40 ECLIPSES. SECT. V. 



into the earth's shadow. The breadth of the space occupied by 

 the penumbra (N. 112) is equal to the apparent diameter of the 

 sun, as seen from the centre of the moon. The mean duration 

 of a revolution of the sun, with regard to the node of the lunar 

 orbit, is to the duration of a synodic revolution (JST. 113) of the 

 moon as 223 to 19. So that, after a period of 223 lunar months, 

 the sun and moon would return to the same relative position with 

 regard to the node of the moon's orbit, and therefore the eclipses 

 would recur in the same order were not the periods altered by 

 irregularities in the motions of the sun and moon. In lunar 

 eclipses, our atmosphere bends the sun's rays which pass through 

 it all round into the cone of the earth's shadow. And as the 

 horizontal refraction (N. 114) or bending of the rays surpasses 

 half the sum of the semidiameters of the sun and moon, divided 

 by their mutual distance, the centre of the lunar disc, supposed 

 to be in the axis of the shadow, would receive the rays from the 

 same point of the sun, round all sides of the earth ; so that it 

 would be more illuminated than in full moon, if the greater por- 

 tion of the light were not stopped or absorbed by the atmosphere. 

 Instances are recorded where this feeble light has been entirely 

 absorbed, so that the moon has altogether disappeared in her 

 eclipses. 



The sun is eclipsed when the moon intercepts his rays (N. 

 115). The moon, though incomparably smaller than the sun, is 

 so much nearer the earth, that her apparent diameter differs but 

 little from his, but both are liable to such variations that they 

 alternately surpass one another. Were the eye of a spectator 

 in the same straight line with the centres of the sun and 

 moon, he would see the sun eclipsed. If the apparent diameter 

 of the moon surpassed that of the sun, the eclipse would be 

 total. If it were less, the observer would see a ring of light 

 round the disc of the moon, and the eclipse would be annular, 

 as it was on the 17th of May, 1836, and on the 15th of March, 

 1858. If the centre of the moon should not be in the straight 

 line joining the centres of the sun and the eye of the ob- 

 server, the moon might only eclipse a part of the sun. The 

 variation, therefore, in the distances of the sun and moon from 

 the centre of the earth, and of the moon from her node at the 

 instant of conjunction, occasions great varieties in the solar 

 eclipses. Besides, the height of the moon above the horizon 



