If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

Michelson–Morley Experiment introduction

Simple idea behind the Michelson‒Morley experiment to detect ether wind.

Want to join the conversation?

  • leaf grey style avatar for user Austin Jäger
    If you have a fiber-optics cable running the circumference of the Earth at the equator, would a signal traveling with the rotational velocity of the earth complete one circuit in the same amount of time as a signal traveling against the rotational velocity of the earth? Is the same true at the scale of the solar system, galaxy, galaxy cluster, etc. ?
    (27 votes)
    Default Khan Academy avatar avatar for user
    • female robot ada style avatar for user Rodrigo Campos
      Depends on who's observing. Suppose you're on the equator and sends a signal travelling both directions. Since you sent the signal simultaneously and both rays have the same angular speed relative to you , then they'll perform a lap (360 degrees) simultaneously.
      However, if you observe this from outside the earth, simultaneity is lost. Both rays have the same speed, but this time you can see that the equator is moving, so they'll perform a lap in different times.
      Keep in mind that simultaneity is not absolute.
      (23 votes)
  • piceratops ultimate style avatar for user Jim & Jennie
    Did the Michelson-Morley experiment fail because it was unable to detect the effect of a luminiferous ether or because they failed to understand that light travels at a constant speed regardless of the reference point?
    (11 votes)
    Default Khan Academy avatar avatar for user
    • leaf green style avatar for user Mark Zwald
      Failed for both those reasons.
      1. If there was a universal frame of reference (ie. ether), then their experiment would have detected a change in speed of light relative to Earth's motion.
      2. Since special relativity was unknown to them, then there was no reason to believe ahead of time that spacetime changes relative to motion to keep the speed of light constant in all reference frames.
      (13 votes)
  • hopper cool style avatar for user Luis Soares
    If the speed of light speed up when travel in the same directions of the ether and slow down when it does the opposite travel, the blue photon will reached the the center, the middle mirror at same time that the orange photon does. If this is truth, both waves will reached the detecter at same time. So no way to prove the existence of the luminiferous ether. What is wrong in this my interpretation?
    (11 votes)
    Default Khan Academy avatar avatar for user
    • female robot ada style avatar for user Rodrigo Campos
      Let's say that in ether light would move at c. Also, consider Earth moving at a velocity v relative to the ether, and that, at some day, the ray that was drawn vertically is parallel to v (this last consideration is just for simplicity). Furthermore, the distance travelled by this ray from the middle to the other mirror is s1, while the ray that did not reflected at first travelled s2 until it reached the mirror.
      time it takes for the reflected ray to reach the middle again, with ether: s1/(c+v) + s1/(c-v)
      time it takes for the reflected ray to reach the middle again, without ether: s1/c + s1/c
      s1/(c+v) + s1/(c-v) =/= 2.s1/c
      Thus, it does make a difference.
      However, that's not the important thing about the experiment, testing this only once would prove nothing, since you don't know the speed of Earth relative to the ether, nor the speed of light (with enough precision).
      What matters is the fact that you can spin the apparatus around and observe no change at all. Were there a luminiferous ether, the difference in time taken for light to travel through s1 and s2 would change according to the orientation of the apparatus. To explain further, let's say the reflected ray takes a time Tr to hit the middle mirror and come back, while the other ray takes a time Tp to do the same. For each value of Tr - Tp, you would obtain a unique wave pattern. So, let's say in a given day, you measure a wave pattern corresponding Tr - Tp = 2 nanoseconds (random number). If you were to realize the experiment with the same equipment 3 months later, due to Earth's movement, you should measure a different value of Tr - Tp, because the rays' orientation relative to the ether would be totally different, thus giving you a different wave pattern. Or, more pragmatically, you could just spin the apparatus around, and you would still have no change at all.
