised parallel to the plane of incidence giving far more current than perpendicular) is clearly shown only by liquid Na-K alloy; wavelength maxima vary with surface condition and preparation, contradicting their earlier assignment to specific elements. Elster and Geitel found a parallel/perpendicular ratio of 70 for Na-K at 70 degrees incidence; Pohl and Pringsheim named the sharp maximum near 0.39 micron the 'selective' effect and the gradual rise the 'normal' effect. Potassium distilled directly on an optically flat bulb surface did not show the selective effect (ratio 3-4) [read: digest p.1-3].

Key claims

• The selective photoelectric effect is a property of the film or surface state, not of the element.

Relevance: Background to the standing-wave explanation (peripheral to the claims; context for Q10).

Mainstream conflict: None.

Thin films of Na, K, Rb and Cs deposited spontaneously on clean metal (mostly platinum) in high vacuum all show a striking selective photoelectric effect, previously known only for Na-K alloy. It occurs only at a certain stage of film development: absent for very thin films and again for thick pure-metal layers; the wavelength maxima ascribed to the selective effect in pure alkali metals (rough or colloidal surfaces) are absent in these films. Cells: pyrex bulb with polished platinum plate and tungsten filament, baked to 400 C, mercury diffusion pump with liquid-air trap; metals purified by repeated distillation; currents by D'Arsonval galvanometer [read: digest p.1-4]. Per QE round 2 section 3.10, selective/normal emission ratios were 15-25 against at most about 4 expected from absorption differences [read: qe.md round 2 section 3.10].

Key claims

• Spontaneously formed thin alkali films show a selective photoelectric effect with emission ratios (15-25) far above the absorption ratio (up to about 4).

Relevance: Primary data for the standing-wave photoelectric work that the NAS memoir calls Ives's 'most important contribution' (context for Q10).

Mainstream conflict: None.

Simultaneous photoelectric and thermionic emission from caesium vapour films on a heated platinum ribbon (2-mil, heated by AC; temperature by optical pyrometer; cell walls packed with CO2 snow to reduce thermionic current). Both currents rise to a maximum and then fall to negligible values; their relative magnitudes indicate the thermionic electrons are not due to internal photoelectric excitation [read: digest p.1-6]. Per QE round 2 section 3.10, photoemission peaked 5-10 times above room temperature near 650 K [read: qe.md round 2 section 3.10].

Key claims

• Photoelectric and thermionic emission from caesium films on hot platinum both pass through a maximum; thermionic emission is not internal photoemission.

Relevance: Peripheral background (photoelectric work, context for Q10).

Mainstream conflict: None.

Why photoelectric cells were chosen for the Bell System's commercial picture-transmission system (instant light-to-current conversion, against 1-2 hours for photo-mechanical processing), why alkali-metal cells of the Elster-Geitel type suit it, and how such cells behave: hermetic bulb with low-pressure gas, stopping potential and Einstein's equation eV = h(nu - nu0), gas amplification with an optimum pressure of a few tenths of a mm of mercury, saturation voltages, dependence on angle, polarisation and wavelength; polarised light with the electric vector in the plane of incidence gives 4-8 times greater absorption but photocurrent ratios up to 20-30 to one, especially for liquid Na-K alloy and deposited films [read: digest p.1-9].

Key claims

• The alkali-metal photoelectric cell is the practical light-sensitive element for picture transmission.

Relevance: Ives's commercial photoelectric engineering (peripheral; context for his standing among alternative-science readers and for Q10).

Mainstream conflict: None.

