596 



MR WILLIAM SWAN ON THE GRADUAL PRODUCTION OF 



within the Umits of the observations the brightness of the light increases in exact 

 arithmetical proportion with the time during which it acts on the eye. Since the 

 observed intensities of the lights when projected, as in the figure, are all nearly 

 included in a straight line passing through the origin, it may naturally be in- 

 fen-ed, that the impression of light commences at the instant of its incidence 

 on the retina. This conclusion is strengthened, when it is recollected that the 

 preceding experiments prove that light, which is incident on the eye only 



18432 



of a second, produces a distinct impression, while, according to Professor Wheat- 

 stone, less than the millionth part of a second is necessary for this effect. It has 

 also been proved (see p. 594), that up to ^^ of a second, the impression produced 

 by light is proportional to its duration. It seems, therefore, highly probable, that 

 ft-om 0" up to 0"-05, the brightness of a luminous impression is exactly proportional 

 to tlie tims during which the light has acted on the eye.* 



These experiments, therefore, confirm, in a very satisfactory manner, the 

 inference which has already been drawn from the previous investigation, as the 

 observed intensities of the flashes are very nearly proportional to their duration ; 

 while, at the same time, they exhibit the actual numerical ratio of the apparent 

 Ijrightuess of a flash of a certain duration, to that of the light which produces it 

 acting continuously on the eye. 



VI. The time required for the complete production of Luminous Impressions is independent 

 of the apparent intrinsic brightness of the light. 

 The following series of experiments was made partly to confirm the result al- 

 ready obtained ; but more especially in order to ascertain whether the time re- 

 quired for the complete development of luminous impressions varies with the 

 brightness of the light by which they are produced. In this set of experiments, the 

 same sectors were used as in the last ; and the cu-cumstances were identical, except 

 that the fixed light was placed 8-5 inches (G V 2) from the screen, so that the 

 brightness of the incident light was reduced to half its former intensity. The fol- 

 lowing Table exhibits the mean of ten observations with each sector ; the bright- 

 ness of the light seen by continuous vision, being expressed by unity. 



* In an experiment made since this paper was read, I have found that the same law extends to 

 impressions lasting for jVth of ^ second, of which the ohserved brightness was 0-6118. 



