374 Transactions of the Canadian Institute. [Vol. VII. 



the dominant hues of the sensation curves ; that is they are a particular 

 red, green, and blue-violet. Abney has chosen as the most suitable 

 position for these primary, fundamental, or reproduction colours about 

 /6700 in the red, /I5120 in the green, and /4600 in the blue; these 

 particular wave lengths, which are in ten millionths of a millimetre, 

 coincide very approximately with the wave lengths due to the red and 

 blue lithium and the green magnesium lines and are hence easily 

 located in the spectrum. 



Before discussing and illustrating the mixing of spectrum colours, it 

 will be advisable to clearly distinguish between two methods of colour 

 mixture commonly employed which will be frequently referred to in 

 this treatment of the subject. These are positive and negative mixture, 

 or mixture by addition and mixture by subtraction. The former is 

 effected by adding coloured light to coloured light, and the latter by 

 adding absorption to absorption ; while the resultant colours produced 

 in the two cases are, in general, essentially different. Perhaps the 

 commonest fallacy on the subject of colour mixture is the prevailing 

 belief that the mixture of )'ellow and blue gives green, but it can 

 easily be shown that the mixture of }'ellow and blue lights can not by 

 any possibility give green. If a piece of yellow glass be placed in one 

 lantern, forming a yellow patch upon the screen, and a piece of blue 

 glass in another lantern, forming a blue patch, the overlapping of these 

 patches and the consequent mixture of the coloured lights, as is at once 

 seen, [s/iown] gives us white light, which, although it may be yellowish 

 or bluish in hue, is without any approach to a greenish tinge ; by no 

 variation in the intensities of these colours, provided the hues remain 

 yellow and blue, can green light result. If, however, one lantern be 

 stopped and the yellow glass placed in front of the blue glass, a green 

 patch at once appears on the screen, [s/wu>u] showing that the colour 

 produced by superposing the absorptions, as is always done in the 

 mixture of paints or pigments, is essentially different to that resulting 

 from the mixture of coloured lights. To prevent confusion, it is very 

 necessary that this distinction be carefully borne in mind in the 

 subsequent treatment ; and it will at once be evident that positive 

 mixture is the only kind that can be employed in mixing spectrum 

 colours. 



The mixing of the three primary spectrum colours is effected by a 

 modification of Abney's well known colour patch apparatus. This 

 consists, first of all, in a means of forming a pure spectrum, and, 

 secondly, of an arrangement for combining any of the spectrum colours. 