From three stated assumptions (the equivalence of a stationary gravitational field to motion in field-free space; rods and clocks in the field behave like those on the equivalent moving particle, i.e. FitzGerald contraction and Larmor-Lorentz slowing, called A1 in the deep read; invariance of an interferometer's null reading, A2), Ives derives for a small interferometer in a static spherically symmetric Newtonian field ds^2 = c^2 y dT^2 - R^2 dtheta^2 - dR^2/y, with y = 1 - 2GM/(R c^2); he calls it 'the Schwarzschild formula of general relativity'. Deep read: the derivation is correct and gives the exact Schwarzschild line element in Schwarzschild coordinates (not a 1/c^2 approximation), valid for an interferometer small enough that the field is uniform across it [read: deep read, short answer 1; p.2-4]. Predictions, stated in words only: a 'natural clock' runs slower in a field (red shift) and a light wave front passing a large mass changes direction. No numbers are given, but the speed field implied (c y radially, c y^(1/2) sideways) gives a limb deflection of 1.7516 arcsec against 4GM/(c^2 R) = 1.7515 arcsec, the full general-relativity value; dropping radial contraction gives the half value 0.8758 arcsec; Sun red shift 2.12e-6 [calc: deep reader]. Ives also considers the alternative of keeping only equivalence (constant, isotropic light speed): atomic clocks still run slow but stars near the Sun would not be displaced, 'in conflict with observation' [read: p.4]. 'Anisotropy of the ether' means that, measured with Lorentz's 'stationary' rods and clocks, light speed near a mass is c y radially and c y^(1/2) sideways; apparatus-carried rods and clocks measure c in every direction, so the anisotropy adds nothing measurable beyond general relativity [read + calc: deep read]. His stated limits: the formula 'is of purely optical significance', gives no information on dynamics, has 'no non-Euclidean geometry, nor new law of gravitation', and 'calls for anisotropy of the ether due to the presence of a center of gravitational attraction, but makes no suggestion as to how this anisotropy affects other than optical phenomena' [read: p.4-5]. One door left ajar: a sufficiently large and sensitive interferometer 'could be used to detect motion in a gravitational field' if the invariance fails for large sizes, which he expects will not happen (no calculation) [read: p.4]. Footnote 5 quotes Levi-Civita relaying Einstein as knowing no physical meaning for ds; Ives gives it one (the apparatus's own undistorted size). Presented at the OSA meeting, New York, 24 February 1939.

Key claims

• A fixed ether with contraction and clock slowing reproduces the exact Schwarzschild line element for a small interferometer in a static spherical field.
• The gravitational red shift and the bending of light follow from the same assumptions (qualitatively stated by Ives; the full general-relativity deflection follows from his equations).
• An isotropic-light-speed ether is ruled out by the eclipse observations; the anisotropic ether is required.

Relevance: Primary source for claim 4. Replaces the earlier inference from an abstract: the content of the 'Ives gravitation theory' is here, and it coincides with general relativity's optics.

Mainstream conflict: Ives's own words: no new law of gravitation; the anisotropy is of purely optical significance. Mainstream (general relativity) already gives the same optical predictions; in its terms the 'ether anisotropy' is a coordinate effect [read + calc: deep read; memory: high].

Ives derives the mass-velocity law, length and clock-rate changes and velocity addition from Maxwell radiation pressure and moving-mirror reflection, radiation momentum E/c, Doppler compression in the medium, Galilean light times c +/- V and the Newtonian conservation of mass (with mass equivalent E/c^2 for radiation), momentum and energy, in an ether with 'the space and time concepts of Newton and Maxwell retained'. D1 (clock factor), D2 (length along motion) and D3 (across) are fixed by a light-clock argument and come out as (1-V^2/c^2)^(1/2), (1-V^2/c^2)^(1/2) and 1. Received 16 October 1944, Bell Laboratories, 28 September 1944; called 'logically the first of the series' [read: deep read sec. 1-3, 7]. Deep read, no special-relativity formula is put in by hand, so the derivation is not circular; but three premises do work the conservation laws alone do not: (a) the inertia of energy (m = E/c^2 applied to matter), (b) an unexplained step that the extra term 'must drop out' (equivalent to the relativity principle for this process; a counter-example with a non-zero function of V satisfies energy and momentum with w = 0.709 instead of 0.696), and (c) clock universality and isotropy, equivalent to the Michelson-Morley and Kennedy-Thorndike results, which Ives cites [calc + read: deep read sec. 4]. 'The fact that the Lorentz transformations with all their consequences are deducible from these classes the "Special Principle of Relativity" as a superfluous hypothesis': in context a claim about logical economy, not predictions; the principle is moved into the premises. Ives states no measurable difference from special relativity; the 'indeterminancies and impotences' of the restricted theory 'all follow as consequences'; he names aberration and the Michelson-Gale and Sagnac experiments as 'direct proofs' of the motion of matter through the medium (note 11), which special relativity also predicts [read: deep read sec. 5-6]. Note 10 calls light-speed constancy 'a useful tool', 'a paradox' and 'an unsatisfactory foundation'. It is the paper Ruderfer cites for the most direct verification of a Lorentz ether 'independent of relativity'.

