arXiv:physics/0205065v1 [physics.gen-ph] 23 May 2002

MICHELSON-MORLEY EXPERIMENTS
REVISITED and the COSMIC BACKGROUND
RADIATION PREFERRED FRAME
Reginald T. Cahill and Kirsty Kitto School of Chemistry, Physics and Earth Sciences
Flinders University GPO Box 2100, Adelaide 5001, Australia
(Reg.Cahill@ﬂinders.edu.au) (Submitted to Nature)
Abstract
We report a simple re-analysis of the old results (1887) from the Michelson-Morley interferometer experiment that was designed to detect absolute motion. We build upon a recent (1998) re-analysis of the original data by Mu´nera, which revealed a small but signiﬁcant eﬀect after allowing for several systematic errors in the original analysis. The further re-analysis here reveals that a genuine eﬀect of absolute motion is expected, in what is essentially a quantum interference experiment, but only if the photons travel in the interferometer at speeds V < c. This is the case if the interferometer operates in a dielectric, such as air, as was the case, incidently, of the Michelson-Morley experiment. The re-analysis here of the Michelson-Morley experimental data, correcting for the refractive index eﬀect of the air, reveals an absolute speed of the Earth of v = 359±54 km/s, which is in excellent agreement with the speed of v = 365±18 km/s determined from the dipole ﬁt, in 1991, to the NASA COBE satellite Cosmic Background Radiation (CBR) observations. Other experiments where the interferometers operated in air (Miller 1925,1933) or helium (Illingworth 1927) give similar results when re-analysed. These experimental results refute Einstein’s assertion that absolute motion through space has no meaning. This re-analysis was motivated by developments in a new information-theoretic modelling of reality, known as Process Physics1, in which the Einstein Special and General Relativity formalisms arise as consequences of an emergent quantum-foam explanation for space. This amounts to an absolute and preferred frame of reference, in what is a uniﬁed quantum theory of space, gravity and matter. So Einstein’s theoretical phenomenology survives, and is now explained, even though he used a false premise in his derivation, but his spacetime construct is now seen to have no ontological signiﬁcance. We predict that new interferometer experiments, operating in dielectric mode, will reveal an inward velocity component of magnitude 11.2km/s due to the quantum-foam ﬂow caused by the matter of the Earth. This is a fundamental test of the quantum theory of gravity that has emerged from Process Physics.
Key words: Michelson-Morley interferometer, Cosmic Background Radiation (CBR), COBE, preferred frame, quantum foam, quantum gravity, process physics, aether.
PACS: 03.30.+p, 04.80.-y, 03.65.-w, 04.60.-m
1Process Physics Web Page: http://www.socpes.ﬂinders.edu.au/people/rcahill/processphysics.html
1

One key unchallenged folklore in physics is that the Michelson interferometer laboratory experiment of 1881 [1], and repeated by Michelson and Morley in 1987 [2], that were designed to detect absolute motion, gave a null result, vindicating Einstein’s assumption that absolute motion (motion relative to space itself) has no meaning; it is in principle not detectable in a laboratory situation. Motion of objects is always relative to other objects, according to Einstein. Using this assumption Einstein went on to construct the Special and General Theory of Relativity, which uses the notion of spacetime to avoid any notion of absolute space. Of course Einstein’s formalism has been abundantly conﬁrmed both by the extensive use of the special theory in particle physics experiments and theory, and by the general theory in various experimental and observational situations. Nevertheless we report here experimental evidence that absolute motion is detectable in laboratory experiments, such as those done by Michelson and Morley and others, but that this requires a re-analysis of the operation of their interferometer, as reported herein. This analysis leads to a speed which agrees with that found from the NASA COBE satellite observations on analysing the dipole anisotropy of the Cosmic Background Radiation. These two speed observations, 104 years apart, agree within experimental errors. Together these results show that absolute motion has been detected. New interferometer experiments are needed to conﬁrm that the direction of that motion is the same as the direction discovered by the COBE mission. These results are profoundly signiﬁcant to our understanding of reality. It follows from recent work that these experimental outcomes will not be in conﬂict with the Einstein phenomenology, but require a major re-assessment of what that phenomenology describes [3].
Recent developments in the new information-theoretic modelling of reality, known as Process Physics [3, 4], indicate that the Einstein Special and General Relativity formalisms arise as a consequence of an emergent quantum-foam explanation for space, but with this quantum foam amounting to an absolute and preferred frame of reference, in what is a uniﬁed quantum theory of space, gravity and matter. In [3] it was shown how the general theory account of gravity arises from an amalgam of two distinct quantum foam eﬀects, one being the eﬀective diﬀusion of the quantum foam towards ‘matter’ that acts as a ‘sink’, together with the classical measurement protocol based on the apparent invariance of the speed of light, which uses the radar method to retrospectively assign spacetime coordinates to distant events, that is, spacetime is in fact a historical record of reality, and not reality itself. Here the invariance of the speed of light is caused by the genuine dynamical eﬀect of the quantum foam on moving rods and clocks, as suggested long ago by Fitzgerald and Lorentz to explain the, here now disputed, null result of the Michelson-Morley experiment. Observers adopting this phenomenological property of light within the classical measurement protocol will ﬁnd, according to Process Physics, that the formalism of the General Theory of Relativity describes their historical records which have the form of a spacetime geometrical construct, but that this spacetime construct has no ontological signiﬁcance. In this construct the modelling of time by geometry is entirely appropriate; it is essentially a pagination of the historical record. The resolution of the confusion between the geometrical modelling of time in historical records and time as an actual process has now been achieved within the new Process Physics. The older but still current Non-Process Physics modelling of reality, which has been with us since Galileo and Newton introduced the geometrical modelling of time, is now superseded.
An implication of these developments in Process Physics is that it should be possible in principle to overcome the classical measurement protocol eﬀects by using the long-known nonlocality of quantum processes which should be sensitive to absolute motion through this quantum foam. Indeed ever since the Aspect [5] experiment showed a violation of the Bell inequalities, in the context of the Einstein-Podolsky-Rosen (EPR) nonlocal eﬀects it has been understood that quantum
2

