Michelson–Morley experiment: Did it disprove the existence of ether?

[cragar]

Did this experiment show that there was no ether . Because one way the light had to move up stream and it would take longer. But both beams arrived at the same time , so does this show there is no ether . I'm just trying to make sure I understand this experiment or am I way off .

 

[russ_watters]

It does not show that there is no ether, it shows that the Earth is not rotating through an ether. There are other possibilities such as an ether that is dragged along with Earth that would show the same result, but these are mostly considered superfluous.

 

[Born2bwire]

I think it would be more accurate to say that it showed that there was no ether wind (or if there was it was on an order far smaller than any theories at the time). The presence of an (a)ether stipulated a set of consequences like ether wind, ether drag, and I also seem to recall that movement through a medium affects the index of refraction. MM was just one of the many experiments that chipped away at the properties of various ether theories until you arrive at the point where an ether theory must be exceedingly complex to comply with experimental results.

 

[cragar]

ok so the MM experiment does not prove that light is always perceived at c.
Thanks for your replies by the way .

 

[russ_watters]

ok so the MM experiment does not prove that light is always perceived at c.

No! That is a much different question than your first question - and a poorly phrased one at that. One can never prove anything absolutely, particularly for all situations, so it is nonsensical for that to be the conclusion of an experiment. But the MMX did prove (to a reasonable certainty) that light is always perceived to travel at C...in an MMX.

 

[cragar]

ok I am just trying to make sure I understand the MMX. Now I am not quite sure why it proves that light is perceived at c. Is it because that one of the light beams is moving toward the mirror and then away from it and the other one is moving parallel to the motion. And did both light beams travel the same distance because of length contraction. I really appreciate you helping me understand this .

 

[Born2bwire]

Basically the idea, at least from what I recall (which is getting less and less these days it seems), is that we have a beam of light split off and travel the same distance but along paths of normal incidence. We then recombine the beam and if the time of travel is different then this manifests itself as an interference due to the phase shift. So with the idea of an ether, if we were moving through an ether then speed of light would be dependent upon the path through the ether and our relative motion with respect to the ether. So we run the MM experiment and (ideally) we run this throughout the course of a year. The Earth's orbit moves around the Sun and thus the idea being that our relative motion through any ether will change as the year passes. So we would expect to see a change in the interference pattern between observations but this was not observed.

So, as russ puts it, the experiment only proves that the speed of light is invariant (and c) in the MM experiment which was meant to test a specific set of ether theory ideas. But it was an indicator that the speed of light may be a universal constant.

As for length contraction, I do not think that length contraction came into the original experiment. The apparatus and its results are observed in the same reference frame. The point was that the apparatus (and its observer) may be moving relative to some ether.

 

[cragar]

ok , But when we take into account the Earth turning the one light beam has a longer path to take , But the light beams end up at the interferometer at the same time . And isn't this what gave Hendrik Lorentz his idea for length contraction.

 

[russ_watters]

ok I am just trying to make sure I understand the MMX. Now I am not quite sure why it proves that light is perceived at c. Is it because that one of the light beams is moving toward the mirror and then away from it and the other one is moving parallel to the motion. And did both light beams travel the same distance because of length contraction. I really appreciate you helping me understand this .

The argument is that the light from one arm should move a longer distance because the ether causes it to move at an angle. It's a geometry issue similar to the speed of a boat going across a river: A boat takes longer to go across a faster moving river because it has to angle upstream.

 

[Feldoh]

But when we take into account the Earth turning the one light beam has a longer path to take , But the light beams end up at the interferometer at the same time.

If you do the Michelson and Morley experiment you will find that relative to the spinning Earth the "ether wind" is not moving.

The experiment of itself does not explicitly imply that there is no "ether wind", rather it the experiment implies that the "ether wind" is at rest relative to the Earth frame of reference.

When Lorentz came up with his theory of length contraction he believed there was still an "ether wind" it was just that light happened to become length contracted. In fact, the theory was developed to save the concept of an "ether wind".

However the idea that the "ether wind" is constant relative to the Earth appeals to us as scientists in a different way. It appeals to the idea that we have no reason to be special. If the "ether wind" exists why does it just happen to be here on Earth that it is constant? In much the same way that Copernicus absolved that idea that the Earth was the center of the universe it appeals to the rational mind that in the case of an "ether wind" there is no reason that the Earth should be considered so special as to have an "ether wind" be completely at rest relative to the earth.