Key claims

• The mass-velocity law, length contraction, clock slowing and velocity addition follow from classical radiation physics and the conservation laws in an ether with Newtonian space and time.
• The Special Principle of Relativity is a superfluous hypothesis (a logical-economy claim, not a claim of different predictions).
• Aberration and the Sagnac and Michelson-Gale experiments are direct proofs of motion of matter relative to the light-transmitting medium.

Relevance: Primary derivation for claim 3 and the paper Ruderfer cites; settles that no new force or measurable effect is claimed.

Mainstream conflict: Ives names no experiment that separates his result from special relativity; the mass-energy premise and the 'must drop out' step carry relativistic content (deep read sec. 4). J. S. Bell (1976) argued a similar dynamical reading in mainstream physics [memory: high].

Ives rewrites G. D. Birkhoff's flat-space gravitational equations (Birkhoff died 1944) in ordinary time and ordinary force, and finds that Newton gravity acting on the Lorentzian velocity-dependent mass leaves out a transverse 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.' Newton-Lorentz alone gives only a fraction of Mercury's advance of perihelion (the printed fraction is low-resolution; probably 1/6, [memory: medium], and confirmed as one sixth by the 1948 paper); the added transverse terms change the areal constant from r^2 thetadot = h exp[k/2 - GM/(rc^2)] to h exp[k - 2GM/(rc^2)] and give the full observed advance, 'the key contribution of Birkhoff's theory' [read: deep read, short answers 1-4; p.3]. The terms do no net work (the energy equation is unchanged) and 'an infinite number of such additional terms could be set down', so energy conservation does not select this one [read: p.3]. For Mercury v/c is about 48 km/s over c, v^2/c^2 about 2.6e-8 [calc]. Ives does not endorse Birkhoff as physics: the force was not introduced with physical understanding, 'an independent physical derivation of these transverse forces would be welcome' [read: p.4]. Gauss analogy: the force is 'similar to' the posthumous 'fundamental law' of Gauss for moving electric charges, 'the occurrence of c^2 stems from the idea of the attraction being transmitted with the speed of light' (a suggestion of origin, not a derivation) [read: p.4]. He also rejects Birkhoff's emphasis on local time: 'Physically, forces act on masses' (footnote 5 cites Ives's own 1945 Phil. Mag. paper) [read: p.2-3]. Ruderfer's note describes this paper as 'the proper correction of Newtonian gravitation, which has an infinite speed of propagation, for propagation at the speed of light' [read: qe.md round 2 section 3.2, from Moray pp.271-272].

Key claims

• Rewritten in ordinary force, Birkhoff's equations contain a transverse velocity-dependent force of size v^2/c^2 times the normal part of gravity, which brings Mercury's perihelion advance to the full observed value.
• The transverse force resembles Gauss's velocity-dependent law, with c^2 arising from finite-speed transmission of the attraction.

Relevance: The one place Ives names an extra force; by the 1948 paper's own account this is the same effect as the position-dependent mass factor in GR (see the 1948 entry). Important to the owner's question whether a Lorentz ether adds forces.

Mainstream conflict: A reinterpretation of Birkhoff, not a new proposed force; Birkhoff's theory as a whole was later criticised on other grounds (e.g. its matter model) [memory: medium; web check not done]. Prediction on Mercury equals general relativity's.

