Studies in Visual Sensation. 



465 



it was found that the three grey rings required very unequal amounts 

 of proportional reduction in order to render them of uniform shade to 

 the eye. As the mean of three sets of observations, the dark grey 

 ring required 50 per cent, reduction of the white at its outer border ; 

 the mid-grey ring 40 per cent. ; the light-grey ring 25 per cent. 

 These figures give only a rough and preliminary approximation to a 

 quantitative estimate in terms of physical stimulation of the effects of 

 contrast under certain conditions of illumination and for speeds of 

 rotation sufficiently rapid completely to get rid of any flicker effect. 

 If the illumination be materially reduced or if flicker occur, the 

 contrast effects within the rings reappear. In other words, with 

 reduced illumination or with that flicker effect which has recently 

 been studied by Professor Sherrington,* a large proportional amount 

 of reduction is required. 



The quantitative estimate of contrast and its physiological bearing, 

 cannot here be further discussed. The markedly different effects in 

 the several rings is sufficient to suggest that we have here a sufficient 

 cause for the discrepancy between results obtained by the method of 

 ring grading and those reached through continuous shading. For the 

 present, however, I am not prepared to do more than suggest that the 

 curve for continuous shading affords a more trustworthy scale for 

 comparing the relative values of stimulus and sensation than is 

 afforded by graded rings which do not appear of uniform shades of 

 grey throughout their width. I provisionally accept therefore the 

 curve through 12 per cent, mid-point as a basis for further experi- 

 mental work. 



I must here confess that in a previous paperf I gave far too high a 

 percentage for the mid-point. But the black I then used was not 

 nearly so deep, the white was not quite so brilliant ; I failed to make 

 due allowance for the so-called optical illusions before mentioned ; and, 

 the worst error as I now see, I used ring grading as a check on 

 continuous shading, not realising that the effects of contrast vitiated 

 the results in the manner in which I have just attempted to indicate. 



I may now pass on to consider another fact which shows the import- 

 ance of conducting observations in visual sensation under approxi- 

 mately uniform conditions of illumination. Suppose that with a given 

 illumination we have obtained even shading or fairly equal steps on a 

 ring disc, and suppose that the illumination be then materially 

 diminished. The one disc no longer gives even shading; the other no 

 longer gives rings with equal sensation steps. Delbceuf| drew atten- 

 tion to this fact for discs with grey rings, and accounted for it by a 

 somewhat far-fetched hypothesis of physiological tension. No such 



* 'Journal of Physiology,' vol. 21, p. 33 (1897). 



t ' Psychological Eeview,' May 1900, p. 217 (vol. 7). 



X ' Exaruen Critique de la Loi Psychophysique,' 1883, pp. 147-48. 



