408 Dr. R. A. Houstoun on a 



To explain the use of this Table take, for example, the 

 wave-length 5842 and the columns headed Kn. Then the 

 Table states that light of this wave-length, according to 

 Konig, can be regarded as made up od 40 per cent, of red 

 of wave-length 6302 and 60 per cent, of green of wave- 

 length 5281. Or take wave-length 4488 and the three 

 columns headed A*00. Then the table states that light of 

 this wave-length, according to Abney's 1900 paper, can be 

 regarded as made up of 9 per cent, of red of wave-length 

 6302, —37 per cent, of green of wave-length 5281, and 

 128 per cent, of blue of wave-length 4569. The minus sign 

 simply means that the colour is to be on the other side of 

 the equation ; thus the colour in question superposed on the 

 green "37 times as bright gives the same hue and luminosity 

 as the red "09 times as bright superposed on the blue 

 1*28 times as bright. 



It will be noticed that Maxwell's two observers differ 

 from the others in stating that the extreme red has one per 

 cent, of blue. This point has been tested specially by 

 Rayleigh in the paper cited, and he has quite definitely 

 decided that Maxwell is wrong, that the extreme red has no 

 tendency to blue whatever. Maxwell's observers' last three 

 wave-lengths are also extremely irregular ; the values for 

 K's last wave-length are omitted on account of their being 

 absurd. Maxwell himself mentions the accuracy of the 

 figures for these wave-lengths as being " doubtful." Also 

 Maxwell's own colour-vision was somewhat abnormal in 

 having an excessive absorption in the yellow spot. I have 

 therefore rejected the observations of K and J. 



The observations of Konig and Dieterici agree fairly well, 

 except for an abnormality in Konig' s at 5728, stated as due 

 to absorption in the yellow spot. I have therefore chosen 

 Dieterici's in preference to Konig's. Abney states with 

 reference to the difference between his two sets of obser- 

 vations (p. 333, 1906 paper) " some slight alteration in the 

 sensation-curves was the result, and, though small, ought to 

 be recorded." His 1906 set must be regarded as superseding 

 his 1900 set. 



The chief difference between his 1906 set and Dieterici's 

 set lies in the violet. Dieterici's last three wave-lengths 

 show a marked tendency towards the red, the last one 

 having 72 per cent, of red. Abney's last three wave-lengths 

 do not show this tendency at all, the last one having —4 per 

 cent, of red. Now this tendency of the violet towards red 

 was investigated by Rayleigh in the paper already cited, and 

 he has decided that there is no doubt about it ; that 



