Michelson–Morley experiment

Michelson's 1881 interferometer. Although ultimately it proved incapable of distinguishing between differing theories of aether-dragging, its construction provided important lessons for the design of Michelson and Morley's 1887 instrument.[note 1]

Wikisource has original text related to this article:

The Relative Motion of the Earth and the Luminiferous Ether (1881)

Michelson had a solution to the problem of how to construct a device sufficiently accurate to detect aether flow. In 1877, while teaching at his alma mater, the United States Naval Academy in Annapolis, Michelson conducted his first known light speed experiments as a part of a classroom demonstration. In 1881, he left active U.S. Naval service while in Germany concluding his studies. In that year, Michelson used a prototype experimental device to make several more measurements.

The device he designed, later known as a Michelson interferometer, sent yellow light from a sodium flame (for alignment), or white light (for the actual observations), through a half-silvered mirror that was used to split it into two beams traveling at right angles to one another. After leaving the splitter, the beams traveled out to the ends of long arms where they were reflected back into the middle by small mirrors. They then recombined on the far side of the splitter in an eyepiece, producing a pattern of constructive and destructive interference whose transverse displacement would depend on the relative time it takes light to transit the longitudinal vs. the transverse arms. If the Earth is traveling through an aether medium, a light beam traveling parallel to the flow of that aether will take longer to reflect back and forth than would a beam traveling perpendicular to the aether, because the increase in elapsed time from traveling against the aether wind is more than the time saved by traveling with the aether wind. Michelson expected that the Earth's motion would produce a fringe shift equal to 0.04 fringes—that is, of the separation between areas of the same intensity. He did not observe the expected shift; the greatest average deviation that he measured (in the northwest direction) was only 0.018 fringes; most of his measurements were much less. His conclusion was that Fresnel's hypothesis of a stationary aether with partial aether dragging would have to be rejected, and thus he confirmed Stokes' hypothesis of complete aether dragging.[4]

However, Alfred Potier (and later Hendrik Lorentz) pointed out to Michelson that he had made an error of calculation, and that the expected fringe shift should have been only 0.02 fringes. Michelson's apparatus was subject to experimental errors far too large to say anything conclusive about the aether wind. Definitive measurement of the aether wind would require an experiment with greater accuracy and better controls than the original. Nevertheless, the prototype was successful in demonstrating that the basic method was feasible.[A 6][A 10]

Wikisource has original text related to this article:

On the Relative Motion of the Earth and the Luminiferous Ether (1887)

Figure 5. This figure illustrates the folded light path used in the Michelson–Morley interferometer that enabled a path length of 11 m. a is the light source, an oil lamp. b is a beam splitter. c is a compensating plate so that both the reflected and transmitted beams travel through the same amount of glass (important since experiments were run with white light which has an extremely short coherence length requiring precise matching of optical path lengths for fringes to be visible; monochromatic sodium light was used only for initial alignment[4][note 2]). d, d' and e are mirrors. e' is a fine adjustment mirror. f is a telescope.

In 1885, Michelson began a collaboration with Edward Morley, spending considerable time and money to confirm with higher accuracy Fizeau's 1851 experiment on Fresnel's drag coefficient,[5] to improve on Michelson's 1881 experiment,[1] and to establish the wavelength of light as a standard of length.[6][7] At this time Michelson was professor of physics at the Case School of Applied Science, and Morley was professor of chemistry at Western Reserve University (WRU), which shared a campus with the Case School on the eastern edge of Cleveland. Michelson suffered a nervous breakdown in September 1885, from which he recovered by October 1885. Morley ascribed this breakdown to the intense work of Michelson during the preparation of the experiments. In 1886, Michelson and Morley successfully confirmed Fresnel's drag coefficient – this result was also considered as a confirmation of the stationary aether concept.[A 2]

This result strengthened their hope of finding the aether wind. Michelson and Morley created an improved version of the Michelson experiment with more than enough accuracy to detect this hypothetical effect. The experiment was performed in several periods of concentrated observations between April and July 1887, in the basement of Adelbert Dormitory of WRU (later renamed Pierce Hall, demolished in 1962).[A 11][A 12]

