The Michelson-Morley Experiment Explained

In July 1887, Albert Michelson and Edward Morley set up the most sensitive light-measuring instrument of the nineteenth century in the basement of a dormitory at what is now Case Western Reserve University in Cleveland, Ohio. They were trying to detect Earth's motion through a substance nobody had ever directly observed: the luminiferous ether, the medium physicists assumed carried light waves through space the way air carries sound. They found nothing. That null result is one of the most consequential negative findings in the history of physics, and it helped clear the ground for Einstein's theory of relativity eighteen years later.

The problem physicists thought they had

Every wave known to nineteenth-century science needed a medium. Sound needs air; ocean waves need water; a plucked string needs, well, the string. When James Clerk Maxwell worked out in the 1860s that light was an electromagnetic wave, physicists assumed light must be riding some equivalent medium too, and they called it the luminiferous ether — a hypothetical substance imagined to fill all of space, perfectly transparent, essentially massless, and stiff enough to support waves moving at 300,000 kilometers per second.

The ether idea did real explanatory work. It gave physicists a physical picture for something that otherwise seemed to travel through pure vacuum, which felt conceptually uncomfortable at the time. But a medium that fills all of space creates a testable prediction: if Earth orbits the Sun at roughly 30 kilometers per second (about 18.6 miles per second), and the ether stays still while Earth plows through it, then Earth should experience an "ether wind" analogous to the way a moving car creates a wind against a stationary atmosphere. A beam of light sent in the same direction as that wind should behave slightly differently than a beam sent across it, the way swimming with a river current is faster than swimming against it or across it.

The trouble was scale. Earth's 30 km/s orbital speed is roughly one ten-thousandth of the speed of light. Any effect on light's travel time from an ether wind that fast would be minuscule, well below what ordinary timing instruments of the 1880s could resolve. Detecting it required an entirely new class of instrument, one built specifically to catch a difference too small for a stopwatch or a telescope to see directly.

Michelson had already built a reputation for extraordinarily precise measurements of the speed of light before this experiment, and he would go on to become the first American to win a Nobel Prize in a physical science, awarded in 1907 largely for the optical precision instruments, including the interferometer, that grew out of this line of work (Nobel Prize in Physics 1907, Nobel Foundation).

Michelson's earlier attempt, and why it needed Morley

Michelson had actually tried a version of this experiment once before, alone, in Potsdam, Germany, in 1881, while on leave from the U.S. Navy. That earlier apparatus was sensitive in principle but plagued by a real-world problem: vibration. Passing carriages, footsteps in nearby rooms, and ordinary building tremors jostled the instrument enough to blur the tiny interference pattern it was trying to measure. The 1881 result was inconclusive, not because the physics was wrong, but because the instrument couldn't hold still enough to trust.

Six years later, Michelson, by then teaching at the Case School of Applied Science, partnered with Edward Morley, a chemist at the neighboring Western Reserve University, to rebuild the experiment properly. Morley contributed both his experimental rigor as a chemist accustomed to exacting measurement and a critical engineering fix: mounting the entire interferometer on a massive sandstone slab roughly 1.5 meters (about 5 feet) square, floating in a trough of liquid mercury. Mercury's near-frictionless buoyancy let the whole assembly rotate smoothly on its own axis without picking up vibration from the building around it, and it could be spun continuously rather than reset between readings.

The rebuilt instrument, installed in the basement of Morley's dormitory building to minimize thermal drift and vibration from foot traffic above, used mirrors to fold the light path back and forth, extending the effective arm length to roughly 11 meters (about 36 feet) even though the physical footprint of the device was much smaller. That extended path length mattered enormously: the longer the light travels before recombining, the more any tiny speed difference accumulates into something the human eye can actually see in the interference pattern (Case Western Reserve University, Encyclopedia of Cleveland History).

How the interferometer actually worked

The core idea behind Michelson's interferometer is elegant enough to sketch on a napkin, even though building one precise enough to matter took years of engineering. A beam of light from a monochromatic sodium lamp hits a half-silvered mirror set at 45 degrees, called a beam splitter. Half the light passes straight through; the other half reflects at a right angle. The two resulting beams travel down separate perpendicular arms of equal length, each ending in a fully silvered mirror that bounces the light straight back the way it came. Both beams return to the beam splitter and recombine into a single beam that travels on to an eyepiece, where an observer watches the result.

