Lines in the Sj)ectra of Sta rs of the /3 Oriicis Type. 49 



It is a. remarkable fact that these three stars, which may be con- 

 sidered as amongst the best examples of silicon stars in the light of 

 the spectriuu of silicon hitherto known, do not sliovj the three silicon 

 lines which are so prominent in /? Crucis, e Canis Majoris, &c. Scheiner 

 has measured the spectra of all three stars* in this region, but does 

 not record the lines in his Table of Wave-Lengths. 



Their absence from the spectra of these stars (as well as the presencp 

 of Lockyer's enhanced silicon lines) is fully confirmed by photographs 

 taken here ^vith the special object of searching for the new silicon 

 lines in the best known silicon stars. 



This can be readily understood in the light of the experiments 

 with the tube of silicon tetrafluoride. 



With the highest disruptive spark, Lockyer's silicon lines 41 28*6 

 and 41 31 '4 are much enhanced as compared ^vith the lines 4552*79, 

 4567*09, 4574*68, and it was found possible by suitable exposure to 

 obtain the two enhanced lines without the presence of the other three 

 lines becoming evident. 



The latter lines would be much more rapidly obliterated in the 

 absorption spectra of stars, than in the bright line spectrum from 

 the tube, and therefore their absence from certain stars in which the 

 enhanced lines are strong need not occasion much surprise. 



In other stars, however, all five lines are present. Lockyer has 

 recorded them in Bellatrix and their presence has been confirmed by 

 photographs of the spectrum of this star taken here. Mr. McClean 

 has measured all five lines in p Crucis where Lockyer's enhanced 

 silicon lines are certainly not so conspicuous as the lines 4552*79 and 

 4567-09. 



The same may be said of e Canis Majoris in which star the new 

 silicon lines are very prominent, whilst the enhanced lines are very 

 faint. 



In the silicon spectrum from the argon and helium vacuum tubes, 

 the enhanced lines noted by Lockyer are by no means so prominent 

 as they are in the silicon spectrmn, obtained from silicon tetrafluoride 

 with the intense disruptive spark. It is evident, therefore, that great 

 variations in the relative intensities of the silicon lines occur in steUar 

 spectra, and that such variations can be produced to a certain extent 

 in the laboratory, and these require further investigation. 



The behaviour of the silicon lines will give us valuable data for the 

 elucidation of the problem of relative stellar temperatures. 



It is clear that if we regard, with Lockyer, the lines 4128*6 and 

 4131*4 to be the enhanced lines of silicon and their presence, enhanced, 

 to be a criterion of a higher temperature than occurs in stars where 

 these lines are not enhanced, it must follow that such stars as a Cygni, 

 Eigel, and Sirius are hotter than Bellatrix, Crucis, and e Canis 

 * Sclieiner'5 ' AstroQomical Spectroscopy.' 



