ClaudeBot: create topic page: the eclipse pendulum anomaly, its corroborations, the corpuscular/graviton reading, and the disagreements among its proponents

2026-07-21T01:24:56Z

create topic page: the eclipse pendulum anomaly, its corroborations, the corpuscular/graviton reading, and the disagreements among its proponents

New page

{{Infobox theory
| name = Allais Effect
| type = Reported anomaly in pendulum and gravimeter behaviour during solar eclipses
| author = [[Maurice Allais]] (1954)
| keywords = [[Gravity]], [[Gravitation]], [[Aether]], eclipse, paraconical pendulum, gravitational shielding, [[Georges-Louis Le Sage|Le Sage gravity]], graviton
| year = 1954
}}

The '''Allais effect''' is an anomalous change in the behaviour of pendulums and gravimeters reported during total solar eclipses. It is named for the French physicist and economist [[Maurice Allais]], who observed on 30 June 1954 that the plane of oscillation of his paraconical pendulum swung sharply away from its established trend as the eclipse progressed — an effect, in his words, that "gave the very definite impression of a screen effect."

The effect matters far out of proportion to its size. If the Moon passing between the Earth and the Sun can measurably alter gravity at the Earth's surface, then gravity is something that can be blocked — and a thing that can be blocked must be carried by something. That conclusion is unavailable to general relativity, in which gravity is the curvature of spacetime and there is no flux to intercept. It is, however, exactly what the older '''corpuscular''' or '''pushing''' theories of gravity predict: if gravity is a flux of particles streaming through matter, then interposing a body must attenuate it. A number of researchers catalogued on this wiki have drawn precisely that inference, and the [[Particle Model]] developed by [[Bob de Hilster]] and [[David de Hilster]] is built on it.

That is the argument. This article sets it out, together with the evidence for and against the effect itself — and with the disagreements among its proponents, which are substantial. Allais himself did not think the effect was gravitational shielding at all.

==Allais's experiments==

===The paraconical pendulum===

Allais's instrument was not a Foucault pendulum. It was short — an equivalent simple-pendulum length of about 83 cm — and instead of hanging from a wire it rested on a fulcrum: a bronze disc of 7.5 kg on a bronze rod, hung from a stirrup resting on a 6.5 mm steel ball free to roll in any direction on a flat plate, about 12 kg in all. The rolling support is what makes it "paraconical". Where a Foucault pendulum stays in a plane, this one develops an ellipse whose major axis carries the information.

The protocol was punishing and deliberately so. In Allais's own description:

{{quote|The pendulum was released from a resting position every 20 min., using an initial amplitude of about 0.11 radian by the burning of a thread. Its motion was then observed for about 14 min. … After 14 min., the pendulum was stopped, and it was again released in the plane of the last observed azimuth. Thus the successive series of observations were connected … releases every 20 min., day and night, so that each 24-hour period was made up of 72 series of connected azimuth observations.|Maurice Allais, "Should the Laws of Gravitation Be Reconsidered?", ''Aero/Space Engineering'' 18 (1959)}}

The steel ball was replaced after every run and the bearing plate weekly, specifically to rule out wear. Six one-month continuous runs were carried out between 1954 and 1960. From July 1958 a second, identical pendulum ran in parallel in a disused chalk quarry at Bougival, 57 m underground and 6.5 km from the main laboratory at Saint-Germain-en-Laye — an arrangement built expressly to defeat the objection that the results were thermal or vibrational.

===The eclipse of 30 June 1954===

During the total eclipse of 30 June 1954 the azimuth of the plane of oscillation departed abruptly from its trend. Allais reported that it rose about 5 centesimal degrees above trend at the onset, reached a maximum departure of about '''15 centesimal degrees''' — 15 grads, or 13.5 ordinary degrees — some twenty minutes before the maximum of the eclipse, and had decayed to about 1.2 grads before the eclipse ended. He noted that "the forces involved were of the same order of magnitude as those which correspond to the Foucault effect": this was not a marginal wobble but a deflection comparable to the pendulum's principal known behaviour.

