The Michelson experiment and its family of misunderstandings

Dierck-E.Liebscher, Potsdam

There are only a few experiments in physics which are cited by name in textbooks of general education. The Michelson experiment is the most known, because it is rated to be the foundation of the theory of relativity, which carries the attributes difficult and ingenious. In these rationales, didactic reductions are hidden which are up to 100 years old. Their disentanglement poses important questions and yields surprising answers.
   Whenever you already heard something about, you know that Michelson intended to measure the headwind in the ether which is felt by the earth on its orbit around the sun. You also know that the theory of relativity is referring to it because the expected headwind was not found. You will not have been informed, why you need an ether at all, why the headwind was expected in a basement, and why this all has to do with relativity.
There is no need for an aether to follow the guiding thread to relativity. Nevertheless, the aether concept had both a leading and a misleading impact on the discussion.

What is the ether, what is his use?

We usually read that the ether must be invoked to explain the propagation of light waves. Light waves, we read, need a carrier medium whose oscillations constitute light and in which they propagate as waves. However, from a severe Newtonian point of view, this is not the case: The gravitational field of Newton propagates through void space without any carrier. The desire for mechanical models of the intermediate steps for a theory is not part of the scientific method.
   A mechanical model like the ether is not necessary for the time being, although it is suggested by the expected analogy to the propagation of acoustic waves. As we shall see in a moment, one has every reason to suppose an ether.

Misunderstanding No 1: The ether breaks the principle of relativity, because it yields a universal object of reference.

The opposite is true. The principle of relativity states that any velocity needs an object of reference. There is no velocity against void space. However, when one supposes the composition of velocities to be additive for wave propagation too, then we obtain exactly one state of motion in which the propagation is independent of the direction. The principle of relativity (attributed to Galileo) requires that this distinction is due to a material object of reference, which was later called ether. Any anisotropic propagation then indicates a motion relative to this ether. Galileo's principle is restored.
   Distinct universal reference objects do not violate the relativity of velocity. In astronomy, we know different universal reference objects, for instance the background radiation, or the large-scale distribution of galaxies and quasars. Our velocity with respect to these objects is about 500 km/s. There is no connection with the laws of the propagation of light.

Why Michelson hoped to find the headwind of the ether in a basement of all places?

At first, one expects that a material medium can be influenced somehow, can be stopped by appropriate walls. The question whether this ether should exhibit other physical properties or whether a mechanical model existed was subject to scientific discussion for a long time. This discussion turned out to be fruitless. The search for mechanical and other physical properties led to contradictions or to nothing at all. Far worse: A small effect, the aberration of starlight was observed. This aberration is usually explained with help of an umbrella in the rain, inclined forward when we walk. When on starts motion, the souces of emanation seem to shift toward the direction of motion. The effect for starlight is small, v/c approx 10-4.
   However, light is a wave, and when simultaneity is not questioned, the direction of the wave fronts does not change: Light waves do not show any aberration of wave fronts. Fresnel found an excuse. He referred to the fact that a telescope does not measure wave fronts, but the motion of particle-like wave groups cut out by the aperture of the telescope. They behave like crests on a wave and show aberration. However, the carrier of the waves must not be disturbed, stirred or stopped by the structure of the telescope, by walls, or even by the earth itself. This was difficult to swallow for Fresnel's contemporaries, more difficult than for us, who know about neutrinos. Hence Michelson intended to observe an ether which can pass the walls, and he installed his equipment in a basement.
    He did not find an ether which does. He tried again, and he tried to measure in free air, on mountains and ballons, all was in vain. The ether is included by the walls like air, it is retained by the earth like the atmosphere. Fresnel's explanation of the aberration falls to pieces.

Misunderstanding No 2:
The result is explained through the Lorentz contraction of length

This is spread in lots of books. However, when one accepts Einstein's Axiom of composition of velocities with the speed of light, the experiment is not expected any more to show some effect. Any further explanation is superfluous and wrong (any wife knows: when the husband produces two excusions for coming late, at least one of them is wrong). Any length contraction must be supposed only in case we stick to the additive composition of velocities in order to avoid the relativity of simultaneity. But in this case too, the length contrasction is no explanation, because only the interferometer is appropriately precise to measure it .

Misunderstanding No 3:
The Michelson experiment shows that no ether exists.