As shown in Fig. 5, the light was repeatedly reflected back and forth along the arms of the interferometer, increasing the path length to 11 m (36 ft). At this length, the drift would be about 0.4 fringes. To make that easily detectable, the apparatus was assembled in a closed room in the basement of the heavy stone dormitory, eliminating most thermal and vibrational effects. Vibrations were further reduced by building the apparatus on top of a large block of sandstone (Fig. 1), about a foot thick and five feet (1.5 m) square, which was then floated in a circular trough of mercury. They estimated that effects of about 0.01 fringe would be detectable.

Figure 6. Fringe pattern produced with a Michelson interferometer using white light. As configured here, the central fringe is white rather than black.

Michelson and Morley and other early experimentalists using interferometric techniques in an attempt to measure the properties of the luminiferous aether, used (partially) monochromatic light only for initially setting up their equipment, always switching to white light for the actual measurements. The reason is that measurements were recorded visually. Purely monochromatic light would result in a uniform fringe pattern. Lacking modern means of environmental temperature control, experimentalists struggled with continual fringe drift even when the interferometer was set up in a basement. Because the fringes would occasionally disappear due to vibrations caused by passing horse traffic, distant thunderstorms and the like, an observer could easily "get lost" when the fringes returned to visibility. The advantages of white light, which produced a distinctive colored fringe pattern, far outweighed the difficulties of aligning the apparatus due to its low coherence length. As Dayton Miller wrote, "White light fringes were chosen for the observations because they consist of a small group of fringes having a central, sharply defined black fringe which forms a permanent zero reference mark for all readings."[A 13][note 3] Use of partially monochromatic light (yellow sodium light) during initial alignment enabled the researchers to locate the position of equal path length, more or less easily, before switching to white light.[note 4]

The mercury trough allowed the device to turn with close to zero friction, so that once having given the sandstone block a single push it would slowly rotate through the entire range of possible angles to the "aether wind", while measurements were continuously observed by looking through the eyepiece. The hypothesis of aether drift implies that because one of the arms would inevitably turn into the direction of the wind at the same time that another arm was turning perpendicularly to the wind, an effect should be noticeable even over a period of minutes.

The expectation was that the effect would be graphable as a sine wave with two peaks and two troughs per rotation of the device. This result could have been expected because during each full rotation, each arm would be parallel to the wind twice (facing into and away from the wind giving identical readings) and perpendicular to the wind twice. Additionally, due to the Earth's rotation, the wind would be expected to show periodic changes in direction and magnitude during the course of a sidereal day.

Because of the motion of the Earth around the Sun, the measured data were also expected to show annual variations.

Figure 7. Michelson and Morley's results. The upper solid line is the curve for their observations at noon, and the lower solid line is that for their evening observations. Note that the theoretical curves and the observed curves are not plotted at the same scale: the dotted curves, in fact, represent only one-eighth of the theoretical displacements.

After all this thought and preparation, the experiment became what has been called the most famous failed experiment in history.[A 1] Instead of providing insight into the properties of the aether, Michelson and Morley's article in the American Journal of Science reported the measurement to be as small as one-fortieth of the expected displacement (Fig. 7), but "since the displacement is proportional to the square of the velocity" they concluded that the measured velocity was "probably less than one-sixth" of the expected velocity of the Earth's motion in orbit and "certainly less than one-fourth".[1] Although this small "velocity" was measured, it was considered far too small to be used as evidence of speed relative to the aether, and it was understood to be within the range of an experimental error that would allow the speed to actually be zero.[A 2] For instance, Michelson wrote about the "decidedly negative result" in a letter to Lord Rayleigh in August 1887:[A 14]

The Experiments on the relative motion of the earth and ether have been completed and the result decidedly negative. The expected deviation of the interference fringes from the zero should have been 0.40 of a fringe – the maximum displacement was 0.02 and the average much less than 0.01 – and then not in the right place. As displacement is proportional to squares of the relative velocities it follows that if the ether does slip past the relative velocity is less than one sixth of the earth’s velocity.