collapse events caused by quantum detectors appear to require a preferred frame, and so appear

to be in conﬂict with Einstein’s assumption. The work of Hardy [6] and Percival [7] suggested

that ‘double-EPR’ experiments could reveal an absolute frame, and hence absolute motion, though

such experiments for this purpose would be extremely diﬃcult. However the old Michelson-Morley

interferometer experiment is actually a quantum interferometer, that is, in principle it could be

done with one photon at a time, and we inquired why it is believed that it was unable to detect a

preferred frame, and so absolute motion. In fact it can do so.

As described in Fig.1 the beamsplitter/mirror A sends a photon ψ(t) into a superposition

ψ(t) = ψ1(t) + ψ2(t), with each component travelling in diﬀerent arms of the interferometer, until

they are recombined in the quantum detector which results in a localisation process, and one spot

in the detector is produced. Repeating with many photons, reveals that the interference between

ψ1 and ψ2 at the detector results in fringes. Before the quantum theory (and in particular before

the new quantum foam physics) Michelson designed the interferometer with the idea that any

motion through absolute space (then called the luminiferous aether) would result in diﬀerent path

lengths for light waves in the two arms, which would show as a shift in the fringe. Of course as an experimental expediency one rotates the apparatus through 90o so that the roles of the two

arms are interchanged, this rotation then should result in a shift of the fringes, which is an eﬀect

more easily observed. However Michelson and Morley reported a null result, despite a small eﬀect

actually being seen. Fitzgerald and then Lorentz then oﬀered an explanation for the null result,

namely that the failure to get an eﬀect was caused by the actual contraction of the arm moving

lengthwise through the absolute space, as we show below. However long ago it was decided by

physicists that the more elaborate Einstein explanation in terms of spacetime transformations was

superior to this dynamical explanation.

In reviewing the operation of the Michelson-Morley interferometer (see below) it was noticed

that the Fitzgerald-Lorentz contraction explanation only implies a null eﬀect if the experiment

is performed in vacuum. In air, in which photons travel slightly slower than in vacuum, there

should be a small fringe shift eﬀect when the apparatus is rotated, even after taking account of the

Fitzgerald-Lorentz contraction. This appears to have gone unnoticed. As well Mu´nera [8] in 1998

has corrected the original results for various systematic errors, and shown that these interferometer

experiments do indeed reveal small but signiﬁcant non-null results, but as expected now, only when

they are operated in a dielectric.

We have applied the elementary correction required for the eﬀects of the air, or in one case

helium, to various interferometer results as corrected by Mu´nera for other systematic errors. The

correction is in fact large, being some two orders of magnitude. Applied to one set of data the

Michelson-Morley experiment now results in a speed of v = 359 ± 54 km/s. Even more amazing is

the excellent agreement of this speed with the speed of v = 365 ± 18 km/s determined from the

dipole ﬁt, in 1991, to the NASA COBE satellite Cosmic Background Radiation observations [9].

