﻿1136 Prof. M. N. Saha on Temperature Ionization of 



7 electrons in the outermost shell ; its steric factor is there- 

 fore expected to be much higher than that of calcium, 

 which probably accounts for its comparatively high degree 

 of ionization. 



According to Lockyer, Mn + is represented in stellar 

 spectra by X — 4344*19. It does not occur in King's 

 furnace spectra *, though groups of lines due to Mn + occur 

 about X= 3442-3497, 2914-2940. 4344*19 is certainly not a 

 resonance-line of Mn + . Probably it is of the same type as 

 He + 4686. According to Lockyer, this line vanishes from 

 the stage f Tauri or ft Persei. Thus in stellar ranges we have 

 not only Mn and Mn + , but Mn ++ as well. None of the 

 lines of Mn ++ seems to be known. 



Let us now treat some of the other elements in the order 

 in which they occur in the periodic tables. For elements of 

 the third group, e. g. Al, Sc, Y, La, neither the series 

 classification nor the variation in intensity of lines in stellar 

 spectra is satisfactorily known. We pass to the next group. 



Group IV. Carbon, Silicon, Titanium. 



These elements are extremely interesting, because they 

 have got 4-valency electrons, and the steric factor is expected 

 to be unusually large not only for the neutral, but also for 

 the singly- and doubly-cbarged atoms. As a result, the 

 successive stages of ionization will follow each other in 

 rather quick succession. Unfortunately, the knowledge or! 

 the spectra of these elements, as well as of the variation of 

 their intensity in stellar classes, is not so well known. 



Carbon. — The spectrum of this important element is one of 

 the least known. The line A, = 4267 is supposed to be due 

 to C + ; Lockyer, Baxandall, and Butler f treat the lines 

 4650*8, 4647*4 as specially enhanced : we can assume that 

 they are due to C ++ . One wonders what lines are to be 

 attributed to carbon itself. There are two strong lines, 

 X= 6583*0, 6577*5, which may be due to carbon. The 

 variation in intensity of the lines of C + and C ++ in stellar 

 spectra is given below : — 



B9. B6. B5. B3. B2. Bl. Bo. Oe5. Oe. 



\=4267(C + ) 



B6. 



B5. 



B3. 



B2. 



Bl. 



Bo. 



2 



3 



5 



5 



3 



2 



— 



— 



1 



4 



10 



25 



A=: 4647-6( C+ + )- ~ - - - 1 4 10 25 15 © 

 {' Present, but intensity not exactly measured. 



* King, Astro. Journal, vol. liii. (1921). 



t Lockyer, Baxandall, and Butler, Proc. Roy. Soc. Lond. vol. lxxxii. 

 p. 352. Fowler, Report on Series in Line Spectra, p. 163. 



