436 Mr. A. W. Porter and Dr. F. W. Edridge-Green. [May 28, 



being vertically over one another, the horizontal spectrum on the screen was 

 viewed through the Wratten and Wainwright red filter, so that only red 

 and orange were visible. The spectrum being at its greatest luminosity was 

 viewed for about 20 seconds, and then, at the same time as the filter was 

 removed and the eye was turned to its normal position, the Nicol was rotated 

 and the intensity of the spectrum much reduced. The blue band under these 

 conditions stretched right across the spectrum, whereas when the spectrum 

 was not reduced so much (though still much reduced) the orange and yellow 

 were predominant, and showed no change. 



13. The same process was tried by fatiguing the eye with light through a 

 Wratten and Wainwright blue filter (blue-green). There was no effect 

 unless the spectrum on the screen was very much enfeebled ; in that case a 

 practically black after-image, but if a stronger fatiguing source were used a 

 feeble red after-image, was seen. 



14. A Wratten and Wainwright green filter was now used in the same 

 way, and the screen spectrum was afterwards much reduced. A purple 

 band, having least effect on the brightest parts, yellow and orange, was seen 

 stretching across the spectrum. 



15. A yellow band was formed on screen of about A. 570, obtained by use 

 of a second slit placed in the focal plane where the spectrum was first 

 formed ; the light let through being then projected on screen. 



It was not quite pure, but contained a little yellow-green and orange 

 mixed with the yellow ; it, however, appeared as pure yellow on the screen. 

 When the eye was fatigued by light through the red filter, the after-image 

 on the yellow band was grey or black ; no tinge of colour was seen unless 

 fatiguing light was very bright (direct arc seen through red filter), when the 

 yellow image on screen became green. 



The eye was now fatigued by light from a separate spectroscope, and the 

 effects noted on the subsequent appearance of the after-image after the spectrum 

 on the screen had been viewed. The process adopted was as follows : — 



The eye was fatigued by light of a definite wave-length for 20 seconds 

 and the appearance of the after-image on the spectrum on the screen noted. 

 The eye was then moved so that the after-image was visible on a dark part 

 of the screen. 



16. Fatiguing light of the region of the D lines was first selected. After 

 the eye had been fatigued with this, on looking at the spectrum on the screen 

 a blue band appeared except on the brightest region (green, yellow, and 

 orange). On then looking at another portion of the screen the blue band 

 appeared continuous, i.e. without any change of tint or intensity, while the 

 after-image of the screen spectrum appeared above and below it. 



