F. W. Very — A Cosmic Cycle. 51 



or helium stars. Div. 2. Sirian or hydrogen stars. Div. 3. 

 Procjon or hydrogen-iron stars. Div. 4. Solar or calcium 

 stars. Div. 5. Stars with unknown flu tings. Div. 6. Stars 

 with hydrocarbon flutings. 



The division of stellar spectra proposed by Sir William and 

 Lady Huggins* rests upon the character of the hydrogen spec- 

 trum alone, and is hardly complete enough for my purpose, 

 but its simplicity recommends it. Three conditions are noted: 



" (1.) All the lines thin, defined and very distinct up to the 

 end of the series. 



(2.) All the lines winged and strong, but more or less indis- 

 tinct, and incomplete in number, as the end of the series in 

 the ultra-violet is approached. 



(3.) In the photographic region the first three lines only, 

 namely H#, Hy, Hj, more or less thin and defined, the line 

 He diffuse, and the lines beyond practically absent, probably 

 through excessive diffuseness." 



In addition to these features, hydrogen types (1) and (2) 

 show a broad diffuse band of absorption from 0*35/* to 0*37^, 

 which is wanting in (3), and which does not seem to have been 

 noted before. Since the intensity of this diffuse absorption is 

 proportional to the breadth of the ultra-violet hydrogen lines, 

 and since its position is associated with the vanishing point of 

 their harmonic rhythmical succession, there can be little doubt 

 that the band is due to the same substance. 



Huggins' hydrogen type (2) will include McClean's divisions 

 (2) and (3). It is suggested by Sir William and Lady Huggins 

 that the fading out of the more refrangible members of the 

 hydrogen series in the solar stars is due to interference with 

 these atomic vibrations by the simultaneous vibrations of 

 associated metals which have powerful absorption in the same 

 region (loc. cit., p. 104). 



Heat of the Stars. 



The question of relative stellar temperatures is of funda- 

 mental importance. Sir William and Lady Huggins, from 

 their photographs of ultra-violet and violet stellar spectra, con- 

 clude that solar stars have hotter photospheres than Sirian. 

 The argument is as follows : Comparing the violet spectra of 

 such stars as Capella, which culminates at 5-J-° from the zenith 

 at Tulse Hill, and Vega, whose meridian zenith distance is 

 12j-°, and where, consequently, the spectra have been but little 

 changed by the absorption of the terrestrial atmosphere, it is 

 seen from photographs taken under as nearly as possible iden- 

 tical conditions that, although, as a whole, the violet region 



* An Atlas of Representative Stellar Spectra, p. 154-155. 