— Albert Abraham Michelson, 1887

From the standpoint of the then current aether models, the experimental results were conflicting. The Fizeau experiment and its 1886 repetition by Michelson and Morley apparently confirmed the stationary aether with partial aether dragging, and refuted complete aether dragging. On the other hand, the much more precise Michelson–Morley experiment (1887) apparently confirmed complete aether dragging and refuted the stationary aether.[A 6] In addition, the Michelson–Morley null result was further substantiated by the null results of other second-order experiments of different kind, namely the Trouton–Noble experiment (1903) and the experiments of Rayleigh and Brace (1902–1904). These problems and their solution led to the development of the Lorentz transformation and special relativity.

After the "failed" experiment Michelson and Morley ceased their aether drift measurements and started to use their newly developed technique to establish the wavelength of light as a standard of length.[6][7]

Expected differential phase shift between light traveling the longitudinal versus the transverse arms of the Michelson–Morley apparatus

The beam travel time in the longitudinal direction can be derived as follows:[A 15] Light is sent from the source and propagates with the speed of light ${\textstyle c}$ in the aether. It passes through the half-silvered mirror at the origin at ${\textstyle T=0}$. The reflecting mirror is at that moment at distance ${\textstyle L}$ (the length of the interferometer arm) and is moving with velocity ${\textstyle v}$. The beam hits the mirror at time ${\textstyle T_{1}}$ and thus travels the distance ${\textstyle cT_{1}}$. At this time, the mirror has traveled the distance ${\textstyle vT_{1}}$. Thus ${\textstyle cT_{1}=L+vT_{1}}$ and consequently the travel time ${\textstyle T_{1}=L/(c-v)}$. The same consideration applies to the backward journey, with the sign of ${\textstyle v}$ reversed, resulting in ${\textstyle cT_{2}=L-vT_{2}}$ and ${\textstyle T_{2}=L/(c+v)}$. The total travel time ${\textstyle T_{\ell }=T_{1}+T_{2}}$ is:

${\displaystyle T_{\ell }={\frac {L}{c-v}}+{\frac {L}{c+v}}={\frac {2L}{c}}{\frac {1}{1-{\frac {v^{2}}{c^{2}}}}}\approx {\frac {2L}{c}}\left(1+{\frac {v^{2}}{c^{2}}}\right)}$

Michelson obtained this expression correctly in 1881, however, in transverse direction he obtained the incorrect expression

${\displaystyle T_{t}={\frac {2L}{c}},}$

because he overlooked the increase in path length in the rest frame of the aether. This was corrected by Alfred Potier (1882) and Hendrik Lorentz (1886). The derivation in the transverse direction can be given as follows (analogous to the derivation of time dilation using a light clock): The beam is propagating at the speed of light ${\textstyle c}$ and hits the mirror at time ${\textstyle T_{3}}$, traveling the distance ${\textstyle cT_{3}}$. At the same time, the mirror has traveled the distance ${\textstyle vT_{3}}$ in the x direction. So in order to hit the mirror, the travel path of the beam is ${\textstyle L}$ in the y direction (assuming equal-length arms) and ${\textstyle vT_{3}}$ in the x direction. This inclined travel path follows from the transformation from the interferometer rest frame to the aether rest frame. Therefore, the Pythagorean theorem gives the actual beam travel distance of ${\textstyle {\sqrt {L^{2}+\left(vT_{3}\right)^{2}}}}$. Thus ${\textstyle cT_{3}={\sqrt {L^{2}+\left(vT_{3}\right)^{2}}}}$ and consequently the travel time ${\textstyle T_{3}=L/{\sqrt {c^{2}-v^{2}}}}$, which is the same for the backward journey. The total travel time ${\textstyle T_{t}=2T_{3}}$ is:

${\displaystyle T_{t}={\frac {2L}{\sqrt {c^{2}-v^{2}}}}={\frac {2L}{c}}{\frac {1}{\sqrt {1-{\frac {v^{2}}{c^{2}}}}}}\approx {\frac {2L}{c}}\left(1+{\frac {v^{2}}{2c^{2}}}\right)}$

