Prof. Magnus on the Constitution of the Sun. 377 



the remark that the photosphere denned the outermost border 

 of the sun, but that the sun, as he inferred from observations 

 on the protuberances seen during a total eclipse, is surrounded 

 by a transparent atmosphere. Herschel said that the photo- 

 sphere was neither a liquid nor an elastic fluid, but that it con- 

 sisted of luminous clouds*. According to our present know- 

 ledge of the radiation of light and heat, it is difficult to suppose 

 that the photosphere, which sends out the heat of the sun, should 

 not have the nucleus which it encloses heated to ignition. The 

 reflecting atmosphere assumed to exist between it and the nucleus 

 may indeed hinder the illumination of the latter, but not its 

 becoming gradually heated. Kirchhofff is therefore right when 

 he says that this hypothesis for the explanation of the spots on 

 the sun is so completely in opposition to the higher physical 

 principles that we must abandon it, even if we are not in a posi- 

 tion to offer any other explanation which would render the spots 

 on the sun at all intelligible. His investigations of the solar spec- 

 trum have led Kirchhoff to the supposition that the sun is com- 

 posed of a solid or liquid nucleus which is at the highest tem- 

 perature of ignition, and is surrounded by a transparent atmo- 

 sphere of somewhat lower temperature. 



Up to the present time no one, so far as T know, has drawn 

 any conclusion relative to the constitution of the sun from the 

 heat which it gives out; the observations of SecchiJ upon the 

 smaller quantity of heat which is radiated from the poles than 

 from the equator of the sun may, however, be regarded as an 

 example in point. Some experiments which I have made upon 

 the radiation of heat give, I believe, a new insight into the con- 

 stitution of this heavenly body. If we observe the heat radiating 

 from a non-luminous gas-light and then introduce some soda 

 into the flame, whereby, as is well known, it will become very 

 luminous, we shall find that the radiation of heat will also in- 

 crease. The disposition of the experiment was so that a fixed 

 spot in the soda-flame was always compared with the same spot 

 in the non-luminous flame, and that, moreover, the soda intro- 

 duced into the flame could not radiate against the thermo-pile 

 which served for the observation. Although obviously a part of 

 the heat of the flame was taken up in heating to redness the 

 soda and the platinum wire which carried the soda, and in con- 

 verting the soda into vapour — although therefore the flame had 

 altogether a lower temperature than before when it was non- 

 luminous, yet nevertheless it radiated about a third more heat 

 than before. 



* Phil. Trans. 1/95, p. 71. 



t Denkschriften der Berliner Acad, der Wiss. 1861, p. 85. 



% Comptes Rendus, vol. xxxv. p. 606 ; and vol. xxxvi. p. 659. 



