1901.] 



and Melting Points to Atomic Mass. 



7 



yet been observed corresponding to these in the spectrum of barium. 

 On the more refrangible side of these three lines, two lines, corre- 

 sponding to the solar lines H and K, wave-lengths 3968*8 and 3933*8, 

 are found in calcium; two lines, wave lengths 4215*7 and 4077*9, in 

 strontium, and seven lines in barium. The two calcium lines doubt- 

 less correspond to the two strontium lines, but it is very difficult to 

 select the corresponding lines of barium. The two selected, and 

 included on the diagrams, have wave-lengths 3890 and 3500, but they 

 are the only lines about which serious doubt is held. We must not 

 therefore attach much importance to them until the spectrum of 

 barium has been fully examined and the selection confirmed. 



Humphreys and Mohler* found that the H and K lines of calcium 

 were only shifted half as far as the 4226*9 line by changes of pressure. 



Reesef has shown that under the Zeeman effect the 3933 line of 

 calcium becomes a triplet and the 3968 line a quadruplet. 



Lithium and Sodium. — There is no doubt but that the red lithium line 

 corresponds to the yellow (D) lines of sodium. These lines are the 

 strongest in the spectra, and there are no other lines of the principal 

 series of these two elements nearer than wave-length 3303. 



The two subordinate series may be easily arranged, according to the 

 intensity of the lines : the lines of the diffuse series are much stronger 

 than those of the sharp series. 



Potassium, Eubidium, and G cesium. — Lines which, from their relative 

 intensities in all spectra, appear to correspond to one another are : — 



Potassium — 

 4047 



Intensity. 



Bunsen. 



Oxy hydrogen. 



Arc. 



I «? { 



8 

 10 



Qr 



8r 



Eubidium — 



4216 



4202. . . . . . 



9 

 10 



8 

 10 



Qr 

 8r 



Csesium — 









4593 



9 



6 



Qr 



4555 



10 



10 



8r 



These are all given by the same value of m (= 2) in Rydberg's 

 formula for the principal series, and the other lines of the two series 

 follow in order. The lines in the diffuse subordinate series were 



* ' Astrophys. Jour.,' vol. 3, 1896 p. 114. 

 f Ibid., September, 1900. 