Interference pattern — the alternating bright and dark bands, called fringes, produced when two light waves recombine slightly out of step with each other, revealing timing differences far smaller than any clock of the era could measure directly. If the two beams take exactly the same amount of time to make their round trip, their light waves arrive back in phase, crest lining up with crest, and they reinforce each other into a bright, stable pattern. If one beam is delayed by even a fraction of a wavelength of light, relative to the other, the waves arrive slightly out of step, and the interference pattern shifts visibly. Because visible light has a wavelength of roughly 500 to 600 nanometers, this technique is sensitive to timing differences corresponding to a tiny fraction of a wavelength, an extraordinary level of precision for 1887 and, in modified form, still one of the most sensitive measurement techniques in physics today.

This is the same underlying principle later scaled up to planetary dimensions in the LIGO gravitational wave detectors, which use laser interferometry across kilometers-long arms to catch distortions in spacetime thousands of times smaller than a proton. Michelson's basement apparatus and LIGO's twin observatories are separated by more than a century of engineering, but they share the same trick: split a beam, send it two ways, and read the mismatch in how it comes back together.

What result would have confirmed the ether?

If the ether hypothesis were correct, rotating the interferometer through 90 degrees should have shifted the interference pattern by roughly 0.4 of a fringe width, as the arm that had been aligned with Earth's motion swapped places with the arm that had been perpendicular to it. Michelson and Morley's apparatus was sensitive enough to reliably detect a shift as small as roughly 0.01 of a fringe, an order of magnitude below the predicted effect. They measured essentially nothing.

Why did they rotate the whole device?

Rotating the apparatus let the two arms continuously swap roles relative to Earth's direction of travel, so any real ether-wind effect should have shown up as a smooth, predictable, repeating shift in the fringe pattern as the device turned. A rotating design also canceled out most instrumental errors that stay fixed relative to the apparatus itself, since only an effect tied to the ether wind would rotate in sync with the instrument's orientation.

The null result that would not go away

Michelson and Morley published their findings in the American Journal of Science in November 1887. Their instrument was calibrated to detect an ether wind as slow as roughly 8 kilometers per second, well under half of Earth's actual 30 km/s orbital speed, and across repeated trials at different times of day and different seasons of the year, it detected nothing resembling the expected effect. They deliberately varied the timing to rule out the unlikely coincidence that Earth might, at the exact moment of testing, have been moving in a way that canceled the ether wind by chance.

The result held steady regardless of when they looked. This was not a matter of the instrument being slightly under-sensitive; it was sensitive enough, and it still came up empty. For physicists steeped in the ether framework, this was genuinely disorienting. An entire generation of optical and electromagnetic theory rested on the ether's existence, and here was direct experimental evidence that, whatever was going on, Earth did not seem to be moving relative to it in the way theory demanded.

The immediate response from the physics community was not to discard the ether but to patch it. Irish physicist George FitzGerald in 1889 and Dutch physicist Hendrik Lorentz independently, working through the 1890s, proposed that objects moving through the ether physically contract, ever so slightly, along their direction of motion. This idea, now called the Lorentz-FitzGerald contraction, was constructed specifically so that the contraction of the interferometer's arm would exactly cancel the timing difference the ether wind should have produced, hiding the effect from detection. It was a mathematically consistent patch, and it turned out to contain a kernel of real physics that Einstein would later derive from entirely different first principles. But as an explanation offered in the 1890s, it had an uncomfortable feature: it explained away a null result by inventing a new, otherwise unverified physical effect, rather than questioning whether the ether existed at all.

The interferometer Michelson and Morley used could detect a fringe shift equivalent to about one-hundredth of the wavelength of light, roughly 6 nanometers of path-length difference between the two beams, at a time when the most precise mechanical measuring tools of the era operated at scales thousands of times coarser (American Journal of Science, 1887; APS News retrospective, 2007).

