222 Capt. W. de W. Abney. [June 21, 



The colour of the light of these two sources, can, as far as the eye can 

 distinguish, be very closely matched if not exactly by an orange 

 ray in the spectrum. Evidently then in such a case there can be na 

 complementary to this ray. We know as a matter of observation 

 that these lights do contain a certain quantity of rays of the higher 

 refrangibility, but so small proportionally to that found in daylight 

 that it is negligible. Again, we may take a light, such as the oxy- 

 hydrogen light, and may match its -whiteness by placing in the 

 electric light spectrum three slits, one in the red, another in the 

 green, and another in the blue, mixing the three rays, and altering 

 the apertures of the slits as required. The complementary of the red 

 foi 1 . the oxy hydrogen white light will be obtained by shutting off 

 the red ray, leaving the mixture of the green and blue rays. Keep- 

 ing the slits in the same position in the spectrum, a match may next 

 be made with the white light of the electric (arc) light. The comple- 

 mentary of the red may again be found as before, when it will be 

 found that the mixture of green and blue forming it will be bluer 

 than in the case of the oxyhydrogen light. 



From the above it will be seen that no complementary colours can 

 be definitely stated unless -the quality of the white light be known. 

 Sunlight and daylight being always yellower, at sea level, the lower 

 the altitude of the sun, it follows tliat any attempt to fix accurately 

 the wave-length of the complementary for daylight to any ray of the 

 spectrum must be exceedingly difficult. Wave-lengths of comple- 

 mentaries are given, however, in various text-books, but without any 

 statement as to the quality of the white light to which they refer. It 

 follows that if we do know the complementary colour, then the white 

 light which has to be added to it in order to match the contrast colour 

 must refer to the same white to which the complementary is referred, 



In the experiments which were undertaken as to the true colour 

 produced by contrast the light employed was that which I have 

 always used in colour experiments, viz., that emitted by the crater 

 on the positive pole of the electric light a light which is unchange- 

 able, and which can be relied upon as always being of the same 

 quality, the relative luminosities of the different parts of its spectrum 

 being fi^ed. Two complete sets of apparatus for producing colour 

 patches as described in the Addendum in " Colour Photometry, " 

 (' Phil. Trans.,' 1886) were provided each with its electric light. Each 

 colour patch was thrown on the whitened surface of a cube (No. II) 

 of l^inch side placed 12 inches apart from one another. With No, II 

 instrument the colour contrast was formed between white and a 

 diluted spectrum colour, the colour emerging through a slit placed in 

 the spectrum and forming a patch on the cube, and the white being 

 that reflected from the first surface of the prism, and re-reflected by 

 a silvered mirror, a magnified image being thrown, by means of a 



