In 1887 Albert Michelson and Edward Morley set out to conduct an experiment that would give a measurement of how fast the Earth was moving through the ether. The concept of ether arose from the belief that light, being a wave, needed a medium to propogate through. This was infered from the fact that waves in water have water as their medium, sound waves have air as their medium, a pulse on a string has the string as its medium etc. So it was rational to believe that light also needed a medium to propogate through. This medium was called ether and it was assumed that ether filled every part of the universe.

It should be noted that at the time of this experiment light having an absolute speed in a vacuum was not an accepted notion. This notion only came about from Einsteins Special Theory of Relativity.

The basic principal behind the experiment was:

• The ether is stationary in space.
• The Earth is moving through the ether.
• If light is made to propogate in perpendicular direction, due to the movement of the Earth it will be moving through the ether in different directions with different velocities. One would have the speed of Earth plus the speed of light in the ether, the other would just have the speed of light in the ether.
• This difference would result in a value for the speed of the Earth through the ether.

The experiment was a spectacular failure, repeatedly. But it did result in a brand new theory.

How the interferometer works.

The light enters at the source point on the entrance arm, arm1. It then passes through the beam expander and is reflected off of a mirror, from which the beam splitter cube (BS) is illuminated. At the BS the light splits into two, one half going towards mirror M1, beam1, the other propagating straight through to mirror M2, beam2. The beams reflect and recombine at BS again and they then continue onto the screen.

This seems irrelevant, splitting a beam and recombining it. But in the initial experiment it was thought the ether wind would affect one of the beams, causing interference.

In modern interferometers M1 is attached to a micrometer. Therefore the distance that beam1 travels can be changed, if it is changed by just the right amount interference will be seen on the screen.

M2 is attached to tilt screws so that the centre of interference can be moved so that it is located on the centre of the screen. As the mirrors are circular, circular fringes will be observed.

In most interferometers a compensating plate is needed so that both beams travel the same optical distance. As light propagates at different speeds through different media, if one beam of light has to pass through a medium other than air it will take longer to cross the same distance as a beam that is propagating through air.

In the original interferometer one beam was travelling an extra distance through a glass slide, therefore a compensating slide had to be placed in the path of the other beam so that both were travelling equal optical distances.

The beam splitting cube negates the need for the compensating plate as both beams are already passing through the same optical distance.

Method

This method is how the experiment is conducted in modern times, needless to say Michelson and Morley did not posess He-Ne lasers, which makes there efforts all the more respectable.

1. The first part of the experiment was calibration. The mirrors had to be adjusted so that the interference pattern could be observed on the screen. A pointer was placed directly in the path of a He-Ne laser. Multiple images of the pointer were seen on the screen. The tilt controls for M2 were adjusted until the multiple images coincided to form one image, the pointer was then removed as was the screen at the end of the interferometer arm and the circular fringes were centred on a distant white board screen, when M1 was moved the fringes were seen to collapse into the centre or emerge out from the centre of the pattern depending on which way the micrometer gauge was turned. Both controls were adjusted until a bright sharp pattern was observed on the screen, when this was observed the equipment was set-up correctly and the next part of the experiment could be carried out.
   
2. 
   The second part of the experiment was to find the calibration constant of the micrometer screw gauge using the He-Ne laser. As was already noted when M1 was moved the fringes were seen to collapse into or emerge out of the centre of the pattern, it was decided that measurements would be taken by collapsing the fringes. The initial or zero position of the micrometer was noted; the fringes were then collapsed in 30 at a time noting every thirty what distance the micrometer had moved to. We know that for every fringe collapsed the mirror M1 moved a distance l/2, where l is the wavelength of the light being used, in this case was 632.8 nm, therefore for N fringes collapsed the mirror moved a distance D = Nl/2, if for this distance the micrometer was seen to move a distance x, then it can be said that,
   

   Nl/2 = Kx.

   
   
   
   Where K was the calibration constant of the micrometer. Using the results taken a graph of N vs. x was drawn; the slope of this graph was, from the above equation, m = 2K/l. So it was seen that K = ml/2.
   


3. The third part of the experiment was the original experiment. The two mirrors were set at the exact same distance from BS. The He-Ne laser was turned on and no interference pattern was seen.


4. The last two parts of the experiment were parts that I conducted as part of my undergraduate physics course.


5. The fourth part of the experiment was to obtain a measurement of the wavelength of the mercury green line. To do this first there had to be zero path difference between the mirrors, this was very hard to achieve and was done using the He-Ne laser, the tilt screws were adjusted so that the centre of the interference pattern was seen exactly in the middle of the screen, the mirror M1 was then moved so that the fringes were seen to collapse into the centre of the interference pattern, the micrometer was turned until at one point the fringes were seen to start emerging from the centre even though the micrometer was still being turned in the same direction. This exact point was noted. This was the point of zero path difference. The Mercury discharge lamp with a green filter placed over the output of the lamp was then put in place of the He-Ne laser and the screen put back into the interferometer. The tilt controls of M2 were then adjusted so that the sharpest image possible was observed, the reading on the micrometer was noted and instead of circular fringes parallel lines were observed. A point was picked on the screen and the micrometer was moved in such a way as to pass N fringes past the point picked so that the same equation that was used in the previous part could be used here but this time with l being the unknown instead of K, the same graph was used but at the end the equation l = 2K/m, where m was the slope of the graph in question.


6. For the fifth part of this experiment the aim was to observe white light fringes. The point of zero path difference was again found after adjusting the tilt screws of M2 and moving M1, this time using the mercury lamp. The mercury lamp was then replaced with a white light lamp. The micrometer was then adjusted slightly and the interference pattern was observed on the screen. It was seen to be a black centre line with white light on either side with the full white light spectrum then spreading out on either side.

Conclusions

It can be said with some certainty that the experiment was a complete failure, with regards to their inital goals.

But it did help prove a number of other theories later on.

• It showed that if there is a substance such as ether filling the universe, it does not effect electromagnetic rays noticeably.
• It helped to prove the theory that electromagnetic rays are self-sustaining.
• It helped to prove that light has an absolute speed in a vacuum.

As seen in part 4 it can be used to calculate wavelengths of lines of monochromatic light.

N.B. The differences between the speeds of light in a vacuum and in air are negligible in this case.