 

[cragar]

ok i think i get it now , thanks for your replies

 

[dennisrobert]

MM experiment proves that the speed of light is constant in a specific medium and there is no relativity for it and this was the main principle which lead to Einstein's special theory of relativity. Before the MM experiment, speed of light was thought to be relative to the observer's motion. I don't think it proves that ether DOES NOT exist at all.

 

[Buckleymanor]

It does not show that there is no ether, it shows that the Earth is not rotating through an ether. There are other possibilities such as an ether that is dragged along with Earth that would show the same result, but these are mostly considered superfluous.

Why is this considered superfluous.There are similar circumstances whereby the environment is dragged along and not considered as such.Take the air surrounding the Earth it is dragged along with it but when someone shouts from a distance we don't experience a doppler shift.
Or when a stone is dropped from a height into a still pond.The concentric waves produced are nice and round there is no deformation formed on the side the Earth is being dragged through space.So why was there an expectation that the aether if it existed should behave any different than other substances that are dragged along.

 

[Buckleymanor]

The argument is that the light from one arm should move a longer distance because the ether causes it to move at an angle. It's a geometry issue similar to the speed of a boat going across a river: A boat takes longer to go across a faster moving river because it has to angle upstream.

Nonsense when you consider that it would take a boat longer when it angles upstream in a fast moving river but quicker by exactly the same amount when it travels downstream.
Jump on a moveing walkway and walk in the direction of travell till you get to the end of it.
Consider that the end is where the mirror would be in the MMX.
When you get to the end turn round and walk back at the same pace to where you jumped on.
Time yourself for the journey.
Now walk the same distance without a walkway and time yourself.
The time taken will be equal for both.

 

[Born2bwire]

Nonsense when you consider that it would take a boat longer when it angles upstream in a fast moving river but quicker by exactly the same amount when it travels downstream.
Jump on a moveing walkway and walk in the direction of travell till you get to the end of it.
Consider that the end is where the mirror would be in the MMX.
When you get to the end turn round and walk back at the same pace to where you jumped on.
Time yourself for the journey.
Now walk the same distance without a walkway and time yourself.
The time taken will be equal for both.

No. Assume that you need to travel a distance d and you move with a speed v. Normally, to travel back and forth it will take,

$$t = \frac{2d}{v}$$

Now if you are slowed down by a speed of x on your outward journey and sped up by x on your inward journey, then the time will be

$$t = \frac{d}{v-x} + \frac{d}{v+x} = d\frac{(v+x)+(v-x)}{v^2-x^2} = \frac{2dv}{v^2-x^2}$$

Thus, it will always take you a longer time to traverse the distance. And since the two paths of the MM experiment were orthogonal to each other, they would (except under specific conditions) experience different amounts of ether wind along the paths of travel.

 

[russ_watters]

Why is this considered superfluous.There are similar circumstances whereby the environment is dragged along and not considered as such.

If you look for something in enough places and never find it and it doesn't have any impact on the workings of your equations, then it becomes unnecessary to continue to hypothesize its existence.

Nonsense when you consider that it would take a boat longer when it angles upstream in a fast moving river but quicker by exactly the same amount when it travels downstream.

You misunderstand the geometry: the boat never travels downstream, it only angles upstream in both directions, otherwise it couldn't travel straight across the river and straight back.

Jump on a moveing walkway and walk in the direction of travel...

Perhaps you are unaware of what the experimental setup look like: the goal is to compare the speed of light "along the moving walkway" with the speed of light perpendicular to "the moving walkway".

 

[Buckleymanor]

If you look for something in enough places and never find it and it doesn't have any impact on the workings of your equations, then it becomes unnecessary to continue to hypothesize its existence.

You would not accept that that air or water did not exist because of the lack of impact on the workings of your equations from the examples I gave though.

You misunderstand the geometry: the boat never travels downstream, it only angles upstream in both directions, otherwise it couldn't travel straight across the river and straight back.

Probably did though I took that into account.

Perhaps you are unaware of what the experimental setup look like: the goal is to compare the speed of light "along the moving walkway" with the speed of light perpendicular to "the moving walkway".

