192 A Short History of Astronomy [CH. vii. 



improvement in the customary estimate of the distance of 

 the earth from the sun, from which those of the other 

 planets could at once be deduced. 



If, as had been generally believed since the time of 

 Hipparchus and Ptolemy, the distance of the sun were 

 1,200 times the radius of the earth, then the parallax 

 (chapter n., 43, 49) of the sun would at times be as 

 much as 3', and that of Mars, which in some positions is 

 much nearer to the earth, proportionally larger. But Kepler 

 had been unable to detect any parallax of Mars, and there- 

 fore inferred that the distances of Mars and of the sun 

 must be greater than had been supposed. Having no 

 exact data to go on, he produced out of his imagination 

 and his ideas of the harmony of the solar system a distance 

 about three times as great as the traditional one. He 

 argued that, as the earth was the abode of measuring 

 creatures, it was reasonable to expect that the measurements 

 of the solar system would bear some simple relation to the 

 dimensions of the earth. Accordingly he assumed that 

 the volume of the sun was as many times greater than the 

 volume of the earth as the distance of the sun was greater 

 than the radius of the earth, and from this quaint assumption 

 deduced the value of the distance already stated, which, 

 though an improvement on the old value, was still only 

 about one-seventh of the true distance. 



The Epitome contains also a good account of eclipses 

 both of the sun and moon, with the causes, means of 

 predicting them, etc. The faint light (usually reddish) with 

 which the face of the eclipsed moon often shines is correctly 

 explained as being sunlight which has passed through 

 the atmosphere of the earth, and has there been bent from 

 a* straight course so as to reach the moon, which the light 

 of the sun in general is, owing to the interposition of the 

 earth, unable to reach. Kepler mentions also a ring of 

 light seen round the eclipsed sun in 1567, when the 

 eclipse was probably total, not annular (chapter n., 43), 

 and ascribes it to some sort of luminous atmosphere round 

 the sun, referring to a description in Plutarch of the same 

 appearance. This seems to have been an early observation, 

 and a rational though of course very imperfect explanation, 

 of that remarkable solar envelope known as the corona 



