EXPERIMENTS ON DOUBLE VISION. 207 



or the perception resulting from it : practice however has 

 rendered it very familiar to me, as well as to several per- 

 sons, whom I have employed to repeat them. 



As the perception of the mixed colour in these experi- A1 ' colours do 

 ments results from the impression made by two objects of e o Ua i ease wl 

 different colours, and received simultaneously by each or- 

 gan of vision, it would seem, that all colours, being equal- 

 ly capable of producing such an impression, should occa- 

 sion a sensation equally complete and distinct, and produce 

 it with equal facility. This however is not the case : several 

 of them combine but imperfectly, or not at all. The com- 

 bination of blue and yellow for instance is not only pain- Blue and yel- 

 ful, on account of the continued attention it requires, but and'imperfect. 

 the colour resulting from them is vague, nearly indetermi- 

 nate, and of a disagreeable hue. This singular anomaly, 

 the most remarkable that occurred in these experiments, is 

 not sufficiently accounted for by the extreme difference and 

 heterogeneousness of blue and yellow ; since blue and red, 

 which are equally heterogeneous, combine easily and com- 

 pletely to produce a violet. The property of illuminating, This owine . tQ 

 which these colours possess in different degrees, confirmed a difference in 

 by Newton, and subsequently by Herschel, is the only iHumi n atui^° 

 circumstance, that appears to me capable of giving a plau- 

 sible explanation of it : for this property of illuminating 

 depends on the force with which the colours act on the eye. 

 Thus when two colours possessing this property in different 

 degrees act at once on the two eyes, the too powerful im- 

 pression on one necessarily renders that on the other less 

 sensible, and the mixed colour produced by this double im- 

 pression will not therefore be very distinct. This appears Hence blue 



to me the better founded, as the green is more distinct unites , best with 

 , , . , . , , a weak yellow, 



when the yellow is weaker, and the greenish tint produced 



by the combination of blue and yellow appears to contain the yellow has 



much more of the latter colour than of the former: and ™ ore * han i ts 



. ' ., , , , , due effect, 



further, if the impression of the yellow colour be weakened and thc begt 



by the interposition of a semitrans parent substance, the green is produc- 



green is rendered much more determinate. Both yellow ^yejj^ 11 "" 5 



and orange combine very difficultly with blue and violet, 



while they combine together, or with red, very easily. 



Homogeneous 