Yes I am aware that it is to compare the speed of light perpendicular to the moving walkway.
If you take it that light travels at it's minimum speed in both directions when it is moving perpendicular to the walkway and that any deviation from the perpendicular would result in a loss or gain depending on which direction the light was travelling.
It is possible that the gains and losses cancell out to equal it's minimum speed.

 

[Buckleymanor]

No. Assume that you need to travel a distance d and you move with a speed v. Normally, to travel back and forth it will take,

$$t = \frac{2d}{v}$$

Now if you are slowed down by a speed of x on your outward journey and sped up by x on your inward journey, then the time will be

$$t = \frac{d}{v-x} + \frac{d}{v+x} = d\frac{(v+x)+(v-x)}{v^2-x^2} = \frac{2dv}{v^2-x^2}$$

Thus, it will always take you a longer time to traverse the distance. And since the two paths of the MM experiment were orthogonal to each other, they would (except under specific conditions) experience different amounts of ether wind along the paths of travel.

If we take the points of a compass North South East and West and the centre of the compass as the point where the light originates from in the MMX.
Is it not possible for ease, that when the light is traveling from the centre along the horizontal to the west it gains 1kps and when it travels from the west to the centre it looses 1kps so the value at the centre is 300000kps.
Likewise when it travels east it looses 1kps and when it travels back to the centre it gains 1kps so it's value remains 300000kps.
However when it travels north it does not gain any speed and when it travels back to the centre from the north it also don't gain any speed so it's speed remains at 300000kps at the centre.
If it were to travell to the NW it would gain 0.5 kps and when it travels back to the centre it would loose 0.5 kps.
The reverse would be true of NE and any other directions would be fractions of this.
No matter which direction the light traveled in when it arrived back to the centre the value would be the same but the speeds along the paths different.

 

[ghwellsjr]

ok , But when we take into account the Earth turning the one light beam has a longer path to take , But the light beams end up at the interferometer at the same time . And isn't this what gave Hendrik Lorentz his idea for length contraction.

Yes, this is exactly right. I made a series of animations to illustrate generally what happened in MMX and how Lorentz could have come to his length contraction explanation and how that explanation (along with time dilation) was re-interpreted by Einstein. See these two posts:

https://www.physicsforums.com/showpost.php?p=3059029&postcount=78

https://www.physicsforums.com/showpost.php?p=3059104&postcount=79

 

[Buckleymanor]

Yes I am aware that it is to compare the speed of light perpendicular to the moving walkway.
If you take it that light travels at it's minimum speed in both directions when it is moving perpendicular to the walkway and that any deviation from the perpendicular would result in a loss or gain depending on which direction the light was travelling.
It is possible that the gains and losses cancell out to equal it's minimum speed.

Sorry I can't find the edit button.
I should have said that light travels at a constant when moving perpendicular to the walkway and that any deviation from the perpendicular would result in a loss or gain depending in which direction the light was travelling.etc.

 

[harrylin]

Did this experiment show that there was no ether . Because one way the light had to move up stream and it would take longer. But both beams arrived at the same time , so does this show there is no ether . I'm just trying to make sure I understand this experiment or am I way off .

No. It did show however that the then existing ether theories were wrong. Following those theories, the experimenters took for granted (they even didn't mention it) that Newton's mechanics is exactly correct, according to which the dimensions of an interferometer are completely unaffected by inertial motion.
With the assumption of a stationary ether and by combining Newton's mechanics with Maxwell's electrodynamics, assuming both to be exactly correct, an optical effect was expected from rotating the apparatus. The non-detection of their expected effect was called a "null result", and the solution involved a correction of Newton's mechanics.

You can read their full paper here:

http://en.wikisource.org/wiki/On_the_Relative_Motion_of_the_Earth_and_the_Luminiferous_Ether

Harald

 

[Born2bwire]

If we take the points of a compass North South East and West and the centre of the compass as the point where the light originates from in the MMX.
Is it not possible for ease, that when the light is traveling from the centre along the horizontal to the west it gains 1kps and when it travels from the west to the centre it looses 1kps so the value at the centre is 300000kps.
Likewise when it travels east it looses 1kps and when it travels back to the centre it gains 1kps so it's value remains 300000kps.
However when it travels north it does not gain any speed and when it travels back to the centre from the north it also don't gain any speed so it's speed remains at 300000kps at the centre.
If it were to travell to the NW it would gain 0.5 kps and when it travels back to the centre it would loose 0.5 kps.
The reverse would be true of NE and any other directions would be fractions of this.
No matter which direction the light traveled in when it arrived back to the centre the value would be the same but the speeds along the paths different.

