26 LIGHT SCIENCE FOR LEISURE HOURS. 



ance of the sun's surface is inconsistent with the exist- 

 ence of masses of matter, fixed in position (or even 

 permanent in character, but unfixed in position) over 

 extensive solar regions. 



We have seen the probability that exists that the 

 red prominences are detached from the sun's surface. 



We know, thirdly, that they must exist at an in- 

 conceivably high temperature. 



Lastly, the spectroscope has proved that the sun's 

 light reaches us after passing through an extensive 

 solar atmosphere, consisting of the vapours of many of 

 our best-known metals. The vapour of iron, for 

 instance, forms a part of the sun's atmosphere much 

 in the same way as aqueous vapour appears as a 

 constituent of our own air. 



It seems to me reasonable to conclude from these 

 considerations that the objects called the red pro- 

 minences are, in reality, solar clouds ; only instead of 

 consisting, as our terrestrial clouds do, of visible 

 aqueous vapour (that is, of minute globules of water), 

 they consist of the visible vapours of the various metals 

 which exist in the solar atmosphere. In other words, 

 they are clouds formed by the condensation of the 

 metallic vapours into liquid globules. 



[This was written a few days before evidence was 

 obtained, showing that the prominences are gaseous 

 that is, even more mobile and tenuous than I had 

 inferred. The reasoning was sound, but proved more 

 than was insisted on.] 



Leverrier was led by his observation of the eclipse 



