276 Dr. S. Bidwell. On Negative After-images, and 



white light, while the upper was not. Thus the spectrum and its 

 pulsative image could be seen together, the one above the other. At 

 first sight the pulsative image appeared to contain only two colours — 

 blue-green corresponding to the spectral red and orange, and purple- 

 pink corresponding to the green. Closer inspection revealed a pale 

 grey band between the blue-green and the purple, and a feeble tint of 

 lavender corresponding to the blue of the spectrum. Nothing at all 

 could be seen beneath the violet and the extreme red. The boundaries 

 of the several colours of the pulsative image were found to be roughly 

 as follows :— Blue green, X 6800 to X 6000 ; grey, X 6000 to X 5800 ; 

 purple, X 5800 to A. 5000; lavender, X 5000 to X 4300. 



Observations were also made of the changes undergone by the red 

 and green of the projected spectrum when the illumination was varied 

 by altering the width of the collimator-slit. With a width of 0*06 mm. 

 neither of the spectrum colours was at all affected; they appeared 

 simply as intermittent red and green. With - 125 mm. the green had 

 become transformed into a purple, intermixed with which a little green 

 could sometimes be glimpsed ; this latter completely disappeared when 

 the slit was made 0*2 mm. wide, the apparent colour being with this 

 and all greater widths of slit a steady purple. At the same stage 

 (0'2 mm.) red was still seen as red, though a nicker of blue-green could 

 be detected upon it. At 0*45 mm. red appeared as blue-green with a 

 red flicker, which ceased to be perceptible, except along the extreme 

 edge, when the width of the slit was increased to 0*5 mm. With a 

 slit of 0*94 mm. wide the last trace of red had vanished. Thus the 

 more ready exhaustion of the green sensation is again evidenced. 



VI. Colour Changes with Reversed Cycle. 



If the cycle is reversed by making the zinc disc turn in the opposite 

 direction, most of the spectrum colours undergo remarkable changes. 

 Red becomes rose-purple ; orange a diluted crimson ; yellow is made 

 much paler, as if veiled by a white haze; green appears as blue-green, 

 and blue-green as blue. Blue and violet are very slightly affected. 

 Very similar effects are observed when the disc described in Section I 

 is turned in the reverse direction. They naturally suggest that white 

 light excites a blue or blue-violet sensation, the persistence of which 

 exceeds that of any other fundamental sensation. 



VII. External and Border Phenomena. 



Some very remarkable and interesting phenomena are exhibited in 

 the region of the visual field immediately adjacent to that upon which 

 a " pulsative after-image " is being produced. It is a matter for sur- 

 prise that one should be able to perceive after-images without detecting 



