64 ILLINOIS STATE ACADEMY OF SCIENCE 



mass this maximum density would occur — perhaps 50 

 or 100 times the mass of Jupiter. There is also some 

 minimum gaseous mass which would become incandes- 

 cent due to its generation of heat. It seems natural to 

 identify this with the mass of maximum density although 

 in reality the two may be different. This point of maxi- 

 mum density would then separate large gaseous masses 

 into two classes, the dark gaseous and the bright gaseous 

 or stellar. The sun is 1000 times the mass of Jupiter and 

 it is in a state of intense incandescence, indicating that 

 it is far beyond the point of maximum density. Its dens- 

 ity is 1.4 or slightly greater than that of Jupiter. 



When incandescence first appears in a growing star 

 the mass of the star would be small and the luminescence 

 would be feeble and of a reddish color. It would be a 

 dwarf red star. As the star travels and gathers in the 

 nebuluous material of space its mass increases, its color 

 brightens, and the character of its spectrum changes 

 through the series of types M, G, F, A, and finally to B 

 if the mass increases sufficiently. Passing beyond the 

 massive B-type stars, we arrive at the class of ''giant" 

 stars with small densities and high luminosities. These 

 stars while relatively few are very conspicuous, and are 

 found in all spectral classes. The return from the B- 

 type of spectrum to that of type M with further increase 

 of mass is probably due to the existence of a highly ten- 

 uous envelope surrounding the star, such as the sun's 

 corona, due to radiation pressure. As this envelope in- 

 creases in thickness with excessive and increasing radia- 

 tion the direct light of the star itself, or perhaps better, 

 the nucleus of the star, is cut off and an increasing per- 

 centage of the light radiated by the star comes from this 

 relatively cooler envelope. The enormous diameters 

 recently found by Michelson, 300 times the diameter of 

 the sun for some of the giant red stars, seems to favor 

 this interpretation. 



It will be noticed that this arrangement of the spectral 

 classes from the giant M-type, to the B-type, to the 

 dwarf M-type stars is precisely the same arrangement 

 as that made by Russell. The interpretation of the ar- 



