Intelligence and Miscellaneous Articles. 229 



determined sensation of colour. The results of the three series of 

 experiments are summed up in the following figure. 



3rd yellow-green. 



Green 



Blue-green 



Blue 



3rd blue 



Yellow. 



--Orange-yellow. 



Violet-blue. 



Violet. 



Red. 



Orange. 



It is there seen that the straight lines which join the comple- 

 mentary colours intersect in a single point. Consequently the 

 three separate colours form part of a triad ; here is their position 

 in the spectrum : — 



Orange at f from C to D. 



Green (3rd yellow-green) at | from D to E. 



Blue (3rd blue) at £ from E to G-. 



I repeated the last two series of experiments, methodically repla- 

 cing one of the three colours by another. I did the same with the 

 three colours which Maxwell considers fundamental. None of 

 these systems give regular results. Some colours which to the eye 

 are very close are in them separated by large spaces, while some 

 distinct colours are there brought very near together and almost 

 blended, and the lines joining two complementary colours do not 

 intersect in only one point. 



The two propositions above formulated are therefore in accord- 

 ance with experiment. 



To complete the demonstration of the existence of the triads, it 

 remains for me to prove that all the colours which occupy the vertices 

 of an inscribed equilateral triangle possess the same properties. 



I ascertained that it really is so by recommencing the same series 

 of experiments with the triad formed by the first violet, the blue- 

 green, and a yellow comprised between the first and second yellows 

 of my chromatic circle. These three colours are distant 120° in 

 the above figure. The results of the experiment are there indi- 

 cated. It will be remarked, not only that the complementary 

 colours are situated at the extremities of a straight line passing 

 through the point of intersection of the median lines, but also that 

 the new figure is exactly superposable to the first in what concerns 

 the position of the intervening colours. We may conclude from 

 this that the angular distance of the colours is the same, whichever may 

 be the triad employed as starting-point. 



Phil. Mag. S. 5. Vol. 12. No. 74. Sept. 1881. S 



