Me. Hunt on the Cinephantic Colour Top. 257 



appear to the writer to account sufficiently for the phenomena. A fact 

 in favour of the secondary accommodating power is the gradual enlarge- 

 ment of the pupil after contraction, on emerging from a dark into a 

 light space. The duration of a vivid image on the nerves gradually 

 causes a less intense sensation to he transmitted, in consequence of this 

 accommodating power. In the experiments the rapid alternation of 

 coloured and black or non-acting images prevents the accommodating 

 power from acting to the extent it otherwise would, a reaction taking 

 place during the black or non-acting impression. The nerves thus 

 transmit the intermittent sensations as more intense than a continuous 

 sensation from the same coloured surface at rest. It must not be forgotten, 

 however, that the mental impression produced by the intermittent 

 flashes of colour in the experiments is continuous, the black or dark 

 intervals being unajipreciated. This latter effect is an example of what 

 is termed "persistence of vision ;" but if the increased brilliancy in the 

 writer's experiments is due to some such action at the retina as he has 

 supposed, it is obvious that the " persistence" cannot take place at the 

 same part, but must arise at some more advanced point in the track of 

 the visual sensation towards the sensorium, if not in the latter. The 

 supposition explaining the superior brilliancy of the colours in the 

 experiments over those at rest, also easily explains the still greater 

 vividness when the greater number of sets of colours are used. 



The beautiful gi-adation of the colours arises very simply, and will 

 be most easily understood by taking a coloured and a perforated disc, 

 and placing them in the various successive relative positions which they 

 assume in the experiment. Assuming there are two sets of three 

 colours on the top, the loose disc must have three apertures, and the 

 ratio of the velocities must be as three to two. It will be found that 

 dui'ing the entire period of rotation three apparently stationary radii 

 are exclusively coloured, each with one of the original colours, whilst 

 the intermediate radii are severally tinted with two of the colours, the 

 alternations of which vary in their relative proportions with the positions 

 of the radii. 



If with the same colour disc a two-aperture loose disc is used, the 

 entire series of graduated colours is not seen at once, but a uniform tint 

 occupies the entire disc at every instant, this tint gradually changing 

 through all the various tints of the spectrum previously described. If 

 we next take a two-aperture disc, with the sides of the apertures shaped 

 to volute curves instead of being radial, we have the colours thrown 

 into concentric circles, and flowing out from or in towards the centre, 

 according to the directions of the volute curves, the shading or grada- 

 tion between the colours being still shown. These last two experiments 