One can read this statement occasionally, but the opposite is true. A freely floating ether is equivalent to a universal object of reference. If no ether existed, and velocities are composed by addition, the relativity of velocities could not be valid any more. In this case you must find an anisotropy on the earth. If there is no anisotropy, then an ether must exist, an ether carried along by the walls of the basement respectively the earth itself like the atmosphere. Michelson's conclusion was exactly this. We remember, that the additive composition of velocites is still presupposed. That this premise cannot be sustained was shown by the theory of relativity 24 years later.

Misunderstanding No 4:
Einstein concealed his knowledge about the Michelson experiment.

This was the excitement of the scientific yellow press in 2005, but it is wrong. Einstein cited the Michelson experiment not explicitly, but in summary with the other attempts to observe the ether, and Einstein did not aim at an explanation of this experiment, or the aberration in particular. The latter was the topic of Lorentz and Drude, who found the necessity of a formal time (before 1900) in which simultaneity becomes relative, in which two observers in relative motion differ in assessing events as simultaneous. The relativity of simultaneity and the now famous Lorentz transformations are a consequence of the presupposition of an aberration of a wave front itself.
   Einstein was interested in the establishment of simultaneity and in the Maxwell's equations, which obtained curious and intricate forms when combined with additional velocities. Maxwell's equation contain a constant which is a speed without direction, a speed which cannot composed with a directed velocity in a sensible way. This speed without direction was already found in an experiments of 1854 by R.Kohlrausch und W.Weber. They realised the concept of C.F.Gauß to define electromagnetic units by mechanical means. Here is the point to which Einstein's axiom is linked: A velocity (the velocity of the propagation of light) exists whose size does not change in composition with other velocities.
    After Michelson had won the Nobel Award in 1907 for his interferometer technics, the elementary explanation of the result of the experiment by the theory of relativity obtained a new relevance. This explanation turned out be the simplest access to the theory, and it was acknowledged and hailed in this function by Einstein. Who had no interest in the logical structure of a proof, now saw in the Michelson experiment the proof of Einstein's axiom. Rightly so, Michelson objected this connection for all his life.

Misunderstanding No 5:
The Michelson experiment proves the constancy of the velocity of light.

This is impossible at all. First, Einstein's constancy of the velocity of light only means that its modulus (the speed) does not vary in compositions. Variability in space and time is another question. Second, an experiment never proves hypotheses.
    In gereral, no experiment can prove a theory: A theory has to be valid for uncountably many cases which never can be tested. An experiment can have two results. First, the theory may explain the result, then it does not contradict the theory. If the result has been predicted against prejudice, the experiment supports the theory, of course, but support is not proof. If the result is not compatible with the theory, it is a proof: Then it demonstrates that the theory is wrong, or contains assumptions which have to be dropped or changed. The Michelson experiment with its "negative" outcome constitutes such a proof: It shows that the conclusion of a freely flowing ether uses assumptions which are cannot be applied.
    Experiments may inspire intelligent hypotheses and test them at best. Any test is a comparison and decision between alternative hypotheses. The result of a test depends on the admitted hypotheses. Without hypotheses, a test does not show anything.

Misunderstanding No 6:
The theory of relativity relies on the correct result of the Michelson experiment.

This error lets hundreds of amateurs believe that they could refute the theory of relativity through the construction of hypotheses which produce Michelson's null result, or which analyse the experiment from another point of view. These attempts are in vain because it is not the the primary aim of the theory of relativity to explain the Michelson experiment. The theory of relativity is backed by far stronger experience.

Misunderstanding No 7:
Relativity produces only effects of second degree (in v/c).

The effect which Michelson intended to observe really is of second order in v/c. The popular notions of the length contraction, the time dilation, the variability of the inertial mass with velocity, et cetera are all of second order. Therefore, one reads occasionally that all relativistic effects are of second order. This is wrong.
    First, the aberration of light waves itself is an effect of first order: It is proportional to the velocity of the observer, and it vanishes when velocities are composed via simple addition. Second, the ether drift could be measured also in first order. This has been done in an experiment of G.R.Isaak and his fellows, which reached a precision of 10-10 in v/c.
    Finally, we observe effects of zeroth order, not depending on any observers motion, where the pure structure of the theory led to predictions which were confirmed brillantly. I refer to the prediction of the spin of particles (kind of an angular momentum, which yields twice the magnetic moment than the familiar angular momentum), and to the prediction of antiparticles (an particle-antiparticle pair may decay into only photons). Both predictions are characteristic for the theory of relativity (and the Lorentz transformations in particular). Hence they are the most important footholds of the theory of relativity. Upto now, no alternative theory yielded these predictions.