.S50 SIR w. DE w. ABNPIY : MODIFIED APPARATUS FOR MEASUREMENT OF COLOUR 



90 



80 



\ 



z 



\ 



X 



I 



\ 



"5 



! 



X 



X 



\ 



\ 



20 



10 



\ 



4500 



5000 



5500 6000 



or WAVE tf/var#s 



6500 



Fig. 6. (Normal spectrum) Curves of equally stimulated red, green, and blue sensations to form white. 



PART II. 



(19.) A Colour Defined Inj a Ware-length, &c. 



In a note " On the Numerical Registration of Colour," which I communicated to 

 the Royal Society ('Proceedings,' vol. 49, 1891), it was indicated that any colour 

 could be accurately defined by a wave-length, its luminosity, and the percentage of 

 white light that it contained. In Table III. we have a very ready means of stating 

 all these with extreme accuracy. 



If the percentage of each colour of the spectrum which a coloured medium or a 

 pigment transmits or reflects be known from measurement, then from Table III. we 

 can find the wave-length, the luminosity, and the percentage of white light which 

 the colour contains. 



(20.) Measurement of Spectrum Intensity. 



I have already described the method employed by myself in measuring the intensity 

 of the light transmitted or reflected. Fig. 7 shows the plan. S is the slit moving in 



Fig. 7. 



