Method of Measuring the Luminosity of the Spectrum. G05 



to reflect white light to a similar mirror mounted on the 

 eyepiece H, which reflected it down through a hole in 

 the eyepiece upon the polished sloping top of a steel pointer, 



Fig. L 

 to 



fcfr* 



p ' ' 







U 



with which the instrument was provided instead o£ cross- 

 hairs, which finally reflected the light to the eye. The 

 observer, therefore, looking into the eyepiece could see a 

 small patch of white light, and immediately above it a patch 

 of suitable size of any desired colour of the spectrum. The 

 white light could be reduced in intensity to any desired 

 amount by changing the inclination of the mirrors at C or H, 

 while the intensity of the spectrum was controlled by rotating 

 the polarizer (E). When the disk (D) rotated both patches 

 of light flickered necessarily at the same rate. 



The spectrometer used was of the Hiiger automatic type, 

 w T hich gave a dispersion slightly in excess of twelve degrees. 



In measuring the luminosity of the spectrum the method 

 of procedure was, first, to lower the intensity of the patch 

 of white light until it was of the same luminosity as a patch of 

 violet (X^'414^) of undiminished brightness, the principal 

 sections of the nicols being parallel, as near the end of the 

 spectrum as it was possible to make exact measurements 

 upon. The intensities of the w r hite and violet lights were 

 considered equal when the critical frequency of flicker was 

 the same for both. The white light now became the standard 

 of comparison, and was maintained continuously at this low 

 intensity through all the observations. Each selected part 

 of the spectrum was in turn brought into view and reduced 

 to the luminosity of the white patch by rotating the polarizer 

 an amount depending on the brightness of the part of the 

 spectrum under observation. In every case the same critical 



