Studies in Visual Sensation, 



451) 



Ceoonian Lecture. — " Studies in Visual Sensation." By C. 

 Lloyd Mokgan, F.E.S., Principal of University College, 

 Bristol. Lecture delivered March 21, 1901, — MS. received 

 March 25, 1901. 



Peculiar difficulties are encountered when any attempt is made to 

 express the relative values of sensations in quantitative terms which 

 shall make some approach to exactness. No doubt we commonly deal 

 with the less and the more of sensation ; we say that a surface appears 

 duller or brighter; but on what scale shall we determine with any 

 precision how much the less, or by what amount the more 1 What is 

 to be our unit of sensation in terms of which we can reckon our gains 

 and our losses ? At first sight it may seem reasonable to assume that 

 the unit of sensation is that which corresponds to some definite and 

 constant amount of physical stimulus or physiological excitation. 

 And unquestionably we seem justified in asserting that under constant 

 conditions, physical and physiological, a given amount of stimulus 

 produces an amount of sensation which is constant in quantity. If it 

 be not so the relation of stimulus to sensation is not a subject that is 

 open to scientific investigation. But apart from the fact that there is 

 some variation of sensitiveness among different individuals, and even 

 in the same observer at different times, there are many familiar facts 

 which show that the physical measurement of luminosity does not 

 accord with the estimates we make of the brightness of the illu- 

 minated surface. If a sheet of white paper be illuminated by a 

 standard candle at* a given distance it appears of a given brightness ; 

 if now the distance of the candle be doubled, the physical luminosity* 

 is reduced to one-fourth. But it looks a good deal more than one- 

 quarter as bright. Its brightness may not be even halved. Again, 

 the physical luminosity of coloured paper, as measured by Sir Wm. 

 Abney's methods, does not give values which satisfy sensation. A blue 

 with luminosity 9, as compared with white paper reckoned as 100, 

 appears to have a brightness nearly half-way between black and white ; 

 a red with luminosity 18 does not certainly appear twice as bright as 

 the blue. Furthermore, it is well known that a series of equal incre- 

 ments of stimulus does not produce a similar series of equal increments 

 in sensation. 



This may readily be illustrated by means of a rotating disc. If a 

 disc be prepared with equal sectors of black and white, the effect on 

 the eye, when rotation is sufficiently rapid completely to extinguish 

 flicker, is that of a uniform grey. But it is a grey so light as to be 

 not far removed from white. We may assume that the physical 

 luminosity of the surface is, since the sectors are equal, the arithmeti- 

 cal mean between that of the white and that of the black employed. 



2 K 2 