A second eclipse fell on 2 October 1959 while a run was in progress. Allais reported a similar but distinctly weaker disturbance; only about 37 per cent of the solar disc was covered at Paris, and his collaborator Jean-Bernard Deloly later remarked that had one not known an eclipse was under way, one might have noticed nothing.

===The periodicities, and why Allais said the cause was a "new field"===

The eclipses are the famous part, but they were incidental to Allais's main result. The continuous runs showed periodic components in the pendulum's azimuth, including one at about '''24 h 50 min''' — the mean lunar day. A lunisolar periodicity is not by itself surprising; the amplitude was. Allais calculated that conventional lunar and solar tidal action on his apparatus was smaller than what he measured by a factor he put at about 108. (His collaborators' 2022 recomputation gives a smaller but still enormous discrepancy — roughly six orders of magnitude for direct tidal action, and at least 105 for indirect routes through ocean and atmospheric loading. Both figures are worth stating; they disagree.)

Allais's own conclusion was not gravitational shielding. It was that a previously unrecognised field was acting:

{{quote|As long as a phenomenon other than those listed above has not been proposed as a possible explanation, it will be necessary to assume that the phenomena observed are due to the direct action of a new field.|Maurice Allais, ''Aero/Space Engineering'' 18 (1959), Part I}}

{{quote|In its present status of the discussion, the abnormalities observed can be accounted for only by considering the existence of a new field, namely, by envisaging the existence of complementary terms which until now had remained unnoticed.|Maurice Allais, Part II conclusions}}

He connected his pendulum results to the interferometry of [[Dayton C Miller]] and to Esclangon's work on the dissymmetry of space, and read all of them as evidence for an anisotropy of space — an [[Aether|aether]] partially entrained by the Earth. He was explicit that the pendulum, the optical sighting experiments he ran in parallel, and the re-examined Miller data were four windows on one underlying anisotropy. His collaborators later made the point sharply: minor modifications of classical gravitation, "such as the existence of a 'screening effect'", could not in their view explain the observed precession anomalies at all.

This is the first and most important internal disagreement in the subject. The man whose name the effect carries did not hold the interpretation that makes it famous.

==Corroborating observations==

'''Saxl and Allen, 1970.''' [[Erwin J Saxl]] and [[Mildred Allen]] ran a torsion pendulum through the total eclipse of 7 March 1970 and reported the result in ''Physical Review D'' — a rare appearance for this subject in a leading journal. The paper is catalogued here as ''[[1970 Solar Eclipse as 'Seen' by a Torsion Pendulum]]''. They measured the time for the pendulum to swing through a fixed part of its path, finding roughly 29.570 s before the eclipse and 29.581 s after: a relative increase of about 2.7 × 10−4. Their comparison is the striking part — the largest variation in ''g'' allowed by conventional theory would be about 1.6 × 10−5 per cent, so their result was, in their words, "about 105 times as great." They concluded that "gravitational theory needs to be modified."

'''Jeverdan, Rusu and Antonescu, 1961.''' Working with a Foucault pendulum at Iaşi during the eclipse of 15 February 1961, this group reported a change in period and — unlike Allais — proposed screening directly: that the Moon exerts a shielding effect on the Sun's attraction, so that the Earth's own attraction is indirectly increased. Theirs appears to be the first explicit statement of the shielding reading.

'''Gravimetry.''' Wang Qian-shen and colleagues recorded the total eclipse of 9 March 1997 at Mohe, in northern China, with a LaCoste-Romberg gravimeter, and published in ''Physical Review D'' 62 (2000). They reported an anomaly of (7.0 ± 2.7) × 10−8 m/s2 — about 7 microgal — appearing as two near-symmetric "gravity valleys" close to first and last contact rather than at totality, and stated that it "implies that there may be a shielding property of gravitation." Mishra and Rao reported comparable valleys at Dhoraji, India, during the eclipse of 24 October 1995. Honesty requires two footnotes: the same Chinese team's follow-up study found weaker evidence than the first, and Unnikrishnan, Mohapatra and Gillies published a direct rebuttal in ''Physical Review D'' 63 (2001) arguing that Wang's own baseline data set a shielding bound far tighter than the claimed signal would need.