The time difference between ${\displaystyle T_{\ell }}$ and ${\displaystyle T_{t}}$ is given by[A 16]

${\displaystyle T_{\ell }-T_{t}={\frac {2L}{c}}\left({\frac {1}{1-{\frac {v^{2}}{c^{2}}}}}-{\frac {1}{\sqrt {1-{\frac {v^{2}}{c^{2}}}}}}\right)}$

To find the path difference, simply multiply by ${\displaystyle c}$;

${\displaystyle \Delta {\lambda }_{1}=2L\left({\frac {1}{1-{\frac {v^{2}}{c^{2}}}}}-{\frac {1}{\sqrt {1-{\frac {v^{2}}{c^{2}}}}}}\right)}$

The path difference is denoted by ${\displaystyle \Delta \lambda }$ because the beams are out of phase by a some number of wavelengths (${\displaystyle \lambda }$). To visualise this, consider taking the two beam paths along the longitudinal and transverse plane, and lying them straight (an animation of this is shown at minute 11:00, The Mechanical Universe, episode 41[8]). One path will be longer than the other, this distance is ${\displaystyle \Delta \lambda }$. Alternatively, consider the rearrangement of the speed of light formula ${\displaystyle c{\Delta }T=\Delta \lambda }$ .

If the relation ${\displaystyle {v^{2}}/{c^{2}}<<1}$ is true (if the velocity of the aether is small relative to the speed of light), then the expression can be simplified using a first order binomial expansion;

${\displaystyle (1-x)^{n}\approx {1-nx}}$

So, rewriting the above in terms of powers;

${\displaystyle \Delta {\lambda }_{1}=2L\left(\left({1-{\frac {v^{2}}{c^{2}}}}\right)^{-1}-\left(1-{\frac {v^{2}}{c^{2}}}\right)^{-1/2}\right)}$

Applying binomial simplification;[9]

${\displaystyle \Delta {\lambda }_{1}=2L\left((1+{\frac {v^{2}}{c^{2}}})-(1+{\frac {v^{2}}{2c^{2}}})\right)={2L}{\frac {v^{2}}{2c^{2}}}}$

Therefore;

${\displaystyle \Delta {\lambda }_{1}={L}{\frac {v^{2}}{c^{2}}}}$

It can be seen from this derivation that aether wind manifests as a path difference. The path difference is zero only when the interferometer is aligned with or perpendicular to the aether wind, and it reaches a maximum when it is at a 45° angle. The path difference can be any fraction of the wavelength, depending on the angle and speed of the aether wind.

To prove the existence of the aether, Michaelson and Morley sought to find the "fringe shift". The idea was simple, the fringes of the interference pattern should shift when rotating it by 90° as the two beams have exchanged roles. To find the fringe shift, subtract the path difference in first orientation by the path difference in the second, then divide by the wavelength, ${\displaystyle \lambda }$, of light;[9]

${\displaystyle n={\frac {\Delta \lambda _{1}-\Delta \lambda _{2}}{\lambda }}\approx {\frac {2Lv^{2}}{\lambda c^{2}}}.}$

Note the difference between ${\displaystyle \Delta \lambda }$, which is some number of wavelengths, and ${\displaystyle \lambda }$ which is a single wavelength. As can be seen by this relation, fringe shift n is a unitless quantity.

Since L ≈ 11 meters and λ ≈ 500 nanometers, the expected fringe shift was n ≈ 0.44. The negative result led Michelson to the conclusion that there is no measurable aether drift.[1] However, he never accepted this on a personal level, and the negative result haunted him for the rest of his life (Source; The Mechanical Universe, episode 41[8]).

