176 



Mr. J. Norman Lockyer. 



The evidence for this separation of the stars into two series 

 •depended partly upon differences in the spectra of stars which, it 

 was thought, from other considerations, had about the same mean 

 -temperature. I showed that some of the principal stars might be 

 arranged as follows, in order of ascending temperature, the hottest of 

 hem being placed at the top of the list, 



Bellatrix, 

 Rigel, 

 oc Cygni, 

 7 Cys:ni, 

 a Tauri, 

 % Orionis, 



and that there were other cooling stars besides our sun, some of 

 which could also be arranged in the following order of descending 

 temperature.* 



Sirius, 

 fi Arietis, 

 Procyon, 

 Capella, 



Arcturus (Sun). 



Arcturus is bracketed with the sun, because I have shown that its 

 spectrum is like that of the sun, line for line.f 



The later discussion of the Kensington photographs of stellar spectra 

 fully justified the classification of stars suggested by the previous 

 inquiry into the visual spectra, and the general view that there are 

 bodies of increasing as well as bodies of decreasing temperature 

 was also greatly strengthened. 



Metallic Results, Descending. 



In what has gone before, I have dealt only with stars of increasing 

 temperature, and more particularly with those near the top of the 

 temperature curve. 



I now pass to the discussion of those stars which the previous 

 investigations indicated to be cooling bodies. 



The stars included in the inquiry are as follows, the coolest being 

 placed at the bottom of the list, Bellatrix, for the present being 

 taken as representative of the hottest stars : — 



* < Phil. Trans.,' vol. 184, A, p. 709. 

 f Ibid., vol. 184, A, p. 699. 



