284 THE WORLD MACHINE ' 



is no doubt that the light of the heavenly bodies not only far 

 surpasses the vivid appearance of white, but even the light of 

 any flame (with which we are acquainted) in the vigour of its 

 radiation. 



" The immense velocity of the bodies themselves, which is 

 perceived in their diurnal motion, and has so astonished thinking 

 men, that they have been more ready to believe in the motion 

 of the earth, renders the motion of radiation from them (mar- 

 vellous as it is in its rapidity) more worthy of belief. That 

 which has weighed most with us, however, is, that if there were 

 any considerable interval of time between the reality and the 

 appearance, the images would often be interrupted and confused 

 by clouds formed in the mean time, and similar disturbances of 

 the medium. Let this suffice with regard to the simple measures 

 of time." 



It should be needless to remark that again the italics in- 

 dicated are not Bacon's. While the prophet of the Great 

 Instauration was balancing speculations, the more practical- 

 minded Galileo was attempting a crude measure of the actual 

 speed of transmission. Others perfected his methods ; the 

 velocity still seemed infinite, the passage of light from star 

 to earth instantaneous. 



So it appeared to Descartes ; so it remained until a curious 

 observation by a young Danish observer, Olaus Roemer. While 

 Cassini is making a correct measure of the distance of the sun, 

 revealing the enormous breadth of the earth's orbit, this young 

 Dane is taking note of an extraordinarily minute but certainly 

 observable difference in the times of revolution of one of Jupiter's 

 moons. It seems a far-away cry from the subject in hand ; this 

 is what he observes : Carefully comparing observations of forty 

 immersions and emersions of the satellites, he finds that there 

 is a variation in the moments of their reappearance of some 

 seconds, depending on the time of the year. It is slight, but 

 of the fact there can be no doubt ; how can it be explained ? 

 The times of revolution ought to be absolutely equal. There 

 seems no reason to suppose that the orbital time of the satellites 

 varies ; they should be eclipsed and reappear with the punctu- 

 ality of a clock. Can it be explained by the varying distance 

 of the earth ? 



On the new estimates made by Cassini, when our globe is 