I know this is late but I forgot about this thread. I just demonstrated above that if you traverse the same distance back and forth with a gain and loss in speed then the net effect is that you traverse the distance slower than usual. Because the two paths are orthogonal, the projection of each path onto the ether wind would be different. The only case where the effects to the speeds would be identical would be when both arms are 45 or 135 degrees to the direction of the ether wind. All other cases the ether wind would cause the two beam paths to take differing amounts of time to traverse. The resulting phase delay would have shown up in the measurements of the experiment.

 

[Buckleymanor]

I know this is late but I forgot about this thread. I just demonstrated above that if you traverse the same distance back and forth with a gain and loss in speed then the net effect is that you traverse the distance slower than usual. Because the two paths are orthogonal, the projection of each path onto the ether wind would be different. The only case where the effects to the speeds would be identical would be when both arms are 45 or 135 degrees to the direction of the ether wind. All other cases the ether wind would cause the two beam paths to take differing amounts of time to traverse. The resulting phase delay would have shown up in the measurements of the experiment.

Would this still be the case if there was no aether wind and only extra speed provided by the movement of the Earth through the universe.

 

[Born2bwire]

Would this still be the case if there was no aether wind and only extra speed provided by the movement of the Earth through the universe.

Technically yes, but only because the Earth is accelerating. If the Earth were moving along a constant path then it would not matter since both beams would get the same boost (assuming Galiean transformations which was what they were working with at the time). But the error (if it is even large enough to be considered) from the ever so slight acceleration of the Earth during the time of travel can be accounted for. More importantly, this would also be canceled out by the fact that the Michelson-Morley experiment was repeated again and again at different positions in the Earth's orbit. The experiment was always interested in the relative changes in measurements. Looking at the relative phase shifts between the two paths, looking at these phase shifts relative to the measurements taken along different orientations, looking at these measurements relative to the measurements taken at different times of the year. If the ether wind existed, it would have produced deviations in these results due to the changes in the relative motion of the wind between experiments.

 

[harrylin]

Technically yes, but only because the Earth is accelerating. If the Earth were moving along a constant path then it would not matter since both beams would get the same boost (assuming Galiean transformations which was what they were working with at the time). But the error (if it is even large enough to be considered) from the ever so slight acceleration of the Earth during the time of travel can be accounted for. More importantly, this would also be canceled out by the fact that the Michelson-Morley experiment was repeated again and again at different positions in the Earth's orbit. The experiment was always interested in the relative changes in measurements. Looking at the relative phase shifts between the two paths, looking at these phase shifts relative to the measurements taken along different orientations, looking at these measurements relative to the measurements taken at different times of the year. If the ether wind existed, it would have produced deviations in these results due to the changes in the relative motion of the wind between experiments.

As I elaborated earlier: that's what they thought, based on their assumption that Newton's mechanics is correct. However, that assumption has been disproved.

 

[Born2bwire]

As I elaborated earlier: that's what they thought, based on their assumption that Newton's mechanics is correct. However, that assumption has been disproved.

Nobody is arguing that they weren't.

 

[harrylin]

Nobody is arguing that they weren't.

So you meant that they thought that "If the ether wind existed, it would have produced deviations in these results due to the changes in the relative motion of the wind between experiments."

Sorry to have misunderstood you!

 

[Janus]

Why is this considered superfluous.There are similar circumstances whereby the environment is dragged along and not considered as such.Take the air surrounding the Earth it is dragged along with it but when someone shouts from a distance we don't experience a doppler shift.

Astronomical observation had already ruled that out. In essence, When we look at distance stars, the light from them comes in from angle due to the Earth's relative motion with respect to them. Again, since the Earth revolves around the Sun, its motion with respect to these stars change over the course of a year and so does the angle.

Now if there is an aether and it was being dragged with the Earth, light coming from distant stars upon reaching the Earth would tend to be dragged along with the aether. That would tend to eliminate the above mentioned shift in angle that we see.