If the same situation is described from the view of an observer co-moving with the interferometer, then the effect of aether wind is similar to the effect experienced by a swimmer, who tries to move with velocity ${\textstyle c}$ against a river flowing with velocity ${\textstyle v}$.[A 17]

In the longitudinal direction the swimmer first moves upstream, so his velocity is diminished due to the river flow to ${\textstyle c-v}$. On his way back moving downstream, his velocity is increased to ${\textstyle c+v}$. This gives the beam travel times ${\textstyle T_{1}}$ and ${\textstyle T_{2}}$ as mentioned above.

In the transverse direction, the swimmer has to compensate for the river flow by moving at a certain angle against the flow direction, in order to sustain his exact transverse direction of motion and to reach the other side of the river at the correct location. This diminishes his speed to ${\textstyle {\sqrt {c^{2}-v^{2}}}}$, and gives the beam travel time ${\textstyle T_{3}}$ as mentioned above.

The classical analysis predicted a relative phase shift between the longitudinal and transverse beams which in Michelson and Morley's apparatus should have been readily measurable. What is not often appreciated (since there was no means of measuring it), is that motion through the hypothetical aether should also have caused the two beams to diverge as they emerged from the interferometer by about 10⁻⁸ radians.[A 18]

For an apparatus in motion, the classical analysis requires that the beam-splitting mirror be slightly offset from an exact 45° if the longitudinal and transverse beams are to emerge from the apparatus exactly superimposed. In the relativistic analysis, Lorentz-contraction of the beam splitter in the direction of motion causes it to become more perpendicular by precisely the amount necessary to compensate for the angle discrepancy of the two beams.[A 18]

A first step to explaining the Michelson and Morley experiment's null result was found in the FitzGerald–Lorentz contraction hypothesis, now simply called length contraction or Lorentz contraction, first proposed by George FitzGerald (1889) in a letter to same journal that published the Michelson-Morley paper, as "almost the only hypothesis that can reconcile" the apparent contradictions. It was independently also proposed by Hendrik Lorentz (1892).[A 19] According to this law all objects physically contract by ${\textstyle L/\gamma }$ along the line of motion (originally thought to be relative to the aether), ${\textstyle \gamma =1/{\sqrt {1-v^{2}/c^{2}}}}$ being the Lorentz factor. This hypothesis was partly motivated by Oliver Heaviside's discovery in 1888 that electrostatic fields are contracting in the line of motion. But since there was no reason at that time to assume that binding forces in matter are of electric origin, length contraction of matter in motion with respect to the aether was considered an ad hoc hypothesis.[A 10]

If length contraction of ${\textstyle L}$ is inserted into the above formula for ${\textstyle T_{\ell }}$, then the light propagation time in the longitudinal direction becomes equal to that in the transverse direction:

${\displaystyle T_{\ell }={\frac {2L{\sqrt {1-{\frac {v^{2}}{c^{2}}}}}}{c}}{\frac {1}{1-{\frac {v^{2}}{c^{2}}}}}={\frac {2L}{c}}{\frac {1}{\sqrt {1-{\frac {v^{2}}{c^{2}}}}}}=T_{t}}$

However, length contraction is only a special case of the more general relation, according to which the transverse length is larger than the longitudinal length by the ratio ${\textstyle \gamma }$. This can be achieved in many ways. If ${\textstyle L_{1}}$ is the moving longitudinal length and ${\textstyle L_{2}}$ the moving transverse length, ${\textstyle L'_{1}=L'_{2}}$ being the rest lengths, then it is given:[A 20]

${\displaystyle {\frac {L_{2}}{L_{1}}}={\frac {L'_{2}}{\varphi }}\left/{\frac {L'_{1}}{\gamma \varphi }}\right.=\gamma .}$

${\textstyle \varphi }$ can be arbitrarily chosen, so there are infinitely many combinations to explain the Michelson–Morley null result. For instance, if ${\textstyle \varphi =1}$ the relativistic value of length contraction of ${\textstyle L_{1}}$ occurs, but if ${\textstyle \varphi =1/\gamma }$ then no length contraction but an elongation of ${\textstyle L_{2}}$ occurs. This hypothesis was later extended by Joseph Larmor (1897), Lorentz (1904) and Henri Poincaré (1905), who developed the complete Lorentz transformation including time dilation in order to explain the Trouton–Noble experiment, the Experiments of Rayleigh and Brace, and Kaufmann's experiments. It has the form

