﻿Elements of the Higher Groups. 1135 



Taking the temperature of the spot and the sun to be 

 5000° K. and 7000° K. respectively, and w = 8, we find that 

 for Ba, I ff = 4*22 and 3'8b" respectively, i. e. in the sun 

 barium is ionized like Cs, in the spot like Rb. This very 

 nearly explains the complete ionization of: Ba in the sun, 

 as well as in the spot. 



For Ca, I e =5*18 and 4*82 respectively. Thus in the sun, 

 calcium is more ionized than sodium, while in the spot it ought 

 to be a bit less ionized than sodium. This satisfactorily ex- 

 plains the behaviour of the Ca-lines in the sun and spot 



If we turn to the stellar spectra, we find that calcium S/' 

 disappears from the B8-stage. By using- the original equation 

 of ionization it was found that the temperature could not be 

 less than 13000° K. But, according to "W ilsino- and Schemer's 

 intensity measurements, the temperature is only slightly 

 above 10000° K. Taking n=S, the effective LP. becomes 

 4-21 volts; the ionization is complete at 10000°K*. Thus 

 the introduction of the " steric factor" seems to bring down 

 the temperature of different spectral classes in a line with 

 the temperatures obtained from intensity measurements. 



In my original calculation of the ionization of helium, the 

 I. P. was taken to be 20'4 volts. This is now known to be 

 wrong, for numerous workers have established definitely that 

 real ionization begins at 25*4 volts. The former figures are 

 therefore to be revised. Taking the steric factor = 8, the 

 effective LP. at 25000° K. becomes 21 volts, and ionization 

 is 74 per cent, under concentrations corresponding to one 

 atmosphere pressure. If P = 10 -1 atm., the ionization be- 

 comes complete. Thus the temperature of the Oa stars given 

 in my former paper (Proc. Poy. Soc. Lond., May 1921, 

 p. 151), remains unaltered. 



Manganese. — Let us next consider the element Manganese, 

 because the constitution of its series spectrum has been 

 recently elucidated by Catalan (Phil. Trans, vol. 223). He 

 finds the IS term of Mn = 59937, the I.P. = 7*41 volts. The 

 resonance lines of Mn are the triplet X = 4030*92, 4033*21, 

 4034*62, so that, as far as the variation in intensity of its lines 

 in stellar spectra is concerned, manganese is an ideal element. 

 According to Lockyer, they occur as faint lines in the spectra 

 of Ao-stars (intensity 1 on a scale 1-10, the Ca-line 4227 being 

 of intensity 2). Thire, in spite of the fact that the LP. of Mn 

 is 1*23 volts higher than that of calcium, it. is more highly 

 ionized at T = 10000° K. than calcium. Manganese has got 



* Vide M. N. Saba, " Elements in the Sun," Phil. Mag., Dec. 1920. 