      (2 votes)
  • orange juice squid orange style avatar for user Jan Hense
    How is it possible to build it so precise that the outward mirrors have exactly the same distance from the inside (half-silver) mirror?
    (5 votes)
    Default Khan Academy avatar avatar for user
  • female robot amelia style avatar for user Lionel Powers
    I still don't quite understand how the ether-affected ray travels longer/shorter time compared to the vertical one.. Not sure if we can use average velocity here, but if we do, then V(av)=((speed of light+speed of ether)+(speed of light - speed of ether))/2 = speed of light. For simplicity, I'll use random numbers - ((1000+1)+(1000-1))/2=1000, which is the same as initial velocity. Although when I use WolframAlpha, it gives different time for the two cases. I feel like I'm missing some little thing.. I would be very thankful if someone could explain it to me =)
    (2 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user Ross Henderson
      Let's use an absurd example. You're in a zeppelin, travelling with an airspeed of 50 kph with a 49 kph wind at your back. So your groundspeed is 99 kph, and you go 100 kilometres in a shade over an hour. Then you turn around to go home. Your ground speed is now 50-49=1 kph, and it will take you 100 hrs to complete your trip. So you're comparing 200/50 = 4 hrs for the round trip with no wind, to 101+ hrs with the strong wind.
      (6 votes)
  • male robot johnny style avatar for user ameya_khandekar
    As we know light would travel at different speeds at vacuum than let's say in glass. So, is this entire experiment performed in vacuum, to ensure that the experiment is not affected by the refractive index of any material ?
    (3 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user Dan  Bogar
      I think the test was done in country-side room with mirrors in air. You are right. M-M exp assumed either could be detected in air. That could have been wrong, and I expect M-M knew that. if the light signal travels at a speed relative to the air in the room and not relative to the either that would hide the either. The effect of room air slows light by 1.0003 or 3 parts in 10000. Speed through either would have to be 0.0003C or 90 km/s to have similar effect. I expect M-M understood that air might dominate the either effect but moved ahead with what they could accomplish at the time.
      (1 vote)
  • blobby green style avatar for user patelanish2k
    When the light ray goes up, it speeds up due to the "ether"... and when it returns, it slows down. So, Maybe there is a luminous ether but the experiment was just a fail as the same light ray speeds up and slows down by the same rate...
    (2 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user Ross Henderson
      Do the math - let vy be your speed, and ve be the ether speed, both with respect to the ground. If distance s=vt, then t=s/v. So going out, tg = s/(vy+ve) and coming back,
      tc = s/(vy-ve). Therefore, tg+tc = s/(vy+ve) + s/(vy-ve). Finding the common denominator,
      tg+tc = (svy-sve + svy+sve)/((vy+ve)(vy-ve)) = 2svy/(vy^2-ve^2). With no ether, the total time would be tg+tb=2s/vy. The only way 2svy/(vy^2-ve^2)=2s/vy is if ve=0 - in other words there's no ether effect.
      (2 votes)
  • primosaur seed style avatar for user Burns McLean
    The moon is about 3500 km wide. Suppose we were to direct a powerful laser in the direction of the moon, and then move it across the surface of the moon in 0.001 seconds. From the earth you observe the reflected light rays as the "dot" from the laser moves across the Moon's surface. But from the point of view of an observer on the Moon, would the "dot" appear to move across the Moon's surface at roughly ten times the speed of light?
    (2 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user ubiyen.pettus
    How does light propagate?
    (2 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user loridipa
    At why does Sal say longitudinal wave, isn't that a transverse wave? Plus, light is a transverse wave in general
    (2 votes)
    Default Khan Academy avatar avatar for user

