258 PROCEEDINGS OF THE AMERICAN ACADEMY. 



The Series in the Magnetic Spectrum. 



As we have seen, the complex fluorescent spectrum is made up of six 

 or more series of lines, the individual lines of each series being about 

 38 Angstrom units apart, the spacing becoming less as we pass from 

 yellow towards violet. The fact that the lines in the magnetic 

 spectrum coincide with the lines of the complex spectrum makes 

 it seem certain that the same series will be found there. By com- 

 paring the various fluorescent spectra with the magnetic spectrum, 

 and by measuring carefully the distances between the lines of the 

 latter, I have made a provisional assignment of the magnetic lines thus 

 far observed into five series. 



The wave-lengths and wave-length diff"erences are given in the 

 tables on page 259. 



As will be seen by reference to the chart, the first series has the 

 largest average spacing, and the fifth the smallest, the " scale," if the 

 term is allowed, decreasing gradually from the first to the fifth, the two 

 coming into coincidence at about wave-length 4860. 



Doubtless these series could be extended to wave-length 5500 or 

 thereabouts by making use of the grating photograph of the complex 

 fluorescence obtained by white-light stimulation. The lines are, 

 however, so diffuse in their nature, with overlapping wings and other 

 peculiarities, that I have not yet attempted any further extension. 

 I think that by employing a much denser vapor the magnetic 

 spectrum can be considerably extended, and as the lines are much 

 sharper in this case, an extension of the series will be an easy matter. 

 The other series necessary to give the close spacing found in the 

 yellow may be discovered in this manner. 



A theoretical discussion of the results will have to be deferred 

 to a subsequent paper. In fact, I prefer to have this side of the 

 subject attacked by those who have given especial attention to the 

 theory of molecular radiation. The absence of many lines in each 

 series in the magnetic spectrum, and the absence of certain lines in 

 the fluorescent spectrum, are especially suggestive. We have similar 

 conditions in other series of lines, as is well known, but the present 

 case is, so far as I know, the only one in which we can, by varying the 

 exciting conditions, bring about a change in the position of the absent 

 lines. It appears to me that the data furnished us by sodium vapor 

 ought in the end to enable us to choose between the various theories 

 proposed to account for spectrum series. 



The investigations recorded in this paper have been made possible 



