Il6 LIGHT SCIENCE FOR LEISURE HOURS. 



an inner and brighter portion, which the sesquipedalians 

 have proposed to call the leucosphere, apparently on 

 the lucus a non lucendo principle, for it is neither 

 white nor spherical. And there is the outer portion, 

 much less brilliant, and much more strikingly radiated. 

 Neither one part nor the other presents a single feature 

 suggestive of an atmospheric nature; 1 and the cer- 

 tainty that the two portions belong to a single object 

 affords yet more conclusive evidence against this in- 

 terpretation of the corona. But the rays of the corona 

 are of a somewhat remarkable nature. When well seen, 

 as during the eclipse of 1868, they are pointed ; and 

 even during so unfavourable an eclipse as that of 

 December last, the dark spaces between the rays are 

 seen to widen rapidly with increased distance from the 

 sun. These pointed radiations serve to show that 

 coronal rays must be, in reality, shaped somewhat as 

 cones, having their bases towards the sun. The idea 

 is startling enough, but, admitting the accuracy of the 

 pictures made during well-seen eclipses, and of the 

 Astronomer-Eoyars account of the corona during 

 the eclipses of 1851 and 1860, there is no escape 

 from the conclusion here stated. It is not more certain 

 that the sun is a globe, and not a flat disc as he 

 seems to be, than that the coronal radiations are not 

 flat pointed rays, but cone-shaped. Yet no one will 

 suppose that there are a number of monstrous cone- 



1 I am here referring to the possibility that the corona may be due to 

 some species of solar atmosphere. The theory that the corona is due 

 to light in our own atmosphere has now at length been definitely aban- 

 doned by all astronomers. 



