14 Introductory Chapter [OH. i 



a great difference in absolute brightness, although not necessarily in mass. 

 Any doubt that may have been felt as to the accuracy of this fundamental 

 fact has probably been removed by the investigations of W. S. Adams*, who 

 has found a purely spectroscopic method of determining the absolute brightness 

 of a star. Examining 58 red stars, Adams finds that 48 are of absolute 

 magnitudes between - TO and 3'4, while the remaining ten have absolute 

 magnitudes between 9*8 and 107, the division between Russell's giant and 

 dwarf stars thus being a clear gap of 6'4 magnitudes. More recently Adams 

 and Joyf have spectroscopically determined the absolute magnitudes of 500 

 stars of types F, G, K and M, and their results confirm Russell's facts com- 

 pletely. Of the 500 stars examined, 42 were of type M; of these 29 proved 

 to be of absolute magnitudes brighter than 2'9, one was of absolute magnitude 

 3'6, and the remaining 12 were all of absolute magnitudes fainter than 9'5. 

 Again there is a clear gap of about 6 magnitudes between " giants " and 

 " dwarfs." A similar, although less pronounced distinction, is found to persist 

 through types K and G, but it has almost, if not quite, disappeared for type F. 

 It is thus proved beyond doubt that there exist red stars of extraordinary 

 brightness, for which no place could be found in the older scheme of stellar 

 evolution; for it is, as Russell remarks, very improbable that these stars, 

 some of them 100 times as bright as the sun, are on the verge of extinction 

 through old age. 



Russell accordingly suggests that a star of B type is not at the beginning 

 of its career, but is half-way through. The star is supposed to have originated 

 as a giant star of M type, to have passed through the series of types M, K, 

 G, F, A to the stage B, and then to proceed again through the series A,F,G, K 

 until it becomes a dwarf star of type M. Only the most massive stars ever 

 attain to the degree of incandescence represented by a 5-type of spectrum ; 

 all others turn backwards before this stage is reached, a hypothesis which 

 gives at once a simple and perfectly acceptable explanation of the known fact 

 that 5-type stars are of exceptional mass, while at the same time accounting 

 for the gradual disappearance of the gap between giant and dwarf stars in 

 types K, G and F. 



It will be understood that this brief statement does not give an account of 

 all the details of Russell's theory, neither have we mentioned the many criti- 

 cisms which have been brought against it}. For our present purpose, it is 

 enough to notice that the indisputable facts on which Russell's theory is 

 based cut away to a large extent the original grounds for the belief that stars 

 originate out of nebulae. It is not proved that they do not, but there is no 

 longer any direct evidence that they do ; to retain the theory of nebulous 

 origin, we have now to imagine the nebulous matter to be or become non- 



* Proc. Nat. Acad. Washington, March, 1916. 



t Astrophys. Jouru. 40 (1917), p. 313. 



+ See for instance W. W. Campbell, Science, 45 (1917), p. 547. 



