In 1667,  Galileo suggested a method for actually measuring the speed of light.   The method was to take two people A and B, covered lanterns to the tops of hills that are separeted by a distance of about a mile.  First A uncovers her latern.  As soon as B sees A's light, she uncovers her own lantern.  By measuring the time from when A uncovers her lantern until A sees B's light, then dividing this time by time by twice the distance between the hill tops, the speed of light can be determined.  Galileo was able to determine only that the speed of light was far greater than could be measured using his procedure.  Althougt Galileo was unable to provide even an approximate value for the speed of light, his experiment set the stage for later attempts.


The first successfull measurement of the velocity of light was provided by the Danish astronomer Olaf Romer in 1675.  He based in measurement on observations of the eclipses of one of the moons of Jupiter.  As this moon orbits Jupiter, there is a period of time when Jupiter lies between it and the earth, and blocks it from view.  Romer noticed that the duration of these eclipses was shorter when the Earth was moving toward Jupiter than when the Earth was moving away. He correctlly interpreted this phenomena as resulting form the finite speed of light. 

Geometrically the moon is always behind Jupiter for the same period of time drign each eclipse.  Suppose, however, that the Earth is moving away from Jupiter.  An astronomer on Earth catches his last glimpse of the moon, not at the instant the moon moves behind Jupiter, but only after the last bit of unblocked light form the moon reaches his eyes.  There is a similar delay as the moon moves out form behind Jupiter  but, since the Earth has moved father away, the light must now travel a longer distance to reach the astronomer.  The astronomer, therefore sees an eclipse that ,lasts longer the actual geometrical eclipse.  Similarly, when the Earth is moving toward Jupiter, the astronomer sees and eclipse that lasts a shorter interval of time.

From observations of these eclipses over many years, Romer calculated the speed of light to be 2.1*10^8 m/sec.  This value is approximately 1/3 too slow due to inaccurate knowledge at at that time of the distances involved.   Nevertheless, Romer's method provided clear evidence that the velocity of light was not infinite, and gave a reasonable estimate of its true value-not bad for 1675.


The French scientist Fizeau, in 1849, developed an ingenious mehod for measuring the speed of light over terrestrial distances.  He used a rapidly revolving cogwheel in front of a light source to deliver the light to a distant mirror in discrete pulses.  The mirror reflected these pulses back toward the cogwheel.   Depending on the position of the cogwheel a pulse returned, it would either block the pulse of light or pass it through to an observer.  He measured the rates of cogwheel rotation that allowed observation of the returning pulses for carefully measured distance between the cogwheel and the mirror.  Using the method, Fizeau measurd the speed of light to be 3.15*10^8 m/sec.  This is within a few percent of the currently accepted value.


Foucalt improved Fizeau's method, using a rotating mirror instead of a rotating cogwheel.  Micelson used Foulcalt's method to produce some accurate measurements of the velocity of light.  The best of these measurements gave a velocity of 2.99774*10^8 m/sec.  This maybe compared to the presently accepted value of 2.99792458*10^8 m/sec.