${\displaystyle x'=\gamma \varphi (x-vt),\ y'=\varphi y,\ z'=\varphi z,\ t'=\gamma \varphi \left(t-{\frac {vx}{c^{2}}}\right)}$

It remained to define the value of ${\textstyle \varphi }$, which was shown by Lorentz (1904) to be unity.[A 20] In general, Poincaré (1905)[A 21] demonstrated that only ${\textstyle \varphi =1}$ allows this transformation to form a group, so it is the only choice compatible with the principle of relativity, i.e., making the stationary aether undetectable. Given this, length contraction and time dilation obtain their exact relativistic values.

Albert Einstein formulated the theory of special relativity by 1905, deriving the Lorentz transformation and thus length contraction and time dilation from the relativity postulate and the constancy of the speed of light, thus removing the ad hoc character from the contraction hypothesis. Einstein emphasized the kinematic foundation of the theory and the modification of the notion of space and time, with the stationary aether no longer playing any role in his theory. He also pointed out the group character of the transformation. Einstein was motivated by Maxwell's theory of electromagnetism (in the form as it was given by Lorentz in 1895) and the lack of evidence for the luminiferous aether.[A 22]

This allows a more elegant and intuitive explanation of the Michelson–Morley null result. In a comoving frame the null result is self-evident, since the apparatus can be considered as at rest in accordance with the relativity principle, thus the beam travel times are the same. In a frame relative to which the apparatus is moving, the same reasoning applies as described above in "Length contraction and Lorentz transformation", except the word "aether" has to be replaced by "non-comoving inertial frame". Einstein wrote in 1916:[A 23]

Although the estimated difference between these two times is exceedingly small, Michelson and Morley performed an experiment involving interference in which this difference should have been clearly detectable. But the experiment gave a negative result — a fact very perplexing to physicists. Lorentz and FitzGerald rescued the theory from this difficulty by assuming that the motion of the body relative to the æther produces a contraction of the body in the direction of motion, the amount of contraction being just sufficient to compensate for the difference in time mentioned above. Comparison with the discussion in Section 11 shows that also from the standpoint of the theory of relativity this solution of the difficulty was the right one. But on the basis of the theory of relativity the method of interpretation is incomparably more satisfactory. According to this theory there is no such thing as a "specially favoured" (unique) co-ordinate system to occasion the introduction of the æther-idea, and hence there can be no æther-drift, nor any experiment with which to demonstrate it. Here the contraction of moving bodies follows from the two fundamental principles of the theory, without the introduction of particular hypotheses; and as the prime factor involved in this contraction we find, not the motion in itself, to which we cannot attach any meaning, but the motion with respect to the body of reference chosen in the particular case in point. Thus for a co-ordinate system moving with the earth the mirror system of Michelson and Morley is not shortened, but it is shortened for a co-ordinate system which is at rest relatively to the sun.

— Albert Einstein, 1916

The extent to which the null result of the Michelson–Morley experiment influenced Einstein is disputed. Alluding to some statements of Einstein, many historians argue that it played no significant role in his path to special relativity,[A 24][A 25] while other statements of Einstein probably suggest that he was influenced by it.[A 26] In any case, the null result of the Michelson–Morley experiment helped the notion of the constancy of the speed of light gain widespread and rapid acceptance.[A 24]

