REFRACTION OF LIGHT 353 



after carefully watching the brightest of Jupiter's satellites 

 or moons as it revolved around the planet, noticed that the 

 time of occurrence of its eclipses or passages behind the 

 planet showed a peculiar variation. He accurately deter- 

 mined the interval between two eclipses or the time it took 

 for a complete revolution of the satellite around the planet. 



Using this interval he computed the time at which other 

 eclipses should take place and found that as the earth 

 in its revolution around the sun moved away from Jupiter 

 the eclipses appeared to take place more and more behind 

 time. Determining the exact time at which an eclipse 

 took place when the earth was 

 nearest to Jupiter, and comput- 

 ing the time an eclipse should 

 take place six months later when 

 the earth was farthest from Jupiter, 

 he found that the actual time of 

 the eclipse was 22 minutes behind FIGURE 101 



the computed time. This slow- 

 ness he said must be due to the time required by the light 

 in crossing the earth's orbit. (Figure 101.) 



Many determinations of this kind have been made since 

 those of Roemer, and it has been found that he was some- 

 what in error, as the time required by light in traveling 

 across the earth's orbit is about 16 minutes and 40 seconds, 

 or 1000 seconds. Since the diameter of the earth's orbit 

 is about 186,000,000 miles the speed of light must be about 

 186,000 miles per second. Determinations of the speed 

 of light have been made in several other ways with almost 

 like results. 



Refraction of Light. Experiment 103. Place a penny in the 

 center of a five-pint tin pan resting on a table. Stand just far 



