382 THE CHEMISTRY OF THE SUN. [CHAP. 



flames at some distance from the arc when the arc lines have 

 become so thin as to render indications of absorption almost 

 impossible, or even after the lines reversed in the arc have 

 disappeared entirely from the flame. 



It is perfectly obvious that we have, then, in the electric arc, 

 something not far from an equivalent to a miniature sun ; we 

 have a mass of vapours hotter in the centre and cooler outside, 

 and this being so, we have in these reversal phenomena an 

 additional test to apply to the hypothesis. If the elementary 

 substances volatilised in the arc are broken up by the temperature 

 employed, complex molecules should be reformed at the exterior 

 and the simpler molecules should lie nearer the centre, as I have 

 shown in the case of the sun. This, so far as I can see, is a 

 sufficient explanation of all the reversal phenomena that I am 

 familiar with when the naked arc is used ; and it includes also 

 an explanation of the phenomena afforded by the flame. It 

 would not be difficult to show that, with one or two exceptions, 

 about which a great deal may be said, the arc picks out for 

 absorption those lines which are not seen in prominences, but 

 which are seen at the lowest temperature at which incandescence 

 takes place. 



If the particle subjected to these conditions gives out certain 

 radiations when it is vividly incandescent in the core, we know 

 of no good reason why, if the temperature be sufficient, it should 

 not absorb those radiations when away from the core. The 

 greater the number of exceptions there are to this genera] 

 statement, the less probable does it become that precisely the 

 same molecule exists in the two conditions. 



The simple and sufficient reason of absorption phenomena 

 would seem to be that certain molecules exist in the region of 

 of low temperature more complex than those which exist at a 

 high temperature, and that these molecules are really produced 

 by the association of those finer molecules which exist at the 

 centre of the arc. These finer molecules can exist only in a 