Received 10 October 1947, Upper Montclair, NJ. Postulate: a body at rest in the field has 'the dimensions, clock rate, and mass of a body ... moving with the velocity acquired in falling from infinity'. Ives defines gamma = 1 - V^2/c^2 with V the fall velocity from infinity, so gamma = 1 - 2GM/(rc^2) (his gamma is not the SR Lorentz factor). Gravitational mass, eq. (2): m_g = m0/(gamma - rdot^2/(c^2 gamma) - r^2 thetadot^2/c^2)^(1/2) = m0 dt/dtau. Orbit equation, eq. (25): d^2u/dtheta^2 + u(1 - 6G^2M^2b^2/(c^2h^2)) = K-bar, to terms in 1/c^2 [read: page images p.1-3]. Deep read answers: no adjustable constant (the factor 6 comes from algebra); but two steps are Ives's own choices (transverse force F_theta = 0 by conservation laws; splitting the transverse force (11) into the pair (12)-(13) by analogy with Birkhoff), and the final footnote 9 concedes that on the general-relativity relations alone 'there is no ground for favoring (22) and (23) over the simple and symmetrical equations (12) and (13)' [read: p.1-4]. Perihelion: 6 pi GM/(c^2 a(1-e^2)) = 5.019e-7 rad per orbit x 415.2 orbits per century = 42.98 arcsec per century, the full general-relativity value; plain special-relativity mass increase (Sommerfeld's form) gives 7.16 arcsec, 'one sixth the advance' [read: p.4; calc]. Exactness: the three governing integrals (gamma, h, k) are exactly Schwarzschild's; the printed force equation (23) and orbit equation (25) are first order in 1/c^2 ('agreeing to discard terms in 1/c^4'); eq. (22) is exactly GR's angular-momentum law; eq. (23) differs from the exact Schwarzschild radial equation by terms that give no perihelion advance (a constant times GM/r^2 only rescales GM; a term proportional to rdot^2/r^2 integrates to zero over an orbit); a circular-orbit frequency difference of about 2.5e-8 for Mercury is absorbed in the fitted solar GM [calc: deep reader]. The note added in proof claims some relations 'also exact'; the deep reader confirms this for the integrals but not for (23) as printed. Ives's claim: 'The force equations are actually implicit in the general relativity solution'; the paper is 'an alternative solution to the general relativity and Phillips treatments, from the points where these treatments arrive at Eq. (26)'. He keeps absolute space and time and Newtonian force and remarks that general relativity 'is commonly claimed to negate the concept of force' [read: p.1, 4]. Footnote 6 treats the upward push in a true gravitational field as the physically primary quantity. Cites Eddington 1917 (variable-mass force equations), McCrea 1935 (Sommerfeld's one sixth), Phillips 1922, and Ives 1939, 1945, 1947.

Key claims

• The Schwarzschild orbit results (perihelion advance 43 arcsec per century for Mercury) follow from Newtonian force on a mass, clock rate and length that depend on position as well as speed, in absolute space and time.
• The speed-dependent mass of special relativity alone gives only one sixth of Mercury's perihelion advance; the rest comes from the position-dependent factor.
• The 1947 transverse force is equivalent to the factor gamma^(1/2) in the local mass.

Relevance: Primary source for claim 4. Overturns the earlier inference (from the abstract) that the velocity terms were 'chosen to reach 43'.

Mainstream conflict: Ives claims no prediction that differs from general relativity; the deep read finds none at the order he works. Nothing on radar delay, binary pulsars or gravitational waves (all later) [read: whole paper]. The 'one sixth' is a standard special-relativity-only result [memory: high].