Video transcript

- [Voiceover] We've seen in several videos so far that most of the waves that humans have encountered in nature--waves in the water, sound waves, or just waves traveling along a rope-- they were disturbances traveling through a medium. And so when light displays or has wave-like properties a very natural assumption was, well, light must also be a disturbance traveling through some type of a medium even if that medium wasn't so easy to detect. But they conjectured that there is some type of medium that light is disturbance traveling through. They called it the luminiferous ether. And much of physics in the 19th century was all around proving that the luminiferous ether existed and also figuring out what our relative velocity was in regards to that luminiferous ether. And why do we feel confident, or why did they feel confident, that there was a relative velocity? Well, we talked about that in the last video. The earth is spinning and then it's spinning around, it's orbiting around, the sun at a nice clip and then the whole solar system is orbiting around the center of the galaxy at a nice clip. The galaxy itself might be moving, so if you have some absolute frame of reference that's defined by the ether, well we are going to be moving relative to it. And if we're moving relative to it well maybe you just measure the speed of light in different directions and see whether the speed of light is faster or slower in a certain direction and then that might help you identify-- well, one, validate that the ether exists-- but also think about what our velocity is relative to the ether, relative to that absolute frame of reference. But the problem in the 19th century is that we didn't have any precise way of actually measuring--or a precise enough way of measuring--the speed of light where we could detect the relative difference due to the light going for or against, or into or away from, the actual direction of the ether wind. And so the experiment that is usually cited with first kind of breaking things open, starting to really make a dent in this whole idea of a luminiferous ether, is the Michelson-Morley Experiment. Michelson-Morley Experiment. They recognized, okay, we can't measure the speed of light with enough precision to detect has it gotten slowed down by the ether wind or sped up by the ether wind, but what we could do, and this is what Michelson and Morley did do, and I'm gonna do an oversimplification of the experiment, is that, okay, you have a light source, you have a light source right over here. So, you have a light source. And so that's going to send light in this direction. It's going to send light just like that. And what you do is you have a half-silvered mirror that allows half the light to pass directly through it and half of it to be reflected. So let's put a half-silvered mirror right over here. So, there's a half-silvered mirror. And so half of this light will bounce off like this, and this is just a simplification of it. Let me do it a little neater than that. So half will bounce off like that. And then the other half will be able to go through it. Will be able to go through it. It's a half-silvered mirror. And then we make each of those light rays-- we've essentially taken our original light ray and split it into two-- well then we'll then bounce those off mirrors. Bounce those off mirrors that are equidistant. And there are some adjustments when you actually have to factor in everything, but just as a simple notion, these things are just now going to bounce back. So, this one is now going to bounce back. It's half-silvered, it can go through, or part of it can go through, that mirror. So that's that ray. And then this one is going to bounce back. This one's going to go bounce back. And part of it is going to bounce into this direction. And then you can detect what you see. You can detect what you see. So this right over here is a detector. And you might be saying okay, Sal, well what's the big deal? You've taken a light source, you've bounced, you've split the light rays, you've put them back together, you've bounced them around a little bit. But think about if there was a luminiferous ether these light waves that are going in orthogonal directions will be going at different velocities. Let's say if that luminiferous ether, if that luminiferous ether wind was doing something like, let me see, if the ether wind were in this direction, if the ether wind were in that direction, when the light wave is going that way it should be going faster, and when the light wave is coming back it should be going slower. And so what Michelson and Morley did is they said okay, let's assume... let's adjust our apparatus right over here, so when then these two lights rays bounce off and come back together, if there were no ether you would have some basic interference pattern. So, what do I mean by interference pattern? Well let's say that you have maybe this one bouncing from up here. Let me do that in a different color. So the one bouncing from up here, the one bouncing from up here, let's say that looks like this. I'll just draw it as a longitudinal wave, just like this. Best I had in hand-drawn longitudinal wave. And then the one coming from the other direction, the one that bounces here, and then comes back like this, is another longitudinal wave, like this. And when they overlap they are going to interfere with each other, either constructively interfere or destructively interfere. So you could have something like this. So let me copy and then let me paste it. So depending on how far or how fast each of these traveled, you're going to have different levels of interference. And you would have a difference, depending on the orientation, depending on what the actual ether wind is doing. But what Michelson and Morley observed is that no matter how they oriented this apparatus, and they did it at different times of the year, and they rotated it around, and they rotated it in the vertical direction and the horizontal direction, no matter what they did they always got the same interference pattern. The interference pattern did not change. And because the interference pattern did not change it implied that well, maybe this ether isn't really having an effect on slowing down or speeding up the light waves. So this is often called one of the most famous failed experiments in physics. So let me write this down. What's powerful about it is that it was a failed experiment. Let me get my pen tool out. It was a failed, failed experiment but it made people start to question well, maybe there isn't an ether, a luminiferous ether, maybe light just somehow travels through the vacuum. Maybe there is not this absolute frame of reference that's defined by the luminiferous ether. And I wanna be clear, it wasn't this experiment by itself. This experiment was one of many that started to put that doubt. But even after this experiment and they saw that there was no change in the interference pattern no matter how they oriented this thing, whether is was going in the direction of the hypothetical luminiferous ether or away from it, whatever, when they say no matter how they oriented it they got the same interference patterns, people tried to come up with other explanations that still might have been okay with a Luminiferous Ether. Maybe the length got contracted in the direction of the motion. Maybe other things got affected. But this is a super important experiment in physics because, once again, it started to show that hey, maybe there is no luminiferous ether, that light is just gonna go through that vacuum and, as we'll see, is going to be traveling at the same velocity no matter what frame of reference you look at it from. But we'll explore that more in future videos.