116 



♦ KNO\VLEDGE ♦ 



[Feb. 1, 1886. 



and that as long as we can see them at all, their distinc- 

 tive tint remains invariable. A very little reflec- 

 tion -svill suffice to show that this is not ihe 

 case. Take, as an illustration, a soldier's coat (let 

 us say an officer's, inasmuch as "Tommy Atkins's" 

 tunic is not always of the very brightest or best dyed 

 cloth). If we regard this in brilliant sunshine it will 

 look quite orange-coloured, as compared with its scarlet 

 tint as seen by the light of an ordinary white cloud ; 

 while, as evening draws on, the red gets darker and 

 darker, until it becomes almost indistinguishable. It is 

 noteworthy that the intense red of carmine is much less 

 afl'ected by darkness. Green, when in the full glare of 

 sunlight, becomes a yellowi.sh green, and even a kind of 

 pale yellow. A'iulet, in the same waj', becomes blue and 

 even blue-grey ; darkening, as in the case of red, as the 

 illumination is decreased ; but, unlike red, continuing to 

 show as violet as long as any light falls on it at all. All 

 this may be prettily shown by making a little bouquet of 

 .scarlet geraniums, pansy (" Purple King " answers well), 

 and any bright green leaves, and studying the seeming 

 change in the tints of its components by viewing them 

 fir.it in brilliant sunshine, then in ordinary daylight, 

 and finally in a darkened room. From all this we 

 shall gather that we have not only our three primary 

 colours in their normal state to deal with in forming 

 compound ones, but that we have an indefinitely large 

 series producible by mixing white light with them, as 

 also by altering their luminosity. This may be well 

 illustrated by using- our Maxwell's discs, employing a 

 plain white one to mix white light with the colour to bo 

 tested, and a dead-black one to degrade its luminosity. 

 Suppose, for ex.vmple, that we take a disc painted with 

 chrome yellow. If we combine this with a white one, 

 on rotating the combination rapidly on our whirling 

 table, it will be seen to exhibit a decidedly orange hire. 

 If now we exchange our white disc for a black one, we 

 shall find the resulting colour will be an olive-green. 

 This is very striking, and cannot fail to impress the 

 student with the large range of tints producible by 

 mere change in luminosity. He is strongly urged 

 to prepare a scries of discs as purely and brilliantly- 

 coloured as he can produce them, and try for himself 

 the results of their combinations with the simple 

 black and white ones. 



Now, it might reasonably be supposed that if rtuy 

 given colour changes so manifestly under varying 

 illumination, the difficult}' of judging of it would 

 be well-nigh insuperable. In practice, however, as 

 Helmholtz points out, we find that we are able to 

 judge of local colour without either uncertainty or 

 hesitation, imder the most diverse conditions. Thus, 

 white paper in full moonlight is darker than black satin 

 in daylight, but we never have the smallest difficulty in 

 recognising that the pr.per is really white and the satin 

 black ; though, at the same time, a brilliantly-illumiuTted 

 grej' surface mn,y appear identical in coloiir with a white 

 one in shade. By throwing a strong beam of light (by the 

 aid of a lens) on to a grey disc whose diameter is the 

 same as that of the beam, the disc will appear perfectly 

 white. This is most strikingly shown by mounting the 

 circle of grey paper on a sheet of white cardboard illumi- 

 nated by diffused light. A notable result of the diminu- 

 tion of the intensity of light is to practically reduce the 

 spectrum to the three colours red, green and violet. If 

 the light be still further diminished, the violet disappears, 

 the red turns to a kind of chocolate colour, and the green 

 is seen as a pale green. Finally if we continue to re- 

 duce tlie light passing through our jirism, the chocolate 



colour vanishes, and the green alone remains, itself to 

 turn to a sort of grey before the ultimate extinction of 

 the light altogether. We are indebted to Von Bezold 

 for this curious experiment. Here we get a glimpse of 

 the origin of such colours as brown, which, of course, 

 have no existence in the ordinary spectrum ; and we may 

 farther illustrate their production by attaching not a 

 whole Maxwell's disc to the axis of our whirling table, 

 but a simple card sector painted with vermilion, and 

 placing a sheet of dead-black paper behind it as a back- 

 ground, prior to setting the sector in rapid rotation. The 

 effect of this arrangement on the eye is that of a circle of 

 rich chocolate-brown colour. 



We have seen, then, that the juxtaposition of comple- 

 mentary colours sensibly heightens and intensifies them ; 

 and that any considerable increase or decrease in their 

 illumination also causes the most marked apparent 

 changes in their hues. We may now say a few words as 

 to the practical application of the principles we have been 

 endeavouring to explain. To the artist the knowledge of 

 these principles is, of course, simply indispensable. If, 

 then, we wish in, say, a water-colour sketch to make any 

 given colour "tell " to the utmost, we must contrive to 

 oppose its complementaiy colour to it. Let the reader 

 get a sheet of blue and a sheet of yellow paper, and out 

 of the latter cut two squares or discs. Now, let one of 

 these be laid upon the sheet whence it was cut, and the 

 other on the blue sheet, and regarded side by side. It 

 will not be easy at first for the observer to persuade him- 

 self that they are identical in colour, so much more 

 brilliant will the disc appear which is superposed 

 on the blue sheet. Or, again, he may perform 

 a modification of the experiment of Helmholtz, 

 described on p. 97, and providing himself with 

 a sheet of red paper, place upon it a disc of bluish 

 green, and regard the latter steadily for a minute 

 or so. If, now, an assistant will suddenly blow it 

 away, the space which it occupied will appear of a de- 

 cidedly more intense red than the rest of the sheet, 

 for a reason with which the student is by this time 

 fr.miliar. In Barnard's "Landscape Painting in Water 

 Colours" will be found a series of diagrams of a grey 

 pattern on variously-coloured backgrounds which are 

 very instructive. Or better still, we may go direct to 

 nature herself, and notice the glorious contrast of an 

 orange-red sunset with the vivid purple of distant hills 

 behind which the sun is descending; the distinctly green 

 tint of the moon as seen against one of the lovely rosy- 

 " afterglows " which have prevailed in various parts of 

 the world since the beginning of 1883, and so on. More- 

 over, as we have seen how much paler colours become 

 when very brilliantly illuminated, green, for example, 

 looking almost yellow in the glare of the sun, we 

 hence get a hint as to the most natural method of 

 painting any brilliantly-illuminated landscape, i e., to 

 impart a yellow tone to the bright lights. The careful 

 observer of nature, too, will note how the prevailing 

 colour of the source of illumination of a landscape affects 

 those of the shadows cast by the various objects in it. 

 And, in connection with the colours of shadows, by the 

 way, a very curious and instructive experiment may be 

 made without miich difficulty by any one who can 

 command the use of a room with a single window, closed 

 by a shutter in which a small aperture has been perforated. 

 Through such aperture we admit the daylight reflected 

 from a bright white cloud, and setting up a rod (a broom- 

 stick will do) in the middle of the room, in the path of 

 the rrtys ; its shadow must be made to fall upon the 

 opposite wall of the room, if white ; or, in default of this. 



