The Michelson Morley experiment was designed to detect any movement of the earth through the aether. The aether was a theoretical mass-less incompressible and non-viscous fluid which was said to pervade all space. It was all tied up with the idea that light travelled in wave and waves needed something to move through. One big question mark in the theory was the motion of the earth. Did the earth drag the aether along with it or did it slip through the aether or was there some combination of the two. Did the aether drift past the earth? If it did, then the speed of light would vary with direction.
Interferometry
A little known science called Interferometry was developing just at the right time. Light rays are split by passing them through a part silvered mirror. (We now make some types of sun glasses this way) If the rays are latter recombined after travelling slightly different distances, then the the light waves interfere with each other and fringes of colour, or bright and dark in the case of monochromatic light, can be seen. The most common example of an interference phenomenon is in the reflection of light from a puddle contaminated with oil. The light is reflected from both the top and bottom of the oil film and causes colours to be seen depending on the viewing angle and the thickness of the oil film. Interferometry allows very small distances to be measured with an accuracy down to the size of an atom! Michelson had invented one such interferometer and was putting variants of it to many uses. It was now time to use such great power to detect the earths motion through the aether.
The Experiment
In collaboration with Morley, Michelson built on interferometer on a block of stone 1.5 metres square and 25cm thick which floated on mercury. The light was split into two perpendicular paths and after travelling about 11 meters (backwards and forwards: this is not shown) was recombined. If light travelled through the aether at a constant speed and the earth travelled through the aether then it would result in different values being obtained for the speed of light depending on the direction of the light. The experiment compares the speed of light in two perpendicular directions. As the earth rotates about its axis and moves in its orbit around the sun then there would be times when one arm would point in the direction of the earths velocity through the aether while the other arm would be perpendicular to it. At the other times the arms would swap roles. By setting the whole apparatus in motion revolving slowly there should be times of the day and year when this effect would be detected.
The earth's speed is about one ten thousandth that of light (10-4), but the geometry of the experiment determines that the effect is proportional to the square of this. Michelson and Morley's apparatus would need to detect one part in one hundred million (10-8). This would result in the fringes moving 2/5 of a fringe which would be easily visible.
The experiment failed to be detect any such changes. The light always travelled at exactly the same speed in each arm of the apparatus. The experiment has been repeated several times and its accuracy increase. In the Kennedy Thorndyke experiment the accuracy was increased to on part in 100 billion (10-11), still with the same null result.
It would seem that the movement of an observer through space had no effect on the value he obtained for the speed of light. The speed of light would appear to be a universal constant. Henceforth the aether was out of fashion.