F. W. Very— A Cosmic Cycle. 99 



stars (or possibly even the stars of the fifth division) are the 

 hottest, for there is no probability that the development of 

 heterogeneity has reached its limit. If we ask : Which stars 

 have the hottest photospheres 1 only conditional answers can be 

 given. We need to know the relation between the photospheric 

 level and the layer of mean density in the cooling gaseous 

 sphere. Which of these layers is most rapidly approaching 

 the center? Is there a gradual change in the incandescent 

 material which forms the photosphere, or are all photospheres 

 alike and formed of one highly refractory substance, perhaps 

 carbon ? If the last view is held, it might be surmised that 

 actual photospheric temperatures are closely accordant, and 

 that effective or apparent temperature is an affair of absorption 

 by layers above the photosphere. Since atmospheric absorption 

 in both solar and terrestrial envelopes is greatest for rays 

 of short wave-length, it follows on this assumption that the 

 bluest stars are those whose photospheres are at the highest 

 level, the yellow and red stars being those in which, by con- 

 densation and increased heterogeneity, the level of the photo- 

 sphere has sunk farthest beneath the outer boundary of the 

 stellar atmosphere. 



In his Presidential Address before the British Association 

 for the Advancement of Science in 1891, Dr. Huggins has 

 said : " Passing backward in the star's life, we should find a 

 gradual weakening of gravity at the surface, a reduction of the 

 temperature-gradient so far as it was determined by expansion, 

 and convection-currents of less violence producing less inter- 

 ference with the proportional quantities of gases due to their 

 vapor densities, while the effects of eruptions would he more 

 extensive."* 



With all of this I can agree, except the words which I have 

 italicized. Convection and explosions are not thus correlated. 

 While a star is expanded, the internal temperature is still 

 relatively low, and viscosity is small enough to permit quiescent 

 convectional circulation which does not interfere with a distri- 

 bution of gases approximately according to their densities. It 

 is after viscosity has become so great as to impede circulation, 

 and the transfer of deeper superheated matter to the surface 

 can no longer take place quietly, that explosions bring up the 

 denser gases violently and mingle them with the outer layers. 



The earlier stages of the helium stars, in my view, belong to 

 the quiescent convectional era, but like the quiescence of a 

 slumbering volcano, there is, during this period, a marshaling 

 of forces preparatory to a final catastrophe. After one or 

 more disruptive explosions, a new stage of somewhat quiescent 

 convection is entered upon in the Sirian stars. But as internal 



* Report B. A. A. S. for 1891, p. 16. 