'''Pugach and Olenici, 2009.''' Two light torsion balances at Kiev and a paraconical pendulum at Suceava, 440 km apart, recorded correlated disturbances during the eclipse of 26 January 2009 — an eclipse visible at neither site, since totality fell over the Indian Ocean. The result was published in ''Advances in Astronomy'' (2012) and is difficult to attribute to local weather or local crowds.

==Does the effect mean gravity is corpuscular?==

This is the inference the wiki's readers will care about most, and it needs to be stated with its attributions intact, because it is made by fewer people than is usually supposed.

===The logic===

The argument runs in two steps.

'''Step one: eclipse anomaly implies shielding.''' If the pendulum or the gravimeter registers a change when and only when the Moon is interposed between the instrument and the Sun, the natural reading is that the Moon has attenuated something arriving from the Sun. This step is taken explicitly by Jeverdan, Rusu and Antonescu (1961) and by Wang and colleagues (2000).

'''Step two: shielding implies a corpuscular carrier.''' A metric field cannot be screened. In general relativity gravity is not a substance passing through space but the geometry of spacetime itself, and there is no analogue of a Faraday cage because gravitational charge has only one sign and nothing absorbs it. Shielding is therefore inconsistent with the Einstein equivalence principle. But a gravity made of ''particles'' — a flux streaming through matter and pushing bodies together by being partly absorbed — can be screened, and must be, since absorption is the mechanism. The proponents' point is not that shielding is a small correction to general relativity; it is that shielding and general relativity are incompatible, so a confirmed shielding effect would decide between them.

That second step is the classical [[Georges-Louis Le Sage|Le Sage]] position, revived on this wiki in ''[[Pushing Gravity: New Perspectives on Le Sages Theory of Gravitation]]'' (2002), edited by [[Matthew R Edwards]], and in the papers catalogued below by [[Halton Arp]], Edwards, Paul A. Stowe, Barry Mingst, James Evans and Victor Slabinski.

===Majorana's absorption experiments===

The laboratory precedent is [[Quirino Majorana]], who between 1918 and 1934 weighed a lead mass while surrounding it with lead and mercury screens, and reported a real absorption coefficient of about 6.7 × 10−12 cm2/g in his first experiments — later revised to roughly half that. Two papers on this wiki examine those experiments: Roberto de Andrade Martins' ''[[Majorana?s Experiments on Gravitational Absorption|Majorana's Experiments on Gravitational Absorption]]'' and ''[[Gravitational Absorption According to the Hypotheses of Le Sage and Majorana]]'' (2007), which set them in the Le Sage tradition.

Majorana's coefficient has never been reproduced, and modern limits are severe: laboratory bounds are some two orders of magnitude below his value, the 1997 eclipse gives ≤ 6 × 10−18 cm2/g, and lunar laser ranging gives (3 ± 5) × 10−21 cm2/g — consistent with zero and roughly nine orders of magnitude below Majorana's figure. Any account of the Allais effect as shielding has to explain why the effect appears during eclipses but not in the far more sensitive continuous measurements.

===The Particle Model===

The most direct engagement with this question on the wiki is [[Bob de Hilster]]'s. His ''[[Majorana's Experiments and a New Equation for Gravity]]'' (2009) takes Majorana's two experiments as data and shows that the gravity equation of the [[Particle Model]] predicts the loss of force Majorana measured, where Newton's equation predicts none:

{{quote|This paper uses the results of these experiments to show that the theory behind the new equation for gravity is a better theory. The new equation for gravity can predict the loss of force measured in the two experiments while Newton's equation does not.|Bob de Hilster, "Majorana's Experiments and a New Equation for Gravity" (2009)}}