We now indicate the proper understanding of the operation of the Michelson-Morley interfer-

ometers when operated in a dielectric, and then report the corrected results, as shown in Fig.2.

The two arms are constructed to have the same lengths when they are physically parallel to each

other. For convenience assume that the value L of this length refers to the lengths when at rest in

the quantum foam. The Fitzgerald-Lorentz eﬀect is that the arm AB parallel to the direction of

motion is shortened to

L

=L

1

−

v2 c2

(1)

by motion through the quantum foam.

3

C ❄

L

(a)

✻

✲

✲✛

A ❄D L B

(b) ✲

✄✄✄✄✗✄✄✲✄✄✄❈❈❈❈C❈❈❲❈❈❈❈

✲

v✲ ✛

A1

A2 B

Figure 1: Schematic diagrams of the Micheslon Interferometer, with beamsplitter/mirror at A and mirrors at B and C, on equal length arms when parallel, from A. D is a quantum detector (not drawn in (b)) that causes localisation of the photon state by a collapse process. In (a) the interferometer is at rest in the quantum foam. In (b) the interferometer is moving with speed v relative to the quantum foam in the direction indicated. Interference fringes are observed at the quantum detector D. If the interferometer is rotated in the plane through 90o, the roles of arms AC and AB are interchanged, and during the rotation shifts of the fringes are seen in the case of absolute motion, but only if the apparatus operates in a dielectric. By counting fringe changes the speed v may be determined.

Following Fig.(1), we consider the case when the apparatus is moving at speed v through the

quantum foam, and that the photon states travel at speed V = c/n relative to the quantum foam,

where n is the refractive index of the gas and c is the speed of light, in vacuum, relative to the

quantum foam. Let the time taken for ψ1 to travel from A → B be tAB and that from B → A be

tBA. In moving from the beamsplitter at A to B, the photon state ψ1 must travel an extra distance

because the mirror B travels a distance vtAB in this time, thus the total distance that must be

traversed is

V tAB = L + vtAB.

(2)

Similarly, on returning from B to A, the photon state ψ1 must travel the distance

V tBA = L − vtBA.

(3)

Hence the total time tABA taken for ψ1 to travel from A → B → A is given by

L

L

tABA = tAB + tBA = V − v + V + v

(4)

L (V + v) + L (V − v)

=

V 2 − v2

(5)

=

2LV V

2

1 −

− v2

v2 c2

.

(6)

Now, assuming that the time taken for the photon state ψ2 to travel from A → C is tAC, but that the apparatus travels a distance vtAC in that time, we can use the Pythagoras theorem:

(V tAC )2 = L2 + (vtAC )2

(7)

which gives

tAC

=

√L , V 2 − v2

(8)

and including the return trip (C → A, tCA = tAC , tACA = tAC + tCA) results in

tACA

=

√ 2L , V 2 − v2

(9)

4

giving ﬁnally for the time diﬀerence for the two arms

∆t

=

2LV

1

−

v2 c2

V 2 − v2

−

√ 2L . V 2 − v2

(10)

Now trivially ∆t = 0 if v = 0, but also ∆t = 0 when v = 0 but only if V = c. This then would result in a null result on rotating the apparatus. Hence the null result of the Michelson-Morley apparatus is only for the special case of photons travelling in vacuum for which V = c, as conﬁrmed by the modern vacuum interferometer experiment of Brillet and Hall [10], which in-eﬀect conﬁrms (1). However if the apparatus is immersed in a gas then V < c and a non-null eﬀect is expected on rotating the apparatus, since now ∆t = 0. It is essential then in analysing data to correct for this refractive index eﬀect. Putting V = c/n in (10) we ﬁnd, for v << V and when n ≈ 1, that

∆t

=

−L(n

−

1)

v2 c3

.

(11)

However if the data is analysed not using the Fitzgerald-Lorentz contraction (1), then, as done in the old analyses, the estimated time diﬀerence is

∆t

=

2LV V 2 − v2

−

√ 2L V2−

, v2

(12)

which again for v << V and n ≈ 1, gives

∆t

=

−L

v2 c3

.