It was later shown by Howard Percy Robertson (1949) and others[A 4][A 27] (see Robertson–Mansouri–Sexl test theory), that it is possible to derive the Lorentz transformation entirely from the combination of three experiments. First, the Michelson–Morley experiment showed that the speed of light is independent of the orientation of the apparatus, establishing the relationship between longitudinal (β) and transverse (δ) lengths. Then in 1932, Roy Kennedy and Edward Thorndike modified the Michelson–Morley experiment by making the path lengths of the split beam unequal, with one arm being very short.[10] The Kennedy–Thorndike experiment took place for many months as the Earth moved around the sun. Their negative result showed that the speed of light is independent of the velocity of the apparatus in different inertial frames. In addition it established that besides length changes, corresponding time changes must also occur, i.e., it established the relationship between longitudinal lengths (β) and time changes (α). So both experiments do not provide the individual values of these quantities. This uncertainty corresponds to the undefined factor ${\textstyle \varphi }$ as described above. It was clear due to theoretical reasons (the group character of the Lorentz transformation as required by the relativity principle) that the individual values of length contraction and time dilation must assume their exact relativistic form. But a direct measurement of one of these quantities was still desirable to confirm the theoretical results. This was achieved by the Ives–Stilwell experiment (1938), measuring α in accordance with time dilation. Combining this value for α with the Kennedy–Thorndike null result shows that β must assume the value of relativistic length contraction. Combining β with the Michelson–Morley null result shows that δ must be zero. Therefore, the Lorentz transformation with ${\textstyle \varphi =1}$ is an unavoidable consequence of the combination of these three experiments.[A 4]

Special relativity is generally considered the solution to all negative aether drift (or isotropy of the speed of light) measurements, including the Michelson–Morley null result. Many high precision measurements have been conducted as tests of special relativity and modern searches for Lorentz violation in the photon, electron, nucleon, or neutrino sector, all of them confirming relativity.

As mentioned above, Michelson initially believed that his experiment would confirm Stokes' theory, according to which the aether was fully dragged in the vicinity of the earth (see Aether drag hypothesis). However, complete aether drag contradicts the observed aberration of light and was contradicted by other experiments as well. In addition, Lorentz showed in 1886 that Stokes's attempt to explain aberration is contradictory.[A 6][A 5]

Furthermore, the assumption that the aether is not carried in the vicinity, but only within matter, was very problematic as shown by the Hammar experiment (1935). Hammar directed one leg of his interferometer through a heavy metal pipe plugged with lead. If aether were dragged by mass, it was theorized that the mass of the sealed metal pipe would have been enough to cause a visible effect. Once again, no effect was seen, so aether-drag theories are considered to be disproven.

Walther Ritz's emission theory (or ballistic theory) was also consistent with the results of the experiment, not requiring aether. The theory postulates that light has always the same velocity in respect to the source.[A 28] However de Sitter noted that emitter theory predicted several optical effects that were not seen in observations of binary stars in which the light from the two stars could be measured in a spectrometer. If emission theory were correct, the light from the stars should experience unusual fringe shifting due to the velocity of the stars being added to the speed of the light, but no such effect could be seen. It was later shown by J. G. Fox that the original de Sitter experiments were flawed due to extinction,[11] but in 1977 Brecher observed X-rays from binary star systems with similar null results.[12] Furthermore, Filippas and Fox (1964) conducted terrestrial particle accelerator tests specifically designed to address Fox's earlier "extinction" objection, the results being inconsistent with source dependence of the speed of light.[13]

Optical tests of the isotropy of the speed of light became commonplace.[A 35] New technologies, including the use of lasers and masers, have significantly improved measurement precision. (In the following table, only Essen (1955), Jaseja (1964), and Shamir/Fox (1969) are experiments of Michelson–Morley type, i.e., comparing two perpendicular beams. The other optical experiments employed different methods.)

Figure 9. Michelson–Morley experiment with cryogenic optical resonators of a form such as was used by Müller et al. (2003).[32]

During the early 21st century, there has been a resurgence in interest in performing precise Michelson–Morley type experiments using lasers, masers, cryogenic optical resonators, etc. This is in large part due to predictions of quantum gravity that suggest that special relativity may be violated at scales accessible to experimental study. The first of these highly accurate experiments was conducted by Brillet & Hall (1979), in which they analyzed a laser frequency stabilized to a resonance of a rotating optical Fabry–Pérot cavity. They set a limit on the anisotropy of the speed of light resulting from the Earth's motions of Δc/c ≈ 10⁻¹⁵, where Δc is the difference between the speed of light in the x- and y-directions.[33]

As of 2015, optical and microwave resonator experiments have improved this limit to Δc/c ≈ 10⁻¹⁸. In some of them, the devices were rotated or remained stationary, and some were combined with the Kennedy–Thorndike experiment. In particular, Earth's direction and velocity (ca. 368 km/s (229 mi/s)) relative to the CMB rest frame are ordinarily used as references in these searches for anisotropies.