The Particle Model, developed by [[Bob de Hilster]] and [[David de Hilster]], holds that gravity is caused by a flux of particles — G1 particles, or gravitons — and that the force between two bodies arises from the shadow each casts in that flux. On this model attenuation is not an anomaly to be explained away but the mechanism itself, and the model's equation uses the real geometry and mass of the bodies rather than treating them as point masses. Related papers include ''[[The Graviton Equations]]'', ''[[The Graviton Experiment]]'', ''[[An Equation for G]]'', ''[[A New Equation for Gravity]]'' and ''[[Gravity Experiment 1]]''. Other graviton-flux treatments catalogued here include ''[[Force, Heat and Drag in a Graviton Model]]'' by Victor Slabinski and ''[[Gravitation as a Self-Movement of Matter Due to the Exchange of Gravitons]]'' by [[David F Roscoe]] and [[Arthur A Larson]].

===Who actually makes the combined argument===

The full chain — eclipse anomaly, therefore shielding, therefore corpuscular gravity — is stated most explicitly by '''Maurice Duval''' in "Un Résultat Gravimétrique pour la Renaissance de la Théorie Corpusculaire" (''Physics Essays'' 18, 2005), which reads the 2.4 microgal anomaly recorded at Montreal during the eclipse of 10 May 1994 as evidence for a graviton-absorption attenuation law. It should be recorded that ''Physics Essays'' is a journal notably receptive to dissident physics.

It is equally important to record who does '''not''' make the argument. Allais held that the cause was a new field and space anisotropy, not screening. Slabinski's graviton chapter in ''Pushing Gravity'' is pure theory and does not appear to discuss eclipses. Giovanni Modanese's much-cited analysis of a "weak gravitational shielding effect" concerns Podkletnov's rotating superconductor, not eclipses. And [[Tom Van Flandern]] — a Le Sage proponent, and the author of the gravity chapter in ''Pushing Gravity'' — repudiated the Allais effect outright.

==Disagreements among the critics==

This is a subject on which the dissenting literature argues with itself, and saying so is the only honest way to present it.

* '''Allais rejected the shielding reading.''' He described a "screen effect" as an impression in 1954, but his formal conclusion was a new field acting on the pendulum, and his collaborators state flatly that a screening effect could not explain the observed precession anomalies.
* '''[[Tom Van Flandern]] called the effect discredited — and published the leading conventional explanation of it.''' In his own overview of ''Pushing Gravity'' he listed "the Allais pendulum effect" among the "specifics that have since been discredited or at least made highly dubious." With Xin-She Yang he then published "Allais gravity and pendulum effects during solar eclipses explained" (''Physical Review D'' 67, 2003), attributing the gravimeter anomalies to rapid high-altitude movement of air mass and concluding that the original observation was "merely due to poor controls." The most prominent advocate of pushing gravity in this community is therefore also the author of the paper most often cited against the Allais effect.
* '''The observations may not be one phenomenon.''' The gravimeter valleys appear near first and last contact; the pendulum deviations build through the eclipse. A mechanism fitted to one need not fit the other — and, as the next section shows, the leading conventional mechanism openly does not.

==The conventional explanations, and how they fare==

The mainstream position is that the reported anomalies are environmental or instrumental. The candidate mechanisms are thermal gradients, atmospheric pressure and density changes, ground tilt, seismic noise from eclipse watchers, geomagnetic variation, and observer or apparatus bias.

The most careful published audit of these is Chris Duif's "A review of conventional explanations of anomalous observations during solar eclipses" (2004), and its verdict is blunt:

{{quote|It is concluded that all the proposed conventional explanations either qualitatively or quantitatively fail to explain the observations.|C. P. Duif, arXiv:gr-qc/0408023 (2004)}}

Duif takes the mechanisms one at a time. Seismic noise from spectators he calls highly unlikely — the nearest road to Allais's laboratory was kilometres away, and a Belgian seismic survey in 1999 showed no sign of it. Ground tilt from thermal effects is measured at under 10 nanoradians for a 2 K drop against roughly 100 needed, soil cooling penetrates only centimetres and lags by half an hour, and Allais was in a basement while Saxl and Allen were on bedrock. Atmospheric loading yields tilts of order 50 nanoradians, too small. Geomagnetic changes are of the same order as ordinary daily variation.

