1 82 STUDIES IN LUMINESCENCE. 



From these values the observed curves in Figs. 176 and 177 were plotted. 

 The observed points are indicated by circles. The crosses and dots in this 

 figure refer to corrected values referred to in a subsequent paragraph. In 

 the case of eosin the wave-length of the crest is nearly the same as that of 

 the green mercury line. Measurements in the region of the crest were there- 

 fore rendered uncertain by the presence of stray light from the green line. 

 For this reason all the measurements in this region have been discarded. 

 Observations near the crest of the fluorescein curve were also somewhat 

 discordant, so that we do not regard this part of the curve as determined 

 with much accuracy. 



To obtain from the observed curves the distribution of energy in the 

 fluorescence spectra it was necessary to make corrections for slit-width and 

 absorption, and to multiply the ordinates of each corrected curve by the 

 ordinates of the same wave-length in the curve giving the distribution of 

 energy in the spectrum of the acetylene flame. 



THE CORRECTION FOR SEIT WIDTH. 



In the spectrophotometric comparison of sources of light having con- 

 tinuous spectra and nearly the same luminosity curves the correction for 

 slit width disappears ; but in the case of spectra consisting of narrow bands 

 this is far from being the case. The slit-width correction used in the deter- 

 mination of the energy curves of incandescent solids does not apply, partly 

 because it is the luminosity of the rays rather than their energy that is 

 important, and partly because the distribution of luminosity in two sources 

 has to be considered. The slit correction applicable to the Lummer- 

 Brodhun spectrophotometer may be derived as follows : Let the luminosity 

 curve of the source Si, in front of the slit A, have the equation 



when the distance of the source is such as to give the standard intensity, 

 which we shall call unity. If the distance is varied so that the intensity 

 becomes i, then (X being the wave-length) 



The luminosity of the source S 2 in front of the slit B will then be given 

 by the equation 



L 2 =rf(\) 



where r is the ratio of the energy of S 2 at the wave-length X to the energy 

 of Si at the same wave-length, r is itself a function of X unless the two 

 sources are identical in quality. 



Images of the slit A are formed in the focal plane of the telescope for each 

 wave-length of the spectrum of Si. If the spectrum is continuous, we 

 therefore have a series of overlapping images, forming a spectrum of greater 

 or less impurity according to the width of the slit A . 



The light reaching the eye from the slit A will depend upon the width 

 of A, being proportional to this width if other conditions remain constant 

 But it will also depend upon the width of the aperture C at the principal 

 focus of the telescope, through which aperture the light used in making the 

 setting must pass. If the spectrophotometer is used without an eye-piece, 



