192 ANNUAL REPOJKT SMITHSONIAN INSTITUTION, 1913. 



stars indicates that of the bright, visible stars one out of four belongs 

 to this class, while of the stars of the sixth magnitude there is only 

 1 out of 20, and that few, if any, would be found fainter than the 

 seventh or eighth magnitude. 



It has been strongly intimated by such men as Kapteyn, Campbell, 

 and Boss that the order of spectra B, A, F, G, K, M indicates sub- 

 stantially the order of development of the stars in time, so that the 

 stars of class B may be regarded as the younger stars, and those of 

 classes A, F, G, K, and M, successively, older and older. Prof. Boss 

 goes so far as to say : " There can scarcely be a doubt that the same 

 stars that are now seen of the spectral type A were in past ages of 

 the spectral type B, and then at a mean velocity of approximately 

 only two-thirds or three-fourths that which they have at present. It 

 seems equally probable that A stars of the present will eventually 

 become stars of the second ^ type in the future, and along with that 

 physical development will acquire an increase of mean velocity about 

 50 per cent greater than that those stars now have. This fundamental 

 fact of acceleration in the means of the stellar motions must have a 

 vital bearing on questions of stellar development." 



It is known that the spectrum of the general surface of the sun, 

 which is like that of stars of class G, goes over into the spectrum of 

 a sun spot, which is like that of stars of class K, by a mere lowering 

 of temperature. It is also known of all bodies with which we are 

 familiar upon the earth, that when, as time passes, they lose energy 

 by radiation they cool. Accordingly it seems probable that stars of 

 class G will at length reach the condition of class K by the mere 

 cooling incidental to the continuation of their radiation to space 

 through long periods of time. The gradual progress in form of 

 spectrum from class B to class M, the gradual progress in velocity of 

 motion from class B to class M, the gradual progress in distribution 

 in space from class B to class M, and other lines of gradual progress 

 which could be named, all seem to show that the arrangement of the 

 stars according to this classification corresponds to the march of a 

 fundamental progress in nature. That this progress is in point of 

 time from stars of simpler s{>ectrum to those of the more complex, 

 and not the opposite, is indicated by the consideration with regard 

 to the sun-spot spectrum which I have just cited. 



In contemplation of these various facts Prof. Campbell has re- 

 marked as follows: 



The close relationship of the class B stars to the Milky Way, their low radial 

 and tangential velocities, the apparent absence of class B stars in both near 

 space and distant space, a clustering of many of these stars in apparently 

 related groups — for example, in the Orion region — lead us to believe that the 

 present class B stars assumed stellar form in regions relatively near their 



^ Of Secchrs classification, in which B to F types are I, G to K are II, and M is III. 