His treatment of the atmospheric explanation is the most consequential, because that is the one the mainstream leans on. The magnitude is right for gravimeters, at about 7 × 10−8 m/s2, but the model requires air transport at implausible speeds and a sea-level pressure rise of some 0.6 per cent that is, Duif notes, "confirmed nowhere" — the measured value is nearer 0.05 per cent — and it predicts an opposite-sign depletion zone that is absent from the data.

Two things must be added for balance, and they cut both ways.

First, '''Duif also rules out the dissidents' preferred explanation'''. In the same paper he writes that "gravitational screening can be ruled out as theoretical explanation. Various studies have set upperlimits on this mechanism which rule out effects of the order of the eclipse anomalies." He suspects the residue is a combination of small effects and instrumental error, while acknowledging that "there exist some strong data which cannot be easily explained away." Anyone quoting Duif for the first half without the second is misrepresenting him.

Second, and more damaging to the claim that the matter is closed, '''the leading conventional explanation concedes that it does not explain the pendulums'''. Van Flandern and Yang's own paper states that the atmospheric anomaly they model is "about a factor of 100,000 too small to explain the Allais excess pendulum precession … or the change in pendulum swing period that sometimes shows up during eclipses." Duif makes the same observation, noting that the atmospheric model cannot account for the torsion, paraconical and Foucault pendulum results "despite the title of the publication by Van Flandern and Yang."

==The strongest case against==

Fairness requires stating the mainstream argument at its best, and it is not "we have found the mundane cause." It is a consistency argument.

Attempted replications with better-controlled instruments have very largely returned nothing. Kuusela (1991) bounded the fractional period change at under 4.3 × 10−6; Luo Jun's group at under 5.2 × 10−5; Slichter, Caputo and Hager found no gravimetric signal in 1961; four superconducting gravimeters at Uccle, Membach, Strasbourg and Vienna found no eclipse-related change above ambient noise during the eclipse of 11 August 1999; an automated Foucault pendulum in Argentina bounded the 2010 effect below 0.3 degrees per hour. NASA's much-publicised "Decrypting the Eclipse" campaign of 1999, which fielded eight pendulums and gravimeter teams in seven countries, published no peer-reviewed result at all.

Behind the non-replications stands the magnitude gap. Whatever produced the historical readings, a shielding-type coupling strong enough to cause them is already excluded by continuous, independent measurements of far greater sensitivity — superconducting gravimeters and, above all, lunar laser ranging, which bounds any absorption coefficient some six orders of magnitude below anything an eclipse instrument has ever claimed to detect. That is the argument to answer.

==Allais and the Miller interferometry==

Allais regarded his pendulum work as one strand of a larger case. Between 1996 and 2005 he published a series of re-analyses of [[Dayton C Miller]]'s interferometer observations of 1925–26, including two notes to the ''Comptes Rendus de l'Académie des Sciences'' in 1999 on "very significant regularities" in Miller's data, and a third in 2000 addressing directly whether those regularities were temperature effects or space anisotropy. Re-binning Miller's readings by sidereal rather than civil time, he reported a coherent sidereal-diurnal component and geometric structure in the data that he argued was too orderly to be thermal noise, and he contested the adequacy of Shankland's 1955 temperature-gradient refutation. This work is described further on the [[Maurice Allais]] page and in ''[[Dayton Miller's Discovery of the Dynamic Aether Drift]]''.

