166 



Mr. J. Norman Lockyer. 



Passing to 7 Cygni, the iron lines confirm the idea that the tem- 

 perature is higher than in a Orionis. Here the enhanced lines of iron 

 are stronger than the arc lines representative of the second stage of 

 temperature ; that is, the temperature is approaching the fourth stage 

 (see Map I) . 



In a Cygni, where, to judge by the extension of the spectrum 

 towards the violet, the temperature is much higher, the enhanced 

 lines of iron are among the strongest in the whole spectrum. At the 

 same time, some of the stronger lines of the stage 2 spectrum, in- 

 cluding the triplets, also appear, but they are very feeble as compared 

 with the lines of stage 4 (see Map I). 



Here, then, we probably have absorbing iron vapour at a tempera- 

 ture very nearly approaching the fourth stage. If this result be con- 

 firmed, we at once get an explanation of the great differences of 

 intensity between the lines in a. Cygni which coincide with iron lines, 

 and those which appear in the spectrum of iron as observed terres- 

 trially with an unshielded arc. 



7] Leonis represents a stage of temperature a little higher than 

 that of cc Cygni ; here the second stage iron lines have disappeared 

 altogether, and the enhanced lines appear alone (see Map I). 



At the higher temperature of stars like Rigel, the enhanced iron 

 lines also appear without any trace of the familiar iron spectrum 

 typified by the triplets (see Map I), and there seems to be little doubt 

 that we are here in presence of iron vapour at a transcendental tem- 

 perature corresponding to the fourth stage to which I have drawn 

 attention. 



In £ Tauri, the iron lines are almost identical with those in Rigel 

 (see Map I). 



At the still higher temperature of Bellatrix, all visible traces of 

 the iron spectrum have vanished from the photographs. 



General Results with regard to Iron. 



The general result of the investigation of the enhanced iron lines 

 in stellar spectra confirms the view that the absorbing regions of the 

 hottest stars exist at a higher temperature than any attainable in 

 laboratory experiments, the spectrum of iron consisting solely of 

 those lines which are enhanced in passing from the arc to the spark. 

 At the same time, some of the lines in the spectra of the hottest 

 stars formerly classed as unknown are now shown to be due to 

 iron. 



The fact that even the enhanced lines themselves disappear from 

 the spectra of stars approaching, or at, the acme of stellar temperature 

 raises another question to which reference will be made later. 



As I pointed out in a preliminary note on the occurrence of these 