Since the angle we measure matches what we would expect for the Earth's orbital speed, we can rule out aether drag.

 

[thwle]

The usual interpretation of MMX is biased both by the intended purpose addressing the concept of a luminiferous ether and by the subsequently enunciated special theory of relativity.
MMX does not involve a clock or measurements of time, it does not measure the speed of light, and it does not compare observations of observers in relative motion.
The source of light in MMX moved with the apparatus so an emission theory ("Light speed is relative to its source.") might be taken as the explanation. I have little knowledge of such theories, myself.
But even if we knew for certain that emission theory was wrong and that light speed was known to be relative to an inertial frame of reference, homogeneous and Euclidian in the familiar way, there is a simple and satisfactory way to interpret MMX.
MMX deals in effect with a comparison of two independent methods of measuring distance: (1) distance measurement by a solid material measuring stick, (2) distance measurement by echo ranging using an electromagnetic signal. Two uncalibrated measuring sticks (not necessarily of equal length) arranged approximately normal to each other are used as echo paths for two parts of a split light beam and the interference of the recombined beams is observed. That this interference pattern does not shift when the apparatus is turned 90 degrees demonstrates that the correspondence of these two methods of measurement is undisturbed by changes of orientation or by changes of motion in the solar frame of reference.
If we assume emission theory is false, and anticipate that the two measurement methods will be found altogether equivalent, then we must say that the interatomic and intermolecular distances within a solid (measuring stick) are constant (on average) relative to electromagnetic echo ranging between them. Then, it becomes extremely tempting to propose that the latter is the cause of the former: that the distances between the atomic constituents of solid matter are mediated by forces that operate in a manor akin to radar ranging. This, of course, would account for Lorentz contraction.

 

[Buckleymanor]

Technically yes, but only because the Earth is accelerating. If the Earth were moving along a constant path then it would not matter since both beams would get the same boost (assuming Galiean transformations which was what they were working with at the time). But the error (if it is even large enough to be considered) from the ever so slight acceleration of the Earth during the time of travel can be accounted for. More importantly, this would also be canceled out by the fact that the Michelson-Morley experiment was repeated again and again at different positions in the Earth's orbit. The experiment was always interested in the relative changes in measurements. Looking at the relative phase shifts between the two paths, looking at these phase shifts relative to the measurements taken along different orientations, looking at these measurements relative to the measurements taken at different times of the year. If the ether wind existed, it would have produced deviations in these results due to the changes in the relative motion of the wind between experiments.

I am slightly confused.
You say technically yes, but only because the Earth is accelerating.
If the Earth were moving along a constant path then it would not matter since both beams would get the same boost.
Ghwellsir mentions that.

They believed that they were constantly moving with respect to the ether and also constantly changing their velocity through the ether as the Earth rotated on its axis and as it revolved around the sun. This constant acceleration was very small so for all practical purposes, they could assume that they were moving at a constant speed through the ether.

So was this the reason for the null results that for all practicle purposes they were moving at a constant speed and any gains or losses were practicaly canceled out.

 

[Born2bwire]

I am slightly confused.
You say technically yes, but only because the Earth is accelerating.
If the Earth were moving along a constant path then it would not matter since both beams would get the same boost.
Ghwellsir mentions that.

It doesn't matter if the Earth was moving along a constant path because the observer and apparatus are moving with the Earth. In that case, the entire experiment is performed and observed within the same frame. Assuming Galilean transformations, the movement of the light source will impart an extra boost to the light and this would cancel out the effects of the moving apparatus. Only when we consider Lorentz transformations does this cause a difference since the speed of light is the same in all frames. You might as well ask yourself if you threw two baseballs back and forth between two partners whether the exercise would be different if you were stationary or on a uniformly moving platform. In non-relativistic assumptions, there would not be a difference from your point of view regardless of how you observed it, but things would change if the platform was performing a wide turn.

If the Earth is accelerating, then we can see that the apparatus will move while the beams are travelling. So this can cause a slight shift in travel times because the apparatus will have moved while the beams are traveling thus changing the paths that need to be travelled. So we can see that this would affect the experiment even if we assume Galilean transformations.

The thing to remember here is that the experimentalists and theorists were still assuming Galilean transformations.

