240 PROCEEDINGS OF THE AMERICAN ACADEMY. 



considered provisional : it is the best that I can do at the present 

 time, and I believe that it is correct in the main. We shall see pres- 

 ently, however, that photographs of the fluorescence stimulated by 

 monochromatic radiation will have to be made with the large concave 

 grating before we can be absolutely sure of all of the lines. We will 

 drop the magnetic spectrum for the present and consider 



The Fluorescence Spectrum. 



In the previous paper I have described some of the remarkable 

 changes which take place in the distribution of energy in the fluo- 

 rescence spectrum of sodium vapor when the wave-length of the 

 exciting light is changed. With white-light stimulation the general 

 appearance of the spectrum is shown in Plate 4, Figure 6, A. There 

 is, in addition, a broad double band at the position of the D lines, and 

 a red-orange spectrum which, when the vapor is dense, is distinctly 

 banded. In the present paper we shall be concerned chiefly with 

 the portion figured above, for it is in this region that most of the 

 remarkable changes occur. As will be seen, it is comprised between 

 wave-lengths 460 and 570, and is devoid of any apparent regularity in 

 the distribution of its lines, except in the region above X = 505 where 

 we have lines spaced with considerable regularity, the spacing becoming 

 less as the wave-length increases. The distribution of intensity in this 

 portion of the spectrum is such as to give it a fluted appearance, the 

 flutings being most conspicuous in the region between A = 505 and 

 X = 535. With white-light stimulation the flutings cannot be made 

 out above 540, as can be seen from Plate 4, Figure G, B, in which the 

 upper limit of this part of the spectrum is shown. The fine lines are 

 present in this region, becoming, however, less and less distinct as the 

 upper limit of the spectrum is approached. If, now, instead of stim- 

 ulating the vapor with white light, we employ blue light in the region 

 460-465 obtained from a spectroscope for the excitation, the fluorescent 

 spectrum presents a totally diff'erent appearance (Figure 6, C). The 

 blue region, corresponding in its range to that of the exciting light 

 (indicated by a double arrow), appears as before, and the upper limit 

 of the spectrum between wave-lengths 540 and 565, the intermediate 

 jxjrt'ton being entirely absent, as shown in the lower spectrum of Fig- 

 ure 6, C. Furthermore, at the upper or yellow end, there now appear 

 the flutings, which were absent when the fluorescence was stimulated 

 with white light. 



If, now, we gradually increase the mean wave-length of the exciting 

 light, the region of maximum intensity in the fluorescence spectrum 