(13)

The value of ∆t is deduced from analysing the fringe shifts, and then the speed vMM (in previous Michelson-Morley analyses) has been extracted using (13), instead of the correct form (11). ∆t is typically of order 10−15s. However it is very easy to correct for this oversight. From (11) and (13)
we obtain, for the corrected absolute speed through the quantum-foam vQF ,

vQF

=

√vMM . n−1

(14)

Note that fo√r air at STP n = 1.00029, while for helium at STP n = 1.000036, and so the correction factor of 1/ n − 1 is large. The corrected speeds vQF for four MM experiments are shown in Fig.2, and compared with the CBR speed determined from the COBE data [9].
Mu´nera [8] has reviewed these interferometer experiments and uncovered systematic errors and as well applied standard statistical tests to the values originally reported. He has noted that the Michelson-Morley experiments and subsequent repetitions never were null, and that correcting for invalid inter-session averaging leads to even larger non null results. Mu´nera’s new results are as follows. The original Michelson-Morley data [2] now gives vMM = 6.22km/s with a standard deviation on the mean of 0.93km/s for one set of noon sessions, while giving vMM = 6.80km/s with a standard deviation on the mean of 2.49km/s for 18h observations. For Miller [11, 12], from results at Mt Wilson after moving the Morley-Miller [13] experiment, the new result is vMM = 8.22km/s with an upperbound at 95% CL of vMM = 9.61km/s and a lower bound at 95% CL of vMM = 6.83km/s. The Illingworth [14] data (from a helium-ﬁlled inteferometer) gives smaller values of

5

vMM = 3.13km/s with an upperbound at 95% CL of vMM = 4.17km/s and a lower bound at 95% CL of vMM = 2.09km/s.
Hence the Michelson-Morley and Miller values for vMM appear to diﬀer signiﬁcantly from the value from Illingworth. But this is now an expected outcome as Illingworth used helium (to control temperature variations) instead of air. Because of th√e diﬀerent refractive indices of air and helium the correction factors are substantially diﬀerent (1/ n − 1 = 58.7 for air at STP and = 166.7 for helium at STP), and as shown in Fig.2 the air and helium ﬁlled interferometers now give comparable results, when corrected. The use of helium by Illingworth has turned out to be a fortuitous situation that allows us to conﬁrm the correction-factor dependence on n in (14). We have assumed that Illingworth used helium at STP, but if in fact he used helium at 2 atmospheres then his corrected data, in Fig.2, would coincide with the COBE data. We have also used the n value for air at STP in correcting the Miller results even though the experiments were performed at altitude on Mt. Wilson.
700
600
500
400
300

km per s

1

2

3

4

5

Experiment Code

Figure 2: Speeds vQF in km/s determined from various Michelson-Morley experiments: (1) MichelsonMorley (noon observations) [2], (2) Michelson-Morley experiments (18h observations) [2], (3) Miller, Mt. Wilson [11, 12], (4) Illingworth [14], and ﬁnally in (5) the speed from the COBE satellite observation of the CBR microwave spectrum dipole term [9]. The results (1)-(4) are not corrected for the ±30km/s of the orbital motion of the earth about the sum, though this correction was made, as well as that for the satellite orbital speed, in the case of (5). The agreement between the Michelson-Morley result in (1) and the COBE result in (5) is striking. The horizontal line at v = 365km/s is to aid comparisons with the COBE data.
Of course the vacuum experiment by Brillet and Hall [10] gave a genuine null outcome as now expected. The presence of gases in these early experiments, rather than high vacuum, was of course an experimental expediency, but only because of this can we now realise the full implications of these long forgotten experiments.
Hence the old interferometer data is not only non-null but the extracted speeds agree, within errors, with the speed determined from analysis of the CBR dipole component observed by the COBE mission. These results demonstrate that absolute motion has a meaning and is measureable.
A new class of experiment can now be carried out using say a rotateable ﬁbreoptic laser interferometer, and such quantum-gravity experiments are capable of measuring the absolute speed
6