Ives says the special-relativity derivation assumes the coefficients of the Lorentz transformation depend only on the relative velocity V of two platforms, which holds only if no clock is ever carried to a distant point to set its time. If a clock is carried at finite speed q, the measured one-way light speed is Q_L = c/(1-ck), where k = (sqrt(1+q^2/c^2) - 1)/q, about q/(2c^2), so Q_L is about c + q/2, contradicting special relativity's 'definition' of one-way speed c. He concludes: 'the assumptions used in the Special Theory of Relativity derivation exclude the use of clocks for setting "times" at a distant point', and the Lorentz transformations are 'an approximation, owing their simplicity of form to the exclusion of clocks in the determination of epoch' [read: p.5; deep read sec. 1-2]. Deep read: his equations (19)-(20) are, term for term, the standard Lorentz transformation written with clock readings that lag Einstein time by k r' (confirmed by SymPy); special relativity predicts exactly his one-way result for a carried clock, and Q_L tends to c as q goes to 0, which Ives accepts as an experimental fact in 1950 but rejects because q = 0 'could not be performed'. So the charge does not survive as an inconsistency; what survives is a true, also-standard statement that synchronisation by a finite-speed carried clock differs from Einstein synchronisation by a small computable amount (conventionality of simultaneity; Reichenbach 1928, Gruenbaum 1963) [read + calc: deep read sec. 3]. Sizes: walking clock 1 m/s over 10 m gives an offset of 5.6e-17 s and a fractional light-speed change 1.7e-9; jet at 250 m/s over 10,000 km gives 14 ns; satellite at 7.8 km/s over 40,000 km gives 1.7 microseconds [calc: deep read sec. 4]. Ives says the setting-clock terms are negligible in practice but he is concerned 'not with practical approximations, but with obtaining rigorous expressions' [read: p.4]. Also derives the standard relativistic aberration formula (18) and velocity composition (23). The paper names no separating experiment; the 1950 and 1951 papers cite only Michelson-Morley and Ives-Stilwell, which special relativity predicts identically [read: deep read sec. 5]. A misprint in k on p.3; asymmetry for signals sent toward the origin (c/(1+ck)) is never stated by Ives (deep reader's calculation).

Key claims

• The special-relativity derivation of the Lorentz transformations is internally inconsistent if clocks may be used to set distant times.
• The Lorentz transformations are an approximation valid when c/(1-ck) cannot be told from c.
• Re-expressed in Einstein-synchronised time, Ives's transformations are exactly the Lorentz transformation for every finite q (deep reader's finding, not Ives's claim).

Relevance: Primary source for the inconsistency charge in claim 3, now checked against its own equations: it is a synchronisation convention, with no differing prediction.

Mainstream conflict: Mainstream: slow-transport synchronisation equals Einstein synchronisation in the limit q -> 0 (Eddington 1924; Mansouri and Sexl 1977) [memory: high for Mansouri-Sexl, medium for Eddington's date]. The deep read finds no inconsistency in special relativity.

Same canal-ray apparatus as 1938 with two aims: a much larger velocity range and a test of factors that might have affected the first results. The 1938 caveat: the largest centre-of-gravity shift was only about half the H-beta doublet separation, so a shift of intensity between the two components 'might conceivably' fake it; 'a shift of at least twice that previously obtained was therefore set as the goal'. Multiple electrode gaps keep each below the sparking limit (about 20,000 V), a first gap of about 5,000 V; total voltages 'up to about 43,000', H2 and H3 particles, coarser faster plates (less precise) [read: p.1-2]. Result: at about 43,000 V for H2 'a shift of over 0.11 A as against the previous maximum of 0.047 A' (Table I: observed 0.113 and 0.1145 A at 42,260 and 42,280 V, computed 0.1098 and 0.1073 A); 'the desired goal of doubling the observed shift was attained, and ... over this entire range the relationship previously observed is maintained'. The implied doublet separation is about 0.09-0.095 A (not stated), so the shift now exceeds it by about 1.2 times [read: deep read sec. 1; calc: lambda0 beta^2/2 for H2+ at 43,000 V is 0.1114 A; at 28,300 V 0.0733 A computed vs 0.0717 A printed]. Observed minus computed for nine H2 rows has mean about +0.003 A (about 3 percent), six above and three at or below; the authors attribute 'the slight tendency ... to be above the theoretical curve' to the mirror being less than 100 percent reflecting [read: p.4-5; calc]. Possible misprint: the H3O computed entry .0012 should be about .0118 from its own delta-lambda [calc]. R. C. Jones (JOSA 29:337, 1939) had suggested that energy imparted to the 'stationary' particles gives them a velocity along the motion, shifting the reference line and mimicking a red shift. A comparison-source densitometer scheme was built; with the mirror in, no central-line displacement within error; with it covered 'a very small displacement appears', about 750 m/s at 14,000 V (Table II: 0.0125 A without mirror, 0.002 and 0.000 A with mirror) [read: p.2-4]. Otting's criticism of the mirror arrangement was tested with a double-ended tube at 28,300 V (observed 0.0744 A against 0.0717 A computed): 'completely disposes of this criticism' (one point, no error bar) [read: p.5-6]. Conclusion: 'establish conclusively that the frequency of light emitted by moving canal rays is altered by the factor (1-v^2/c^2)^(1/2). This agrees with the prediction of Larmor and Lorentz', 'unless such a frequency change should be shown to be the result of other factors than the speed of the particles'. The words ether and relativity do not appear in these six pages [read: p.6]. Received 26 February 1941; Bell Telephone Laboratories (the authors' employer; Ives and Stilwell are the interested parties in the interpretation; Jones and Otting are independent).

