406 Messrs. Nichols and Snow on the Influence of 



still much the brighter of the two. The measurements con- 

 sisted in turning the movable Nicol until the spectrum of 

 lamp L was reduced, region for region, to equality with that 

 of the spectrum due to the pigment. 



In presenting the results of measurements with the spectro- 

 photometer it is not always desirable to reduce the intensities 

 to absolute measure, since it is with the luminosity rather than 

 with the distribution of energy that we have to do. One of 

 the most satisfactory ways of defining colour is to compare it 

 with white as a standard, and that is the method which we 

 have adopted for the purposes of this investigation. The 

 ideal white is one which reflects all the wave-lengths of the 

 visible spectrum in the same proportion. Whatever be the 

 character of the illumination to which a body the colour of 

 which is the ideal white is subjected, therefore, the incident 

 and reflected rays will be identical in quality. Since no such 

 body is known to exist, however, it becomes necessary to 

 select some actual pigment as a reference-standard. 



Dr. Arthur Kcenig, in one of his papers on colour-blindness, 

 has suggested the adoption of magnesium oxide, obtained 

 from the smoke of the burning metal, as the normal white *. 

 Between this substance and magnesium carbonate, which one 

 of the present writers has repeatedly used as a standard!, 

 there is little to choose, so far as selective reflexion is con- 

 cerned. Neither of them can be considered a true white, in the 

 sense in which that term is to be understood in spectro- 

 photometry J, but the oxide possesses two great advantages: 

 the film is very readily produced, and the tint and the degree 

 of brilliancy reappear in each new specimen with a constancy 

 which leaves nothing to be desired. In our experiments, then, 

 magnesium oxide was adopted as the reference-white, and the 

 ideal white, in terms of the brightness of which all pigments 

 were to be measured, was defined as a surface which, reflecting 

 all wave-lengths of the visible spectrum equally well, possesses 

 the same reflecting-power for the wave-length '5890 //. (region 

 of the D line) as does the smoke-film of magnesium oxide. 



Our first step was to determine the correction-factor for the 

 select : ve absorption suffered by the light from the lamp L, in 

 traversing the lens A, and the Nicol prism. For this purpose 

 a block of the carbonate of magnesium was cut in two. The 



* Wiedemann's Annalen, xxii. p. 573. 



t E. L. Nichols, American Journal of Science, xxviii. p. 343. 

 % E. L. Nichols, "On Black and White," Transactions of the Kansas 

 Academy of Science, vol. x. p. 1 (1886). 