 

[harrylin]

I am slightly confused.
You say technically yes, but only because the Earth is accelerating.
If the Earth were moving along a constant path then it would not matter since both beams would get the same boost.
Ghwellsir mentions that.
So was this the reason for the null results that for all practicle purposes they were moving at a constant speed and any gains or losses were practicaly canceled out.

Yes - the Michelson interferometer is by design insensitive to rotation, because the enclosed area is nearly zero (in contrast to Sagnac and Michelson-Gale).

 

[ghwellsjr]

I am slightly confused.
You say technically yes, but only because the Earth is accelerating.
If the Earth were moving along a constant path then it would not matter since both beams would get the same boost.
Ghwellsir mentions that.
So was this the reason for the null results that for all practicle purposes they were moving at a constant speed and any gains or losses were practicaly canceled out.

Yes - the Michelson interferometer is by design insensitive to rotation, because the enclosed area is nearly zero (in contrast to Sagnac and Michelson-Gale).

There are actually several different issues going on here:

1) Maxwell, as well as most other scientists of the time, incorrectly believed that his equations supported the detection of a fixed absolute ether rest frame. They did not realize that his equations were Lorentz Transformable and that they actually predicted the null result of MMX.

2) Michelson and Morley designed an experiment, based on their lack of knowledge, that was supposed to detect an ether wind whenever the surface of the Earth (wherever MMX was located) was moving through the ether. But since they didn't know when or if the Earth ever was stationary with respect to the ether, they relied on the fact that the Earth was constantly accelerating, but at an insignificantly low rate, so that if they happened to be stationary in the fixed ether at one point in time, it certainly would be moving through the ether 12 hours later and/or six months later, assuming that the Earth was not dragging the ether along with it.

3) As has been pointed out by Born2bwire earlier in this thread, even if they were moving through the ether, they realized they still would get a null result if their apparatus were aligned so that the two arms of the experiment happened to be at a 45 degree and 135 degree angle with respect to the direction of motion. So they designed their experiment so that it could be rotated very slowly, about one rotation per minute. They then believed that during the course of one revolution of the appartus, they should hit four null spots and two positive peaks and two negative peaks, as long as there was some ether wind. During the course of one revolution, even though the surface of the Earth was accelerating, it wasn't enough to affect the data during one revolution or even several revolutions.

4) Since they always got a null result, in other words, they never could detect an ether wind, Michelson believed that the Earth was dragging the ether along with it. Later, Lorentz discovered that if the ether were causing the lengths of the apparatus to contract along the direction of motion through the ether, this would explain the null result.

5) Finally with the addition of time dilation, the full Lorentz Transformation equations were developed and they realized their original mistake in interpreting Maxwell's equations and they fully understood why MMX produced a null result, all within the context of a continued belief in a fixed, absolute ether that was not dragged by the earth. This interpretation is called the Lorentz Ether Theory and is fully compatible with all experiments. However, they still could not identify the absolute ether rest frame but they believed that their own clocks were dilated and their own rulers were contracted and constantly changing (at a very low rate) as the surface of the Earth was changing direction and speed through the ether on a daily and seasonal basis.

6) Einstein promoted the idea of assuming that any inertial observer was at rest with respect to the ether and everyone else who was moving with respect to that observer was experiencing the time dilation and length contraction. Of course, he didn't word it precisely that way, but that is the equivalent of what he was saying.

 

[harrylin]

There are actually several different issues going on here:

5 x yes

[..]

6) Einstein promoted the idea of assuming that any inertial observer was at rest with respect to the ether and everyone else who was moving with respect to that observer was experiencing the time dilation and length contraction. Of course, he didn't word it precisely that way, but that is the equivalent of what he was saying.

If that is the equivalent of what he was saying at the time, then it was self-contradictory - some people indeed accused him of that, but not Lorentz who knew him rather well.

I would say that Einstein promoted the idea that we should not introduce phantoms that seem to escape us - the "shut up and measure" approach. Later he changed his opinion and became less positivistic, but his approach of that time had great impact.

 

[ghwellsjr]

If that is the equivalent of what he was saying at the time, then it was self-contradictory - some people indeed accused him of that, but not Lorentz who knew him rather well.

In his 1905 paper, he stated, in effect, that it was apparently self-contradictory, but that it wasn't really.

So are you agreeing with Einstein that it is only apparently self-contradictory or do you think it is actually self-contradictory?