Michelson-Morley Experiment (MMX) is one kind of optical interference of two beams of light. So the explanation of the the results from MMX should depend on the conception of light interferenc.

The two coherent light sub-sources (S and S') are made by the splitter mirror in MMX, and a time difference will be produced by moving media. Meanwhile the space difference of the two sub-source will be produced too, i.e. the position of S' will not be at the same point with S, as shown in follow figure.

Optical interferenc at the observation point O in MMX

 

Thereby, the Optical interference fringes at observation O is determined by the difference of optical path between S-B'-S'-O and S-A-S-O. Clearly, this defference of opitcal path includes the effect of the space different of the two coherent light sub-sources, and is different to the current solution of MMX which the space different of the two coherent light sub-sources is not considered.

I don't known the understanding above is correct, or not?

If it is right, and what will happen to the interpretation about MMX?

S and S' are the same point. Even under the historic view of the experiment as a test of the aether, the light from both arms overlaps from S' to O. If everything is moving, it has to include the beamsplitter and the detector. So there is no kink in the light path at S'

S and S' are the same point.

 

Only if the velocity of moving media u is zero. but if u is not zero, the direction of light beam SB will be changed to SB', and will come back to S' after reflected by mirror b. So, S' is not at the same position with S absolutely, isn't it??

Only if the velocity of moving media u is zero. but if u is not zero, the direction of light beam SB will be changed to SB', and will come back to S' after reflected by mirror b. So, S' is not at the same position with S absolutely, isn't it??

 

If you want to describe absolute motion they are not at the same point in the absolute frame, but the point is that you have to draw a second beamsplitter and detector to properly show this. The light does not change direction when it goes through the beamsplitter on its way from B to O, and when it strikes the beamsplitter, it will be where the light returning from A strikes it. i.e. the interferometer works.

 

I suppose there's the view that the light does not deflect at all, in which case simply getting the interferometer to show a fringe can falsify the idea of absolute motion.

If you want to describe absolute motion they are not at the same point in the absolute frame, but the point is that you have to draw a second beamsplitter and detector to properly show this. The light does not change direction when it goes through the beamsplitter on its way from B to O, and when it strikes the beamsplitter, it will be where the light returning from A strikes it. i.e. the interferometer works.

 

I suppose there's the view that the light does not deflect at all, in which case simply getting the interferometer to show a fringe can falsify the idea of absolute motion.

 

Sorry, in my post, the motion of media Indicates the relative motion of media to interferometer but not the absolute motion of it, because the effect of the moving media on propagation of light beam is relate to the relative velocity of media to interferometer.

In any event, it will move in a straight line.

In any event, it will move in a straight line.

 

It could be changed by the disturbance.

 

I think the conclusion of MMX given by Michelson need to be reconsidered, because his analysis was incomplete.

For a moving bar, the measure result of the length do not relates with its motion and speed of light, if we select two points on trajectory of the moving bar and measure the times that the moving bar reach and leave the points.