How Einstein resolved it, and the nuance historians insist on

The resolution most physics textbooks now present came in 1905, when Albert Einstein published his special theory of relativity. Einstein's approach was strikingly different from the FitzGerald-Lorentz patch. Rather than adding a new physical effect to preserve the ether, Einstein simply postulated that the speed of light in a vacuum is identical for every observer, no matter how fast that observer is moving relative to anything else, and that the laws of physics look the same in every non-accelerating reference frame. Once you accept that postulate, the ether becomes unnecessary. There is nothing for Earth to be moving "through" in the way nineteenth-century physicists pictured it, so there is no wind left to detect, and the Michelson-Morley null result stops being a puzzle and becomes exactly what the theory predicts.

Here the historiographical record deserves more honesty than most popular science writing gives it. Einstein's 1905 paper, "On the Electrodynamics of Moving Bodies," does not name the Michelson-Morley experiment at all. It refers only vaguely to "the unsuccessful attempts to discover any motion of the earth relative to the light medium." Historians of physics who have combed through Einstein's letters, later interviews, and the paper's own structure have found a genuinely mixed picture: Einstein gave inconsistent accounts across his life of how much the experiment influenced him directly, at times crediting it and at other times emphasizing that his thinking grew primarily out of theoretical tensions in Maxwell's electromagnetism and thought experiments about chasing a beam of light, along with results like the Fizeau experiment on light in moving water and the phenomenon of stellar aberration (arXiv historical analyses of Einstein's 1905 reasoning; MacTutor History of Mathematics archive, University of St Andrews).

What is not in dispute is the broader context: by 1905, the Michelson-Morley result and its repeated confirmations by other experimenters had already made the classical ether concept scientifically strained, and the physics community was primed for a theory that could explain a stubborn null result without inventing new, unverifiable patches. Whether Einstein was consciously reasoning from that specific 1887 experiment or arriving independently at a framework that happened to explain it, special relativity did the job the Lorentz-FitzGerald contraction could only gesture at. It made the null result the expected outcome of the theory rather than an anomaly requiring rescue.

Reading the Michelson-Morley data against the two competing predictions

The clearest way to see why the 1887 result mattered is to lay the ether prediction and the actual measurement side by side.

Prediction Source	Expected Fringe Shift	What Was Actually Measured	Interpretation
Stationary ether, Earth's 30 km/s orbital motion	~0.4 fringe widths	<0.01 fringe widths	Ether wind effect not detected at predicted scale
Lorentz-FitzGerald contraction (patched ether)	~0 fringe widths (contraction cancels the effect)	<0.01 fringe widths	Mathematically matched the null result, but added an unverified new physical assumption
Einstein's special relativity (1905)	0 fringe widths, by definition of the postulate	<0.01 fringe widths	Null result required no extra assumption; it is the theory's baseline prediction
Modern laser and maser repeats (20th–21st century)	0 fringe shift if relativity holds	No shift detected, to increasingly tight precision	Confirms special relativity's light-speed postulate with far greater sensitivity than 1887

The 1887 instrument was sensitive enough to catch the ether wind predicted by classical theory, and instead measured a result forty times smaller than that prediction, a gap only special relativity would eventually explain without a patch.

Why this experiment still matters

Michelson-Morley is taught less today for what it found and more for what it demonstrates about how science treats a well-designed failure. A negative result from a sloppy or under-sensitive instrument tells you very little; a negative result from an instrument built specifically to catch the effect, and sensitive enough to have caught it if it were there, tells you something real about the world. That distinction is why physicists took the 1887 null result seriously enough to spend nearly two decades wrestling with its implications rather than shrugging it off.

The experiment has also been repeated many times since, with steadily improving technology. Twentieth-century versions using lasers, and later rotating optical cavities and cryogenic sapphire oscillators, have pushed the sensitivity of the same basic test thousands of times beyond what Michelson and Morley could achieve, and every one of them has continued to find no directional dependence in the speed of light, tightening the experimental bound on any hypothetical ether-like effect to extraordinarily small limits (arXiv summaries of modern Michelson-Morley-type tests). The original 1887 apparatus, in other words, was not just an interesting historical dead end. It launched a lineage of experiments that remains one of the most stringently tested claims in all of physics.

There's also a methodological lesson that outlives the specific physics. Michelson and Morley did not set out to prove special relativity, a theory that did not exist yet and that neither of them lived to see fully vindicated in Michelson's case only through decades more evidence. They set out to measure something they were confident existed, built the best instrument of their era to do it, and reported honestly that they found nothing. That honesty, more than the specific null result, is what let the rest of physics eventually build something better on top of it.

