I 



and Absorption Spectra of Sodium Vapour. 369 



tlie iliagranis \vliere the deviated continuous spectrum is 

 shown. To make this diagram exhibit the changes just 

 ilescribed, one has only to move a narrow vertical slit across 

 it from right to left. 



It was impossible to tell whether flutings were present in 

 any portion of ilie fluorescent spectrum or not, owing to the 

 comparatively wide source, and small dispersion employed. 

 There is no reason for believing that they are not, but it does 

 not seem sate to assume that the spectra are identical when 

 the fluorescence is produced by white light and monochromatic 

 blue light. The extent ot" the s])ectrum, and the distribution 

 of the intensity in it, has been shown to be different in the two 

 cases, and the flutings, say in the green, which are present 

 when white light is used, may be absent when the fluo- 

 rescence is produced by light of much shorter wave-length. 

 To test this the spectrum will have to be photographed when 

 the tube is illuminated with monochromatic blue light. To 

 accomplish this with the present a})paratus would probably 

 require an exposure of six or eight hours. More favourable 

 conditions are expected with a new form of tube now^ under 

 construction, and it seems probable that we shall eventually 

 learn a good deal about the dynamics of the sodium molecule. 

 The results already obtained apparently prove that the light 

 of longer wave-length emitted when the vapour is fluorescing, 

 is not the result of damping of the vibration, as assumed in 

 Lommel's theorv, but an emission resultino- from the fact that 

 the ions of longer free period are set in vibration either by the 

 waves of shorter period, or by the ions of short period v/hich 

 are thrown into vibration by these waves. It is not worth while 

 to speculate about this, however^ until the fluorescent spectrum 

 produced by monochromatic blue light has been photographed. 



In general the fluorescence of a substance has its maximum 

 intensitv w^hen the wave-lenpth of the lioht is that of the lio-ht 

 most strongly absorbed. Sodium vapour is an exception tc 

 this rule, for the D-line absorption, which is far more intense 

 than the fluted absorption, has nothing to do with the fluo- 

 rescence. On Lommel's theory of fluorescence the absence 

 of any lateral emission of light by an absorbing medium is 

 explained in one of two ways. Either the absorption is of 

 waves of different period (an octave below for example) from 

 the free period of the ions, or else the damping factor is so 

 large that the emitted light lies in the infra-red region. In 

 the case of sodium vapour neither explanation is sufficient 

 to account for the absence of fluorescence when the wave- 

 length of the exciting light is that of the D-lines, for since 

 Kirchhoff^s law is obeyed in this case, the absorbed waves 

 and emitted waves have the same period, and the absorption 

 Phil. Mag. S. 6. Vol. 6. No. 33. Sept. 1903. 2 B 



