1B2 STELLAR KVOLITTION. 



which has thou.stinds of counterparts juiioiig" other stars of the same 

 spectral class. 



We are now approaching the last chapters in the life history of a 

 star. After the solar stag-e has passed the color changes from yellow 

 to orange, and subsequentl}' to red, as the temperature falls, 'riie 

 spectral lines of hydrogen become fainter and fainter and finally dis- 

 ajopear completely. The lines of the metallic elements, on the con- 

 trary, become more and more complex and the changes in their 

 relative intensities are those which arc characteristic of lower temper- 

 atures. But curiously enough, there are two well-detined classes of 

 these older stars, which until recently were not known to have any 

 points in common except their red color. These are the stars of 

 Secchi's third and fourth types. In general appearance their spectra 

 are whollj^ unlike, particularly on account of the absertce from third- 

 class spectra of the broad dark bands due to the a]:)Sorption of carbon 

 vapor, the most characteristic feature of the fourth t3'pG. But in 

 spite of this apparent dissimilarit}^, photographs recentl}- taken with 

 the •iO-inch Yerkes telescope show that in certain regions of the spec- 

 trum stars of the two types are practicall}^ identical and are thus 

 probably more closely related than formerly appeared to be the case. 

 The measurements and reductions of a long series of photographs of 

 fourth-type spectra now in progress at the Yerkes Observatory should 

 soon permit us to form an opinion of the nature of these interesting 

 stars. 



In both the third and fourth types it is easy to trace the successive 

 stages of development. In stars of the fourth type the signs of increas- 

 ing age are particularly striking. The carbon vapor which produces 

 the broad dark bands becom(\s denser and denser, until it is not ditii- 

 cult to imagine that through the further increase of such absorption 

 the light of the star might be almost completely extinguished (PI. X). 



The phenomena of the red stars indicate that this final stage is close 

 at hand, and curiousl}" enough, in further testimon}^ of the remarkable 

 power of the spectroscope, the total extinction of a star's light does 

 not always prove sufficient to place that star beyond the reach of this 

 instrument. It is true that the spectroscope can not reveal the cheniical 

 composition of a solid body which is devoid of intrinsic light, but such 

 a body may form a system with another object which is still luminous, 

 and its gravitational power may cause the luminous I)ody to move in 

 an orbit. As we have already seen, the spiM-troscope is ca[)able of 

 revealing the motions of such a body. From a knowledge of thes(» 

 motions and Ihe time in \vhich the revolution is effected it is possible 

 to det(M"mine the mass and dinKMisions of the system, and in some spe- 

 cial cases, like that of Algol, the diameter and density of the invisible 

 component of" the pair. 



