348 SIR W. DE W. ABNEY : MODIFIED APPARATUS FOR MEASUREMENT OF COLOUR 



Columns IX., X.,and XT. give the percentage composition of the different colours of 

 the spectrum in terms of the three colours which best represent the colour sensations 

 when white is deducted from them, viz., red lithium, SSN 37 '5, and SSN 23 '2. 



In reference to this table it may be remarked that of the whole spectrum '225 is 

 white and '775 colour. This shows that the white in the colours is by no means a 

 negligible quantity. 



(17.) Significance of the Inherent White. 



In regarding the table, Column V., for white, it will be remarked that the maximum 

 amount of white is near SSN (42). In ' Colour Photometry,' Part III. (' Phil. Trans.,' 

 1892), it was shown that in this region the light disappeared last when the intensity 

 was reduced. It was also shown that the maximum luminosity of a very feeble 

 colourless spectrum was near this point, and in the concluding page of my last paper 

 on the colour sensations, I pointed out that the presence of the fundamental sensation 

 of light, which is white, must be taken into account in any theory of colour vision. 

 The fact that in these slightly revised measures we get more than indications that 

 white exists in the region where the fundamental sensation has been shown also to 

 exist, leads one to believe that we are in some way separating this sensation from the 

 three-colour sensations. What seems to confirm this view is that when a very bright 

 spectrum, such as is given by sunlight, is measured, there is a tendency for the 

 proportion of white in the region SSN's 48 to 16 to diminish. This is what we should 

 expect to find, since fixed amplitude of wave colour vanishes at some, as also does the 

 fundamental light at a lesser amplitude, be the spectrum feeble or brilliant. 



(18.) The Normal Spectrum Curves. 



Table V. gives the sensation curves for the normal spectrum, and is shown in the 

 same manner as it was in my previously quoted paper. 



