DISCOVERY 



117 



Clark, in the clear air of America ; though its light is 

 exceedingly dim owing to its low temperature. Its 

 existence was detected, however, in the first instance, 

 not by its luminosity, but by its gravitational influence 

 on its bright companion. 



And many another double star has thus been detected, 

 some of them with components of equal brightness ; 

 and as they revolve one is receding and the other 

 approaching, which has the effect of causing every 

 single line in their spectrum to appear double — half 

 its light being shifted to the right and half to the left — 

 although the stars themselves may be so close as never 

 to be seen double by even the highest power telescope. 

 Such is the remarkable phenomenon connected with 

 the star /3 Aurigse, which was discovered by one of 

 the assistants workmg under the Draper Memorial 

 for Professor Pickering, at Harvard, when examining 

 photographs of the spectrum of a number of stars 

 taken on successive days. The rate of revolution in 

 this case is so enormous that a complete period was 

 gone through in four days. 



To return, however, to the eclipse of Jupiter's 

 satellites, it is really a phenomenon of essentially the 

 same kind as is noticed in sound or light, though it is 

 observed by altogether different means. For here the 

 period of revolution or vibration, instead of being a 

 minute fraction of a second, is something comparable to 

 a week or a month. And the completion of a period is 

 observed by the periodical plunging of the satellite 

 into the shadow of the planet, that is to say, by an 

 eclipse. Making all proper allowances for shifting of 

 our owTi position — which to an astronomer is simple 

 enough — the period of revolution can be pretty accu- 

 rately determined. The true value is obtained when the 

 earth is nearest Jupiter and moving, as it were, side- 

 ways, not to or from that planet : and the true period 

 is also observable when the earth is farthest from Jupi- 

 ter, and again moving sideways. But at the halfway 

 points, three months distant from either of these, 

 when the earth is moving towards Jupiter, the period 

 of the satellite appears accelerated, seeming to revolve 

 in a shorter time than it really does, because the earth 

 is constantly catching up the light which, by its appear- 

 ance or disappearance, brings the message. On the 

 other hand, when the earth is running away from 

 Jupiter, the light from a reappearing satellite has to 

 overtake an observer ; and accordingly we get the 

 information late : so that, if we were not acquainted 

 with the cause, we should assume that the satellite 

 was revolving more slowly than its average value. 



It ought to be said that when Roemer published his 

 great discovery, as above summarised, it was regarded 

 as fanciful and not accepted, because up to that time 

 it had been thought that light travelled instantane- 

 ously, or perhaps did not travel at all. It simply 



existed. The fact that light really did travel at a 

 definite speed was of fundamental importance. It was 

 the first definite thing ascertained about the Ether of 

 space, and was not to be lightly accepted. Like many 

 another discovery, it had to run the gauntlet of con- 

 troversy. And it was not tiU fifty years later, when 

 Bradley, the Astronomer of Oxford, conceived the 

 explanation of a quite different and still more 

 delicate phenomenon which he had observed, viz. 

 the slight shifting in the position of stars, so that 

 they described some kind of ellipses or circles in the 

 course of the earth's year — an observation which he 

 ultimately explained as due to the finite velocity of 

 light ; it was not, I say, until Bradley arrived at a 



JUPITEiv, lUili-iL.^kV 17, 1906. 

 From a drawing by the Rev- James Baikie, Edinburgh' 



similar result, viz. that the speed of light was ten 

 thousand times that of the earth in its orbit, that 

 Roemer's theory was accepted and incorporated into 

 science. For it was perceived that the two observa- 

 tions mutually sustained each other, although up to 

 that time no terrestrial observation had been sufficiently 

 delicate to enable the speed of light to be directly 

 measured by experimenting with artificial sources on 

 the earth's surface. This was done, as everyone now 

 knows, by the great French physicist Fizeau, in 1849, 

 and afterwards confirmed beyond the shadow of a doubt 

 by Foucault ; so that the speed with which the Ether 

 is able to transmit waves became the familiar and 

 commonplace fact which it is to-day. 



A Popular Misconception 



I began this article by saying that there was much 

 misunderstanding about the behaviour of Jupiter's 