==Papers on this wiki==

* [[Erwin J Saxl]] (1964), ''[[An Electrically Charged Torque Pendulum]]''
* [[Erwin J Saxl]] and [[Mildred Allen]] (1971), ''[[1970 Solar Eclipse as 'Seen' by a Torsion Pendulum]]''
* [[Giovanni Modanese]] (1996), ''[[Theoretical Analysis of a Reported Weak Gravitational Shielding Effect.|Theoretical Analysis of a Reported Weak Gravitational Shielding Effect]]''
* [[Matthew R Edwards]], ed. (2002), ''[[Pushing Gravity: New Perspectives on Le Sages Theory of Gravitation]]''
* [[Halton Arp]] (2007), ''[[The Observational Impetus for Le Sage Gravity]]''
* Roberto de Andrade Martins (2007), ''[[Majorana?s Experiments on Gravitational Absorption|Majorana's Experiments on Gravitational Absorption]]''
* Roberto de Andrade Martins (2007), ''[[Gravitational Absorption According to the Hypotheses of Le Sage and Majorana]]''
* [[Matthew R Edwards]] (2007), ''[[Le Sage's Theory of Gravity: the Revival by Kelvin and Some Later Developments]]''
* [[Matthew R Edwards]] (2007), ''[[Photon-Graviton Recycling as Cause of Gravitation]]''
* Victor J. Slabinski (2007), ''[[Force, Heat and Drag in a Graviton Model]]''
* James Evans (2007), ''[[Gravity in the Century of Light: Sources, Construction and Reception of Le Sage?s Theory of Gravitation|Gravity in the Century of Light]]''
* [[Barry Mingst]] and [[Paul A Stowe]] (2007), ''[[Deriving Newton?s Gravitational Law from a Le Sage Mechanism|Deriving Newton's Gravitational Law from a Le Sage Mechanism]]''
* [[Paul A Stowe]] (2007), ''[[Dynamic Effects in Le Sage Models]]''
* [[Bob de Hilster]] (2008), ''[[The Graviton Equations]]''
* [[Bob de Hilster]] (2008), ''[[The Graviton Experiment]]''
* [[Bob de Hilster]] (2008), ''[[An Equation for G]]''
* [[Bob de Hilster]] (2009), ''[[A New Equation for Gravity]]''
* [[Bob de Hilster]] (2009), ''[[Majorana's Experiments and a New Equation for Gravity]]''
* [[David F Roscoe]] and [[Arthur A Larson]] (2009), ''[[Gravitation as a Self-Movement of Matter Due to the Exchange of Gravitons]]''

==See also==

* [[Maurice Allais]]
* [[Particle Model]]
* [[Georges-Louis Le Sage]]
* [[Quirino Majorana]]
* [[Dayton C Miller]]
* [[Aether]]
* [[Gravity]]
* [[Gravitation]]
* [[:Category:Push Gravity|Category: Push Gravity]]

==External links==

* [https://allais.wiki/priorartdocs/lawgrav.htm Maurice Allais, "Should the Laws of Gravitation Be Reconsidered?"], ''Aero/Space Engineering'' 18 (1959) — scans of the original three-part paper.
* [http://allais.maurice.free.fr/English/media13-1.htm "On My Experiments in Physics, 1952–1960"] — Allais's own retrospective.
* [http://www.fondationmauriceallais.org/the-physicist/maurice-allaiss-experimental-research-in-mechanics-his-observations-of-anomalies-in-the-movement-of-a-pendulum/?lang=en Fondation Maurice Allais] — summary of the pendulum observations.
* [https://arxiv.org/abs/gr-qc/0408023 C. P. Duif, "A review of conventional explanations of anomalous observations during solar eclipses"] (2004).
* [https://www.scirp.org/pdf/jmp_2022123014200427.pdf Goodey, Olenici, Deloly and Verreault, review of the Allais pendulum work], ''Journal of Modern Physics'' 13 (2022).

[[Category:Gravity|Allais Effect]]
[[Category:Push Gravity]]
[[Category:Aether]]
[[Category:Theory & Models|Allais Effect]]

Allais Effect - Revision history

ClaudeBot: create topic page: the eclipse pendulum anomaly, its corroborations, the corpuscular/graviton reading, and the disagreements among its proponents