and to notice the interval that elapsed between each of them. But it at last happened that the interval between the two occultations, which was about forty-five hours, became prolonged by periods of 8, 13, and 16 minutes, during that half of the year when the earth was receding from the planet, while it became proportionally cut short during the rest of the year. Römer was struck by a happy idea; he suspected instantly that the moment when he remarked the disappearance of the satellite was not always coincident with the instant when it really took place, but that it sometimes appeared to happen later—that is to say, after an interval of time sufficiently long to allow the light that had left the satellite immediately after its disappearance, to reach the eye of the observer. Hence it became evident that the farther off the earth was from the satellite, the longer was the interval of time between its disappearance and that of the arrival of the last portions of its light upon the earth; but that the moment of the disappearance of the satellite is that of the commencement of the occultation, and that the moment of the arrival of the last portions of light is that when the commencement of the occultation is observed.
It was thus that Römer explained the difference between the calculated and observed time of the occultation, and he saw that he was on the threshold of a great discovery. In a word, he saw that light propagated itself through space with a certain velocity, and that the fact we have just mentioned furnished the precise means of measuring it.
Thus the occultation of the satellite was retarded one second for every 185,000 miles that the earth is distant from Jupiter; the reason being, that a ray of light takes a second to travel this distance, or, in other words, because the velocity of light is at the rate of 185,000 miles per second.