Author	Year	Description	Δc/c
Wolf et al.[34]	2003	The frequency of a stationary cryogenic microwave oscillator, consisting of sapphire crystal operating in a whispering gallery mode, is compared to a hydrogen maser whose frequency was compared to caesium and rubidium atomic fountain clocks. Changes during Earth's rotation have been searched for. Data between 2001 and 2002 was analyzed.	${\displaystyle \lesssim 10^{-15}}$
Müller et al.[32]	2003	Two optical resonators constructed from crystalline sapphire, controlling the frequencies of two Nd:YAG lasers, are set at right angles within a helium cryostat. A frequency comparator measures the beat frequency of the combined outputs of the two resonators.
Wolf et al.[35]	2004	See Wolf et al. (2003). An active temperature control was implemented. Data between 2002 and 2003 was analyzed.
Wolf et al.[36]	2004	See Wolf et al. (2003). Data between 2002 and 2004 was analyzed.
Antonini et al.[37]	2005	Similar to Müller et al. (2003), though the apparatus itself was set into rotation. Data between 2002 and 2004 was analyzed.	${\displaystyle \lesssim 10^{-16}}$
Stanwix et al.[38]	2005	Similar to Wolf et al. (2003). The frequency of two cryogenic oscillators was compared. In addition, the apparatus was set into rotation. Data between 2004 and 2005 was analyzed.
Herrmann et al.[39]	2005	Similar to Müller et al. (2003). The frequencies of two optical Fabry–Pérot resonators cavities are compared – one cavity was continuously rotating while the other one was stationary oriented north–south. Data between 2004 and 2005 was analyzed.
Stanwix et al.[40]	2006	See Stanwix et al. (2005). Data between 2004 and 2006 was analyzed.
Müller et al.[41]	2007	See Herrmann et al. (2005) and Stanwix et al. (2006). Data of both groups collected between 2004 and 2006 are combined and further analyzed. Since the experiments are located at difference continents, at Berlin and Perth respectively, the effects of both the rotation of the devices themselves and the rotation of Earth could be studied.
Eisele et al.[2]	2009	The frequencies of a pair of orthogonal oriented optical standing wave cavities are compared. The cavities were interrogated by a Nd:YAG laser. Data between 2007 and 2008 was analyzed.	${\displaystyle \lesssim 10^{-17}}$
Herrmann et al.[3]	2009	The frequencies of a pair of rotating, orthogonal optical Fabry–Pérot resonators are compared. The frequencies of two Nd:YAG lasers are stabilized to resonances of these resonators.
Nagel et al.[42]	2015	The frequencies of a pair of rotating, orthogonal microwave resonators are compared.

Figure 10. ⁷Li-NMR spectrum of LiCl (1M) in D₂O. The sharp, unsplit NMR line of this isotope of lithium is evidence for the isotropy of mass and space.

Examples of other experiments not based on the Michelson–Morley principle, i.e., non-optical isotropy tests achieving an even higher level of precision, are Clock comparison or Hughes–Drever experiments. In Drever's 1961 experiment, ⁷Li nuclei in the ground state, which has total angular momentum J = 3/2, were split into four equally spaced levels by a magnetic field. Each transition between a pair of adjacent levels should emit a photon of equal frequency, resulting in a single, sharp spectral line. However, since the nuclear wave functions for different M_J have different orientations in space relative to the magnetic field, any orientation dependence, whether from an aether wind or from a dependence on the large-scale distribution of mass in space (see Mach's principle), would perturb the energy spacings between the four levels, resulting in an anomalous broadening or splitting of the line. No such broadening was observed. Modern repeats of this kind of experiment have provided some of the most accurate confirmations of the principle of Lorentz invariance.[A 36]