Questions to the Michelson experiment


Reconstruction in the basement of the eastern dome of the AOP (now PIK) on the Telegrafenberg

The original place

In Germany, at most six experiments which are individually named are explained in school textbooks for physics. One of them is the Michelson experiment. It is cited even in most elementary presentations of the theory of relativity. The place where it was successfully performed for the first time is the basement under the eastern dome of the former Astrophysical Observatory Potsdam, which now houses the Potsdam Institute for Climate Impact Research.

Albert Abraham Michelson developed techniques of interferometry, which was expected to be able to determine any ether head wind that would be experienced by the earth on its orbit around the sun. Like a swimmer in a river, light would propagate in and drift with this ether wind, and the observation of this drifting should exhibit the motion of the earth around the sun.

Michelson could not find any drifting of light. Later on he characterized this result as negative, and Einstein hailed it as foundation of the theory of relativity. Literally understood, both statements provoke misunderstanding.


Why an ether should exist?

Isaac Newton understood light as an emanation of particles which propagate with high velocity from its source. With todays knowledge, this is rather obvious. In fact, J.Bradley found in the years after 1725 the aberration of starlight. This is kind of an umbrella effect: The incident direction of a stream of particles seems to shift into the field of view when one starts to move. Who starts to move during rainfall has to incline his umbrella forward.

Bradley found that, as Newton expected, the light behaves like a stream of particles. Its incident direction changed with the velocity of the earth on its orbit. This effect, the aberration of star light, was acclaimed as backing of Newton's emanation assumption and of Copernicus' thesis of the motion of the earth as well.

When Thomas Young and later Auguste Fresnel taught the wave theory of light because only a wave could explain interference, it was no problem to explain reflection and refraction through Huygens's construction which we know from school. Because of the success of mechanics, a mechanical model of light propagation was envisaged and a carrier medium of these waves was conjectured.

In addition, since Galilei it was understood, that the velocity of a closed room can only be determined by observing reference object outside (principle of relativity). When one composes velocities by simple addition, oncoming light should be faster than overtaking light. It seemed possible to measure a velocity simply by these differences. When one is bound to rescue the relativity principle, a substantial reference object must exist. It should be the ether.


Why a basement room?

The wave theory could explain interference phenomena (reflection and refraction are explained with a particle's motion too), but the aberration of starlight had disappeared. When the rain falls in horizontal wavy fronts, the incident direction of particles will change, but not the incident direction of the front. The wave fronts do not refer to some motion but to an instantaneous state. Consequently, their observed normal is not influenced by the motion of an observer. Fresnel rated the lack of aberration of wave fronts as the main problem of wave theory until he found the excuse: A telescope does not determine the inclination of a wave front, but the propagation of a small section which the aperture stop of the telescope cuts out of the wave. This section propagates like the crest on a wave and shows aberration again. However, this picture only works when the telescope's material structure does not influence the carrier of the wave field. The ether flow should never be disturbed by matter. When the ether would be disturbed by the structure of the telescope and its vicinity, the determination of the apparent position of a star would be an extremely difficult task in general. If Fresnel would have had access to recent skill of determining wave fronts, he had found the entry to the theory of relativity, the same which Lorentz found after Michelson's experiment (Applet Aberration).

Michelson demonstrated with his experiment, that the hypothesis of the ether being undisturbed by matter could not be sustained. The existence of the ether was not questioned. As everydays prejudice expects, the ether is kept at rest in a closed room like the air. Only Fresnels explanation of the aberration of starlight breaks down. In Michelson's wording:

The interpretation of these results is that there is no displacement of the interference bands. The result of the hypothesis of a stationary ether is thus shown to be incorrect, and the necessary conclusion follows that the hypothesis is erroneous.
This conclusion directly contradicts the explanation of the phenomenon of aberration which has been hitherto generally accepted, and which presupposes that the earth moves through the ether, the latter remaining at rest.


Why Einstein did not cite the Michelson experiment in his construction of the theory of relativity?

Einstein well cited the many ether experiments, but in summary, not individually.
The demand to try a model in which a wave front itself shows aberration had already led to relativity of simultaneity. Lorentz found it before 1899, but he rated this finding as just another excuse and called his new notion of time a formal time only. In this situation, doubts were raised whether Galilean relativity should be valid for mechanics only, not for the interplay of mechanics and electrodynamics. This meant to abandon the ether model.
It was the rethinking of the determination of simultaneity and of the establishing of synchronicity, which led Einstein to his flash of inspiration: To put the speed of light as axiomatically independent of the motion of the source and of the observer as well, and to reinstate Galilean relativity in full. The speed of light is supposed not to depend on the direction and not to change its absolute value in composition with other velocities.

