CELESTIAL CHEMISTRY. 239 



able that, at the time of its emission, the light of all the stars is 

 white alike. The colors observed amongst them must then be 

 caused by some modification suffered by the light, after its 

 emission. 



"Again, it was obvious that if the dark lines of absolution 

 were more numerous, or stronger, in some parts of the spectrum, 

 then those colors would be subdued in power, relatively to the 

 color in which few lines only occur. These latter colors, remain- 

 ing strong, would predominate, and give to the light, originally 

 white, their oavu tints. 



" These suppositions have been confii'med by obseTvations. 



"We have shown that the colors of the stars are produced by 

 the vapors existing in their atmosphere. The chemical constitu- 

 tion of a star's atmosphere will depend upon the elements exist- 

 ing in the star, and ujjon its temperatui'e. 



" Variable Stars. — The brightness of many of the stars is found 

 to be variable. From night to night, from month to month, or 

 from season to season, their light may be observed to be contin- 

 ually changing, at one time increasing, at another time diminish- 

 ing. The careful study of these variable stars, by numerous ob- 

 servers, has shown that their continual changes do not take place 

 in an uncertain or irregular manner. The greater part of these 

 remarkable objects wax and wane in accordance with a fixed law 

 of jieriodic variation, which is peculiar to each. 



" We have been seeking, for some time, to throw light upon this 

 strange phenomenon, by means of observation of their spectra. 

 If, in any case, the periodic variation of brightness is associated 

 with physical changes occurring in the star, we might obtain 

 some information by means of the prism. Again, if the diminu- 

 tion in brightness of a star should be caused by the interposition 

 of a dark body, then, in that case, if the dark body be surrounded 

 with an atmosphere, its presence might possibly be revealed to 

 us by the appearance of additional lines of absorption in the 

 spectrum of the star when at its minimum. One such change in 

 the spectrum of a variable star we believe we have already 

 observed. 



" Betelgeux is a star of a moderate degree of variability. 

 When this star was at its maximum brilliancy in February last, 

 we missed a group of lines, the exact position of which we had 

 determined with great accuracy by micrometric measurements 

 some two years before. 



" Temporary Stars. — With the variable stars, modern opinion 

 would associate the remarkable phenomena of the so-called new 

 stars which occasionallj", but at long intervals, have suddenly ap- " 

 peared in the sky. But in no case has a permanently bright star 

 been added to the heavens. The splendor of all these objects 

 was temporary only, though whether they died out or still exist 

 as extremely faint stars is uncertain. In the case of the two 

 modern teuiiJorary stars, that seen by Mr. Hind in 1845, and the 

 bright star recently observed in Corona, though they have lost 

 their ephemeral glory, they still continue as stars of the tenth and 

 eleventh mao-nitudes. 



