THE BRIGHTNESS OF COLOUR SENSATIONS. 1077 



the brightness curves of normal and abnormal vision are based on this direct 

 comparison. At first sight it seems very difficult to say with any degree of 

 accuracy that a coloured field has the same brightness as a colourless field ; 

 and some, including Helmholtz, have been unable to make satisfactory 

 observations of this kind, but most of those who have used the method agree 

 that it is practicable, and gives more consistent results than might be expected 

 from its purely subjective nature. Rood 1 and Abney tested their results by 

 mixing two complementaries, of which they had determined the white 

 equivalents, and found that the grey so produced was equal to the sum of 

 the white equivalents of the components. 



As might be expected, this direct method of comparison appears to be 

 especially easy in the more simple colour system of the colour-blind. 



The direct method must, however, always be more or less arbitrary ; G. 

 E. Miiller 2 has lately objected that certain colours have a " penetrative " 

 quality, which may be confused with brightness ; and it is certainly very 

 dangerous to draw conclusions in respect to slight individual differences of 

 colour vision from such a method. Numerous indirect methods have now 

 been described to determine the relative brightness in a more objective way. 



One of these methods, devised by Vierordt, 3 consists in finding the amount 

 of white light which must be added to a coloured light, to render the colour of 

 the latter imperceptible. It is, in fact, a method by which the brightness 

 of a part of the spectrum is determined by finding its threshold of saturation. 

 Vierordt's scale of intensities for the spectrum agrees roughly with Fraunhofer's. 

 A similar method was employed by Draper 4 Avith very different results. This 

 observer found that with the prismatic spectrum red was the brightest colour, 

 while with the diffraction spectrum all parts were equally bright, all colours 

 disappearing with the .same amount of added white light. 



Another method is based on the principle that the degree of visual acuity 

 depends upon intensity. This is interesting as the earliest method employed, 

 having been used by Herschel in 1800, 5 who determined the distinctness with 

 which he could with the microscope observe fine details of opaque objects 

 illuminated by the different parts of the spectrum, and placed the maximum 

 intensity in the yellow and green. This method has been most thoroughly 

 investigated by Mace de Lepinay and Nicati, 6 who found for the longer 

 wave half of the spectrum that it gave results agreeing fairly with those of 

 Fraunhofer and Vierordt. They considered two lights as equally bright when 

 the visual acuity in them was of the same value. 



One of the most satisfactory methods is known as flicker photometry. The 

 point at which intermittent stimuli give rise to a continuous sensation seems to 

 depend on brightness and to be independent of colour- tone. The measurement 

 of the intensity of sensation by means of time was suggested by Talbot 7 in 1834, 

 and was first put in practice by Schaf hautl 8 and Rood. 9 Probably the most 

 simple application of the method is to rotate open sectors before a field half- 

 filled with the colour whose brightness is to be determined, while the other 

 half shows colourless light which can be varied in intensity. The two halves 

 are judged to be equally bright, when both show a just perceptible flicker. 

 Results obtained by this method agree closely with those obtained by other 

 methods, and fairly closely with those derived from direct comparison. 



1 Am. Joum. Se., New Haven, 1878. vol. xv. p. 81. 



2 Ztschr. f. Psychol, u. Physiol, d. Sinnesorg., Hamburg u. Leipzig, 1896, Bd. x. S. 28. 



3 " Die Anwendung des Spektralapparats zur Messung d. Starke des farhigen Liclits," 

 Tubingen, 1871. 



4 London, Edinb. and Dublin Phil. Mag., London, 1879, Ser. 5, vol. viii. p. 75. 



5 Phil. Trans., London, 1800, vol. xc. p. 255. 



G Ann. de him., Paris, 1881, tome xxiv. p. 289 ; and 1883, tome xxx. p. 115. 



7 London, Edinb. and Dublin Phil. Mag., London, 1834, Ser. 3, vol. v. p. 327. 



8 Abhavdl. d. bayr. Akad., 1855, Bd. vii. S. 467. 



9 Am. Joum. Sc, New Haven, 1893, vol. xlvi. p. 173. 



