THE ILLUMINATION OF THE ECLIPSED MOON. 



3G3 



corona. Still, enough may remain to contribute a sensible 

 quota towards the illumination of the moon, and in part to 

 explain the variation in the degree of illumination. It has 

 been suggested recently that this question might bo settled 

 by an appeal to the spectroscope. The eclipse of January 

 last was examined spectroscopically by several observers. 

 Some found linos in the spectrum agreeing with those found 

 in the coronal spectrum. In other cases the results were 

 negative or ambiguous. But I doubt if this question, from 

 the nature of the case, can ever bo settled by the spectro- 

 scope. Wo know from theory that light must reach the 

 eclipsed moon both from the sun and from the corona, and 

 it is not easy to see what more the spectroscope can tell us. 

 I know of but one method by which it might bo possible 

 to ascertain whether the solar corona has any appreciable 

 share in the illumination of the eclipsed moon. It is the 

 opinion of some astronomers that the brightness of the 

 corona varies in cycles corrcaponding to those of the sun- 

 spots, only that when the sun-spots are at their minimum, 

 the corona is at its maximum. The sun-spot period is a little 

 over eleven years, and if it wore found in a long succession 

 of eclipses (the relative brightness of wliicli was carefully 

 estimated or measured) that eclipses occurring about the 

 period of sun-spot minimum wore above tlie average bright- 

 ness, there would be evidence in favour of the corona. But 

 this, if ever done, must be a work of the future. The records 

 of past eclipses are not accurate enough for the purpose, 

 except in the rare cases in which either the brightness or 

 the obscurity has been so abnormal as to attract notice. It 

 would bo rash to rely on those few instances. Yet it may 

 not bo out of place to romai-k that the two eclipses which 

 represent the two extremes of modern times in regard to the 

 degree of illumination both tend to favour the idea of the 



