cir. xxi. VELOCITY OF LIGHT. 173 



useful, for certain astronomical reasons, to know exactly when 

 these eclipses happened, and the time of their occurrence 

 was therefore calculated very carefully ever since Galileo 

 first discovered them. There was no difficulty in doing this, 

 and yet, strange to say, the eclipses rarely happened exactly 

 at the right moment. Sometimes they were too early, some- 

 times too late, and they varied according to some regular 

 rule as much as 16 minutes 36 seconds on each side of the 

 exact moment when they ought to have happened. 



At last it occurred to Roemer, and to an Italian astrono- 

 mer named Cassini, that, as Jupiter is farther away from the 

 earth at one time than at another, the eclipses might be seen 

 some minutes later whenever the rays of light from the 

 moons had to cross a greater distance to reach the earth. 

 Cassini seems to have put the thought aside and not to have 

 worked it out ; but Roemer seized upon it, and by careful 

 calculations proved that it was the true answer to the diffi- 

 culty. If the earth was at E (Fig. 32) for example, when Jupiter 



FIG. 32. 



J 



- 



Different Distances at which Jupiter's Light reaches the Earth, 

 j, Jupiter. E E', The earth. 



was at j, the light would not have nearly so far to travel as if 

 the earth was at E' ; and in this last position the 16 minutes 

 36 seconds would be taken up by the light crossing the 

 earth's orbit from E to E'. This distance was known to be 

 about 190,000,000 miles, so that light travels at the rate 



