On the Chemistry of the Hottest Stars. 



1(37 



lines.* "the enhanced lines may be absent from the spectrum of a 

 star, either on account of too low or too high a temperature. In the 

 case of a low temperature, however, iron is represented among the 

 lines in the spectrum, but at the highest temperature all visible 

 indications of its presence seem to have vanished. This result 

 affords a valuable confirmation of my view that the arc spectrum of 

 the metallic elements is produced by molecules of different com- 

 plexities, and it also indicates that the temperature of the hottest 

 stars is sufficient to produce simplifications beyond those which have 

 been produced in our laboratories." 



The facts which are graphically represented in Map I indicate 

 that so far as these stars are concerned the results, with regard to 

 stellar temperature determined by a study of the iron lines, are 

 identical with those to be gathered from the extension of the radia- 

 tion spectrum into the violet. Independently of the extensions of 

 spectrum into the violet for different stars, then, the relative tem- 

 peratures may be determined by a study of the iron lines. Thus, a 

 star, in the spectrum of wmich iron is represented by traces of the 

 triplets characteristic of the arc spectrum as well as by the enhanced 

 lines, must be cooler, so far as the absorbing iron vapour is concerned, 

 than one in which iron is represented by the enhanced lines alone. 

 Similarly, we must conclude that a star in which iron has no repre- 

 sentative lines is hotter than one in which the enhanced lines appear 

 without the arc lines. In practice the iron lines furnish a much 

 more convenient indication of stellar temperature than the continuous 

 spectrum, for the reason that in the case of iron no special photo- 

 graphs are necessary, while for an investigation of the continuous 

 spectrum special photographs of stars with very carefully controlled 

 exposures have to be taken. 



Calcium. 



The calcium lines in the spectrum of « Orioms indicate that we 

 have in that case a temperature not greatly differing from that of 

 the second stage ; the blue line which is characteristic of low 

 temperature, as well as H and K, being very strongly developed (see 

 Map I). 



On passing to the stars of successively higher temperatures which 

 have already been studied in the case of iron, it will be seen from 

 Map I, that the blue line has disappeared at the temperature of 

 x Cygni, while the H and K lines persist with gradually reduced 

 intensities up to the hottest stars. The intensities of the H and K 

 lines as compared with the blue line fully bear out the results as to 

 the order of temperature of the stars which has been derived from a 



* ' Boy. Soc. Proc.,' vol. 60, p. 475. 



