the Spectrum of Sodium. 515 



6. That in Series D there are two new terms corresponding 

 to negative values of n, for one of which n is approximately 

 — 860-6, and for the other —31-17. These positions are the 

 same as \=1162, and A, = 32082. The first of these is 

 probably too far in the ultra-violet and the other much too 

 far in the ultra-red, to be observed. 



7. Similarly, in the hydrogen spectrum there seems to be a 

 new line in the ultra-violet, viz. the line obtained by putting 

 m=l into Banner's formula. This makes n— —822-789, 

 which is the same as X=1215, a position, however, which is 

 probably too far removed in the ultra-violet for observation. 



8. Lines corresponding to negative values of n do not 

 appear to have been observed in any of the monad elements 

 except sodium, but examples of them are met with in some 

 of the triple-line series of the dyads. Kayser and Runge 

 record what is presumably one triple group of this kind in 

 the spectra of zinc, cadmiam, and mercury, and what is 

 perhaps a second group in the spectra of zinc and cadmium 

 (see the photographs they give of a part of each of these 

 spectra, and the observations they make about them on p. 71 

 of their fourth Paper, Ueber die Spectren der Elemente, in the 

 Transactions of the Berlin Academy for 1891). 



It is not yet known what kind of perturbation within the 

 molecules would be competent to affect the partials of the 

 undisturbed motion of an electron so as to resolve the resulting 

 lines into triple lines. But it is, nevertheless, suggestive to 

 find that in the spectra of Zn, Cd, and Hg the constituents 

 of the triple line corresponding to a negative value of n are 

 not reversed, but in the same relative positions to one another 

 as are those furnished by positive values of n. If this non- 

 reversal of position prevails among those double lines of 

 sodium which are due to negative values of n, it will probably 

 be indicated by the less refrangible constituent of the double 

 line, No. 1 of Series S, being the brighter, as is said to be the 

 case with the double lines of the same series, which are due 

 to positive values of n. This would imply a physical fact of 

 importance, viz. that a change in the sign of n induces a 

 change in the direction of the apsidal perturbation as well as 

 in the direction of revolution in the elliptic partial. If, on 

 the other hand, it is found that the more refrangible con- 

 stituent of the double line is the stronger, this will go far to 

 prove that the direction of the apsidal shift is independent of 

 the sign of n. It is therefore desirable to ascertain by observa- 

 tion which constituent is the brighter. 



9. Finally, our investigation makes it probable that there 

 is some connexion between Series D and Series S in the 