Now every problem dissolves. It is simple to derive the aberration of wave fronts, and it becomes clear how the relativity of simultaneity needs to be understood. The result of the Michelson experiment obtained the simplest possible explanation. No ether had to be invoked any more.

When Michelson won the Nobel Prize in 1907 (for the method, not for the particular experiment), the acceptance of theory of relativity was boosted to such a degree that the Michelson experiment was seen as basis of the theory. Michelson resisted this interpretation all his life, and he was right as far as the straight link is meant. The Michelson experiment refutes only Fresnel's construction of an ether which streams unhampered. It does not show the general independence of the speed of light from direction (and from space and time the less).

No experiment can prove the viability of a supposition beyond the circumstances of the experiment. It may back the hypothesis when it meets the expectation, but this is no proof. A real proof arises only in the case when it does not meet the expectation. Then it proves that something is wrong with the hypothesis and its interpretation.

The result of the Michelson Experiment alone does not compel the theory of relativity. It is the aberration of star light which makes the impact: When one insists that the aberration of a rain of photons is equivalent to the aberration of wave fronts we obtain the relativity of simultaneity, and with the axiom of unrestricted validity of the relativity of velocities, we obtain all the kinematics of the theory of relativity.


The experiment

Through observation of the interference pattern of two reflected images of the same light source, relative changes in the two paths can be determined with high precision. The light rays from the source are split, so that one image arises through a path along the motion of the earth and back, the other through a path lateral to this motion (Applet Versuch).
When the apparatus is turned, the paths change corresponding to the position of the arms. The interference pattern should show this. However, the effect did not appear. The changes in the interference pattern were smaller than expected and were supposed to present ordinary errors.


The relativity of simultaneity

When the instantaneously measured position of a moving front depends on the motion of the observer, this dependence can only hide in the notion instantaneously. When an observer sees a front coming down parallel to the ground, all events of touching the ground are simultaneous. When another observer sees another direction of the front, these events are no more simultaneous: Whether two events have to be considered simultaneous depends on the state of the observer. Simultaneity becomes relative.

Einstein's axiom of a composition law for velocities which leaves the speed of light untouched implies a relativity of simultaneity consistent with the aberration of star light.


The absolute speed of light

When velocities are composed by simple addition, no velocity of fixed absolute value exists. Einstein's axiom implies to refute such a rule of composition.

About 50 years before the theory of relativity was constructed, and 25 years before Michelson could perform his experiment, there was an experiment which defined a velocity without direction. This means that a velocity was defined which could not be composed with any directed velocity at all. This velocity had to be constant.

It is the experiment of R.Kohlrausch and W.Weber concerning the definition of the unit of electrical charge by mechanical means. Since the definition of the International System, this experiment is nearly forgotten, although its result is the only one which immediately backs Einstein's axiom.

At least in principle, the electric charge can be determined through two procedures. First, one may use the force between static charges, second, we may measure the force between unit currents (The latter is implicit in the actual definition of the Ampère). Between the two units, there is a factor of proportionality which has the dimension of a velocity, but no direction, of course. This factor could be measured in 1856 only crudely, and it was found to be of the order of the velocity of light. It is the velocity of light, Maxwell stated and used it in his famous equations.

Taking the result more serious than one could dare in 1856, it shows the existence of a directionless and therefore constant velocity, exactly like that which Einstein's axiom postulates.


Michelson,A.A.: The relative motion of the Earth and the luminiferous ether, The American Journal of Science 22 (1881), 120-129.
Michelson,A.A.: Die Relativbewegung der Erde gegen den Lichtäther, Deutsche Übersetzung mit einem Vorwort von A.H.Compton und einem Nachwort von M.v.Laue, Die Naturwissenschaften 19 (1931, Heft 38), 777-784.
Bleyer,U., Gottlöber,S., Haubold,H.-J., Hempelmann,A., Mücket,J.-P., Müller,V., Stoll,D.: Zur Geschichte der Lichtausbreitung, Die Sterne 55(1979), 24-40.

About the historical context:
P.Brosche, D.-E.Liebscher: Fallstricke der Aberration

About recent developments:
Tests of Special Relativity and
What is the experimental basis of Special Relativity?