The speed of light is measured using a time-of-flight method, over a distance of 20 cm, using a high speed pulsed laser, a beam splitter, mirrors and fast detectors and oscilloscope. We then compare the measured speed of light with the measured speed of radio waves. This page supports the multimedia tutorial The Nature of Light.
A high speed pulsed laser (off-screen to the left) produces a regular series of very short pulses. (The pulse rate is 80 MHz and each pulse has a lenght of 0.2 ps. The wavelength is 780 nm, with a 10 nm width due to the Fourier limit and the short pulse.) The red lines drawn over this image show the beam path. (They are not a photograph of the beam, of course: the air in the lab is clean so there is nothing to scatter the beam and so it is invisible from the side view.) The first mirror is just to direct the beam towards our experiment. The beam splitter is just a glass plate. It allows the transmitted beam to come straight to the first detector, at bottom left. The beam reflected by the beam splitter goes to the second and third mirrors before arriving at the second detector.
The screen of the oscilloscope shows the signal from the first detector on the top trace and that of the second detector on the lower.The laser outputs a sustained series of pulses, and the signal from the first detector is used to trigger the oscilloscope trace for both channels. The divisions on the horizontal axis are one nanosecond. The second pulse arrives several ns later than the first, because it has travelled a longer distance. The shape of the pulses on the screen is largely the response of the detectors, rather than showing the time variation of intensity in the light pulse.
Now it is time to move the second mirror.
Moving the second mirror to change the path length difference.
In this film clip, I move the second mirror 10 cm to the left (see the metre rule on the bench). Because the beam has an out and back path, this reduces the path length by 20 cm. The pulse on the second trace arrives earlier by 0.67 ns, with respect to the first, giving us the value for the speed of light: c = 30 centimetres per nanosecond or 3 X 108 m.s−1.
One of the purposes of making this measurement was to show that, in spite of the great speed of light, it is relatively easy to measure it with contemporary equipment.
The first reasonable measurement of the speed of light was made in 1676, by Ole Romer, using observations of the eclipses of the moons of Jupiter. Measurements Fizeau and Foucault in the nineteenth century used mechanical means to 'chop' a light beam, and longer distances of travel.
Another purpose was to allow us to compare the measured speed of light with the measured speed of radio waves.
Maxwell noted that the speed of the wave solution to what we now know as Maxwell's equations of electromagnetism was similar to that measured for light. This was persuasive evidence that light was an electromagnetic wave.
On a personal note, although this was an expensive experiment for a quick experiment with only low precision, I heartily enjoyed making it! Peter Reece and Fan Wang have a fine optoelectronics research lab just down the corridor from mine, and they were able to set this up for me very quickly (thanks Peter and Fan). Optoelectronics is an area of vigorous research, because of the increasing use of optical fibres to carry signals, usually produced by rapidly pulsed lasers.