Do you agree that any experiment with two-way there-and-back journeys will never show any discrepancy? Aren't all differences going to be counterbalanced, levelled in the journey back?.

  • 2
    $\begingroup$ If there were an absolute reference frame for the speed of light, then the two-way round-trip times do get affected by the orientation of the MM interferometer. This is mathematically shown here. The counterbalance for the backward motion doesn't completely cancel out the extra time for the forward motion. If the equations in the link aren't clear, you should consider a numerical example. $\endgroup$ Jan 12 at 10:27
  • $\begingroup$ @MaximalIdeal, right, now suppose there is an aether, what would the experiment show? $\endgroup$
    – user157860
    Jan 12 at 10:36
  • 4
    $\begingroup$ I’m voting to close this question because it is not a question but a provocative statement looking for validation. $\endgroup$
    – noah
    Jan 12 at 11:04
  • $\begingroup$ Just like sound travel at different speed for upwind and downwind, they want to do the same thing for light in ether but they are surprised the result don't match the model so the only conclusion is what we know today ;D $\endgroup$
    – user6760
    Jan 12 at 12:50

2 Answers 2


All differences are going to be counterbalanced, levelled in the journey back.

I think you have misunderstood the Michelson-Morley experiment. It does not attempt to measure the time that light takes to travel along one arm of the interferometer and back. Instead it shows that the difference in travel time along the two perpendicular arms of the interferometer is the same for all orientations of the interferometer. And the only way this can happen is if the speed of the aether (if it exists) relative to the experimental apparatus is zero.

This could happen to be true (by coincidence) on one day in the year. But Michelson and Morley repeated their experiment over a period of over three months. In this period, the Earth's motion in its orbit should have caused a non-zero speed relative to aether at some point in these three months. But no movement of the interference lines was observed.

  • $\begingroup$ your example doesn't fit, because you are fighting against a force, which is alien to our issue, I suggest you consider examples of motions inside a train going at constant speed. If the aether existed such an experiment would not show it. $\endgroup$
    – user157860
    Jan 12 at 10:21
  • $\begingroup$ @user157860 Assuming the existence of the aether, would the orbit and rotation of the Earth not mean that we are accelerating with respect to said aether? $\endgroup$ Jan 12 at 10:32
  • $\begingroup$ now suppose there is an aether, what would the experiment show? $\endgroup$
    – user157860
    Jan 12 at 10:40
  • $\begingroup$ air is the aether for sound, if we send a sound from an ambulance in the direction of motion and in a normal direction and back, would the experiment show any difference in speed? $\endgroup$
    – user157860
    Jan 12 at 10:54
  • 1
    $\begingroup$ Yes of course. Have you not seen the Doppeler formulae for moving sound source and moving air? $\endgroup$ Jan 12 at 11:02

The Michelson-Morley experiment wasn't designed to rule out aether.

It was designed to establish it, according to the terms of how the aether was then conceived. What it showed instead was that the aether could not have the properties ascribed to it at the time.

Einstein later showed that for the purpose of his theory of special relativity, it was not necessary to ascribe any properties at all to free space to explain the situation, and that it could instead be explained purely by movement in time (within a conceptual framework of spacetime).

With his theory of general relativity however, Einstein did concede a return to the postulation that free space has properties, at least concerning gravitational potential. But here gravity is conceived as causing warps in time and space itself, as part of a fundamental relation between matter and spacetime. It doesn't overturn the conceptual model of SR, but extends it further.

By that time, the main proponents of aether theory - of which HA Lorentz is regarded as the foremost - had modified their conception of the aether so as to be wholly compatible with the same evidence as was Einsteinian theory and to make the same predictions.

However, Lorentzian theory was by that time suffering a number of ideological challenges.

The first was that the Lorentzians were seen to be staggering from hypothesis to hypothesis to explain new evidence as it arose - "adjusting the auxiliary belts around the hardcore", to employ Lakatosian language. Einstein's approach had already made predictions ahead of evidence being found to support it, and was somewhat simpler to reason about (which is probably why certain implications were easier to see).

The second was that, in its final incarnation, Lorentzian theory conceded that the aether was not independently measurable, except through the effects it was postulated to explain - the same effects that Einstein did not postulate any aether at all to explain. This was a problem in an era when science was attempting to define itself purely in terms of what it can measure.

The third problem was that the Lorentzians lacked any theory to explain the full welter of relativistic effects in terms of specific mechanical interactions with the aether, which Einstein explained quite straightforwardly in terms of the relativity of time, and in terms of a fundamental link between the geometry of spacetime and gravity. The whole point of the aether in Lorentzian theory was to explain the situation in mechanical terms, so in this sense its proponents were failing even on their own terms.

In summary, Lorentzian theory suffered serious problems of credibility amongst scientists, at a crucial time when an alternative theory - Einstein's - was suffering no credibility problems at all. The Lorentzian theory was never falsified, but work on it was largely discontinued due to the philosophical objections to the unmeasurable aether, and the difficulty of seeing how it should be extended even by those who did not object to it in principle. HA Lorentz himself died in 1928, and most similar advocates of aether were of the same generation as Lorentz.

Developments in subsequent decades showed that the philosophical objections to aether had not been justified, but by that time the Lorentzian approach had already failed to reproduce committed interest amongst professional scientists, whose familiarity with it in recent times often doesn't extend far beyond believing it to be wrong.


Not the answer you're looking for? Browse other questions tagged or ask your own question.