Common misconceptions about the experiment

A frequent misunderstanding is that Michelson and Morley "proved the ether doesn't exist." Strictly, their experiment showed only that Earth's motion through a stationary ether, at the specific magnitude and in the specific way then-current theory predicted, was not detectable by their instrument. It took the cumulative weight of repeated null results, the awkwardness of the Lorentz-FitzGerald patch, and ultimately the explanatory power of special relativity to make the ether concept genuinely obsolete rather than merely embarrassed.

Another common overstatement, ironically the mirror image of the first, is the claim that Einstein built special relativity directly and primarily on the Michelson-Morley result, as though he read their 1887 paper and derived his theory from it point by point. As the historiography above shows, that account oversimplifies a messier real story in which Einstein's own path ran mostly through theoretical concerns about electrodynamics and symmetry, with the experimental null result serving as important background context rather than a documented direct trigger.

A third misconception, more minor but persistent, is that the experiment was some kind of failure or embarrassment for Michelson. The opposite is true: the interferometer he built for this test became one of the most important instruments in the history of experimental physics, and it, along with his other precision optical work, is what the Nobel committee cited when it awarded him the 1907 Nobel Prize in Physics, the first ever awarded to an American in the sciences (Nobel Foundation, 1907).

The takeaway

Michelson and Morley set out to measure Earth's motion through a medium that turned out not to exist, and their scrupulously careful null result became one of the strongest pieces of evidence for a theory neither of them anticipated. Special relativity did not just survive the test; it made the test's failure the expected, unremarkable baseline rather than a puzzle needing rescue. Whether or not Einstein was thinking directly about their 1887 paper as he wrote his own in 1905, the experiment had already done the work of making the old picture of space untenable, which is what a good experiment does even when it "fails": it tells you something is wrong with your assumptions, not with your instrument.

---

Read next:

• What is a photon, explained
• The Rutherford gold foil experiment explained
• Entropy and the arrow of time explained
• Albert Einstein: the eccentric genius who hated socks

FAQ

Q: What was the Michelson-Morley experiment trying to prove?

It aimed to detect Earth's motion through the hypothetical luminiferous ether, the medium nineteenth-century physicists believed carried light waves through space, by comparing the travel time of light beams sent along two perpendicular paths using a precision interferometer.

Q: Why is a null result considered such an important finding?

Because the interferometer was sensitive enough to catch an ether wind far smaller than the roughly 30 km/s effect Earth's orbital motion should have produced, and it still found nothing close to that. A sensitive instrument finding nothing is much stronger evidence against a hypothesis than an insensitive one simply failing to confirm it.

Q: Did Einstein use the Michelson-Morley result to develop relativity?

Einstein's 1905 paper does not name the experiment directly, and historians who have studied his notebooks and later interviews find inconsistent accounts of how much it influenced him personally. What is clear is that the experiment, along with repeated confirmations of its null result, had already made the ether concept scientifically strained by 1905, priming the field for a theory that discarded it entirely.

Q: Has the experiment been repeated with modern equipment?

Yes. Later versions using lasers, rotating optical resonators, and cryogenic oscillators have confirmed the original null result with sensitivity thousands of times better than the 1887 apparatus, further tightening the experimental bound on any possible ether-like effect on the speed of light.

Q: What replaced the ether concept in physics?

Special relativity replaced it. Einstein's 1905 postulate that the speed of light in a vacuum is constant for every observer, regardless of their own motion, removed the need for a medium altogether and turned the Michelson-Morley null result into the expected outcome of the theory rather than an anomaly.

Q: How precise was the original 1887 interferometer?

The apparatus could reliably detect a fringe shift of about one-hundredth of a wavelength of the sodium light used, corresponding to a path-length difference of only a few nanometers between the two arms, achieved using mirrors that extended the effective light path to roughly 11 meters while the physical device itself was much more compact.

Q: Why did Michelson and Morley float their instrument in mercury?

Michelson's earlier 1881 attempt in Potsdam had been undermined by vibrations from passing traffic and footsteps, which blurred the delicate interference pattern. Floating the rebuilt apparatus on a sandstone slab in a trough of liquid mercury let it rotate smoothly and continuously while isolating it from the building's vibrations, which was the key engineering fix that made the 1887 measurement trustworthy.