Key claims

• At about 43,000 V the second-order Doppler shift of H2 canal rays exceeds the H-beta doublet separation and still agrees with lambda0 v^2/(2c^2) (about 0.11 A observed against about 0.11 A computed).
• Stationary-particle velocity (Jones's objection) is about 750 m/s at 14,000 V without the mirror and not detectable with it.
• The canal-ray frequency is altered by (1-v^2/c^2)^(1/2), as Larmor and Lorentz predicted.

Relevance: Answers the main caveat of the 1938 experiment (claim 1): the high-voltage points exceed the doublet separation. Replaces 'Part II not obtained' in earlier files.

Mainstream conflict: Consistent with special relativity; the paper claims no conflict. Pobedonostsev and Parshin (1989) report deviations at 150-300 keV (see their entry).

Optical interference is simulated with capillary waves on a liquid, the sources now air jets (no bow waves) at 30 Hz, with the medium moving at 11.3 cm/s (v/c = 0.51). Doppler wavelengths appear as lambda/(1+v/c) ahead and lambda/(1-v/c) behind. Two synchronised jets mimic a Fresnel biprism or Lloyd mirror; the pattern is concentric circles or confocal hyperbolae at rest and the same but closer when the medium moves. If distances are scaled by (1-v^2/c^2)^(1/2) and periods lengthened by 1/(1-v^2/c^2)^(1/2) the moving pattern becomes identical to the stationary one: these are the FitzGerald contraction and Larmor-Lorentz clock slowing. In a refracting medium the Fresnel convection coefficient c' = (c/mu - v)/(1 - v/(c mu)) also appears; the three classical effects (contraction, clock rate, Fresnel drag) are the conditions that nullify motion. Standing waves of capillary and gravity waves traveling at different speeds add velocity-dependent spacings [read: digest p.1-8]. Digest caveat (QE round 2 section 4): the factor is printed (1-v/c^2)^(1/2); the correct form is (1-v^2/c^2)^(1/2).

Key claims

• In a classical medium, interference patterns with moving sources are restored to the stationary pattern by FitzGerald length contraction and Larmor-Lorentz clock slowing, with Fresnel drag appearing in a refracting medium.

Relevance: Ives's only experiments on a material medium showing 'Lorentz invariance' for waves; a physical picture of what a Lorentz ether means (claims.md item 7).

Mainstream conflict: None claimed; a classical-wave analogue, not a test of light.

Ripples on a mercury surface made by intermittent air jets; a new method images a linear light source at large angle and reflects it from the mercury to show the wave profile. Parameters: jet frequency 30 cycles/s, relative velocity v = 11.28 cm/s, wave velocity c = 19.20 cm/s, v/c = 0.585 [read: digest p.1-2]. Wave amplitude is the same on both sides of a jet despite the high relative velocity, ascribed to offsetting effects of attenuation from expanding circular waves and Doppler compression or expansion; two jets in line with the medium motion give standing waves with node and loop amplitudes duplicating the stationary case [read: digest p.2-4]. One measured ratio disagrees with the ideal prediction: 0.852 measured against (1-v^2/c^2) = 0.658 predicted, attributed by the authors to capillary-wave dispersion; the digest wrongly gives the exponent as 1/2 (which would predict 0.811) [read: qe.md round 2 section 3.7, section 4 check 7; calc: 1 - 0.585^2 = 0.658].

Key claims

• Ripple-tank interference with a moving medium reproduces the stationary pattern under FitzGerald contraction and Larmor-Lorentz time scaling, within a stated deviation attributed to dispersion.

Relevance: Part of the classical-medium analogue (claims.md item 7); relevant to the owner's question what a Lorentz ether means physically.

Mainstream conflict: None claimed.