and direction of motion of the Earth through the quantum foam that is space. As well the very large gravity-wave laser interferometers could be used in a gas-ﬁlled mode rather than a vacuum mode and, if suﬃcient temperature control can be achieved, they could be used as absolute motion detectors. Daily and seasonal changes in vQF will be seen as vQF is the vector sum of various contributions due to the eﬀects of the dissipative and non-uniform ﬂows of the quantum foam (which is gravity, [3]) by the earth, moon and sun, as well as the overall eﬀect of the motion of the Earth and the solar system through the CBR determined quantum-foam frame of reference which is, it now turns out, was one of the main discoveries of the COBE mission.
A task for new interferometers would be to establish the in-ward ﬂow rate of the quantum foam, predicted to be 11.2km/s at the Earth’s surface. This quantum-foam ﬂow is a key test of the quantum theory of gravity that has emerged from Process Physics [3, 4, 15, 16, 17, 18], and in which Newton’s gravitational constant G is a measure of the rate at which ‘quantum matter’ dissipates quantum foam.
The signiﬁcance of this new interpretation of the outcomes of the various Michelson-Morley experiments can hardly be overstated, it will change forever how physicists comprehend reality. These results undermine Einstein’s assertion that absolute motion has no meaning. This Process Physics supersedes the prevailing Non-Process Physics modelling of reality, which is characterised by the geometrical modelling of time. It is remarkable that these experiments were carried out with such diligence and care so long ago that their data, when now properly analysed, yield speeds consistent with those found from satellite technology 104 years later, and even more so when we understand that some of the experimenters themselves believed they had failed to detect non-null eﬀects. In eﬀect the Michelson interferometer, operating with dielectric medium, is a quantum interferometer that performed the ﬁrst quantum gravity experiment as it was capable of detecting absolute motion through the quantum foam that is space.
Again we emphasis that the Einstein Special and General Relativity is essentially a model of the potential history of the universe rather than of the universe itself. Clearly if the universe is lawful, then so is its history, though the ‘laws’ for the latter will clearly be diﬀerent and, as is well known, they have the form of a diﬀerential geometry - the spacetime construct. But in Process Physics we see a deeper understanding of reality in which time is not geometry, it is process.
References
[1] A.A. Michelson, The Relative Motion of the Earth and the Luminiferous Aether, Amer. J. Sci S. 3 22, 120-129(1881).
[2] A.A. Michelson and E.W. Morley, On the Relative Motion of the Earth and the Luminiferous Aether, Philos. Mag S. 5, 24, No. 151, 449-463(1887).
[3] R.T. Cahill, Process Physics: From Quantum Foam to General Relativity, gr-qc/0203015.
[4] R.T. Cahill, Process Physics: Inertia, Gravity and the Quantum, to be published in Gen. Rel. and Grav., gr-qc/0110117.
[5] A. Aspect, J. Dalibard and G. Roger, Experimental Test of Bell’s Inequalities Using TimeVarying Analysers, Phys. Rev. Lett. 49, 1804-1807(1982).
[6] L. Hardy, Quantum Mechanics, Local Realistic Theories and Lorentz-Invariant Realistic Theories, Phys. Rev. Lett., 68, 2981(1992).
7

[7] I.C. Percival, Quantum Measurement Breaks Lorentz Symmetry, quant-ph/9906005. [8] H.A. Mun´era, Michelson-Morley Experiments Revisited: Systematic Errors, Consistency
Among Diﬀerent Experiments, and Compatibility with Absolute Space, Aperion 5, No.1-2, 37-54(1998). [9] G.F. Smoot et al, Preliminary Results from the COBE Diﬀerential Microwave Radiometers Large Angular Scale Isotropy of the Cosmic Microwave Background, Astro. J., Part 2 - Letters 371, April 10, L1-L5(1991). [10] A. Brillet and J.L. Hall, Improved Laser Test of the Isotropy of Space, Phys. Rev. Lett. 42, No.9, 549-552(1979). [11] D.C. Miller, Ether Drift Experiments at Mount Wilson, Nat. Acad. Sci. 11, 306-314(1925). [12] D.C. Miller, The Ether-Drift Experiments and the Determination of the Absolute Motion of the Earth, Rev. Mod. Phys. 5, 203-242(1933). [13] E.W. Morley and D.C. Miller, An Experiment to Detect the Fitzgerald-Lorentz Eﬀect, Philos. Mag. S. 6, 9,680(1905). [14] K.K. Illingworth, A Repetition of the M-M Experiment using Kennedy’s Reﬁnement, Phys. Rev. 30, 692-696(1927). [15] R.T. Cahill, C.M. Klinger, and K. Kitto, Process Physics: Modelling Reality as Self-Organising Information, The Physicist 37(6), 191-195(2000), gr-qc/0009023. [16] R.T. Cahill and C.M. Klinger, Self-Referential Noise as a Fundamental Aspect of Reality, Proc. 2nd Int. Conf. on Unsolved Problems of Noise and Fluctuations (UPoN’99), eds. D. Abbott and L. Kish, Adelaide, Australia, 11-15th July 1999, Vol. 511, p. 43 (American Institute of Physics, New York, 2000), gr-qc/9905082. [17] R.T. Cahill and C.M. Klinger, Self-Referential Noise and the Synthesis of Three-Dimensional Space, Gen. Rel. and Grav. 32(3), 529-540(2000), gr-qc/9812083. [18] M. Chown, Random Reality, New Scientist, Feb 26, 165, No 2227, 24-28(2000).
8

