Lines in the, Spectra of Stars of the /3 Grucis Type, 



45 



in the spectra of P Crucis, e Canis Majoris, and stars of their type, 

 viz., wave-lengths 4552-79, 4567-09, 4574-68. 



Mr. McClean had previously also recorded these lines in his measures 

 9f the spectrum of P Crucis"^ as wave lengths 4552-6, 4567*5, 4574*5, 

 but beyond pointing out the approximate coincidence of the first of 

 these with lines due to barium or titanium, he assigns no origin to 

 them. 



Sir Norman Lockyer frequently records them as unknown lines. In 

 his recent paper " On the appearance of the Cleveite and other New 

 Gas Lines in the Hottest Stars" (June, 1897),t he records all three 

 lines as unknown. The first occurs in a map of the spectrum of 

 Bellatrix as a line in a probable new series found by Dr. W. J. S. 

 Lockyer. The second and third lines, given as 4566-8 and 4574-8, 

 occur in a Table of Lines which Sir Norman regards as belonging, 

 with high probability, to gaseous substances which have yet to be 

 discovered. 



It will be noticed that the connection existing between the three 

 lines is not there recognised. 



In none of the Tables of Wave-Lengths available for reference at 

 the Cape could any satisfactory clue l)e obtained as to the origin of the 

 lines. 



During some experiments made with a view to securing the best 

 elementary line spectrum of oxygen as a comparison spectrum for stars 

 of the /5 Crucis type, I fomid that a tube of carbon dioxide gave the 

 best results, being freer from impurities, and giving stronger oxygen 

 lines than any of the oxygen tubes at my disposal. By the use of a 

 jar and air gap in the secondary circuit of the coil the gas was dis- 

 sociated, and gave the spectra of carbon and oxygen. During use, the 

 carbon dioxide tubes became more vacuous, and, with a view to 

 obtaining a brighter discharge and shorter exposure, I passed the 

 induced current from an 18-inch Apps' coil, using four large jars and an 

 air gap. 



"Whilst using these electrical conditions, I happened to expose an 

 argon tube marked 2 mm. (pressure), and on developing the photo- 

 graph was much surprised to find that it, too, gave the Avell-recognised 

 lines of oxygen. Stronger than these I at once noticed two lines at 

 the green end of the spectrum, which recalled the lines in ^ Crucis, 

 which were unknown terrestrially, whilst the expected argon spectrum 

 was almost entirely absent. 



On comparing the negative, film to film, with one of /5 Crucis, and 

 allowing for the diff'erence of temperature conditions under which the 

 two negatives were taken, the identity of the three unknown lines in 



Crucis with three lines on the argon negative was at once apparent, 



* ' Spectra of Soiitliern Stars ' (Stanford, 1898), p. 13. 

 t ' Key. Soc. Proc.,' toI. 62, p. 60. 



r 2 



