112 LIGHT SCIENCE FOR LEISURE HOURS. 



be true) must exceed the volume of the earth more 

 than fifty thousand times. Besides such displays as 

 these, the most glorious auroras that have ever 

 illuminated terrestrial skies sink into utter nothing- 

 ness. 



But some difficulty was experienced in demonstrat- 

 ing that the spectrum on which these ideas had been 

 based belonged in reality to the ring-formed corona. 

 The study of the sun's surroundings by spectroscopic 

 analysis is not free from certain causes of perplexity. 

 To show how these may arise, we need only consider a 

 case which anyone possessing a small spectroscope (one 

 of Browning's miniature spectroscopes, for instance) 

 can readily test for himself. If such a spectroscope be 

 turned (with suitable precautions) towards the sun, we 

 see the principal solar dark lines, and we know that 

 those lines teach how the sun's light is partially 

 absorbed by the vapours of certain elements existing 

 in his atmosphere. But if next we direct the instru- 

 ment towards the sky, we see precisely the same 

 spectrum, only reduced in splendour. Yet the vapours 

 of iron, copper, zinc, and so on, do not exist in the sky. 

 The fact really is that we receive from the sky reflected 

 sunlight, and therefore we can trace in the spectrum 

 of skylight the dark lines belonging to sunlight. And 

 in exactly the same way, the sky during total eclipse, 

 though not very brilliantly illuminated, is nevertheless 

 lit up to some extent by the corona prominences, 

 and chromatosphere, and therefore the skylight must 

 supply, however faintly, those bright lines which belong 



