stereoscope the figure will appear at one moment blue, then yellow, as though it had no permanent colour of its own; sometimes, again, the observer seems to see one colour through the other, and is distinctly conscious of the presence of both occupying apparently the same place, thus giving rise to the idea that the object might have at the same time two distinct colours. Meanwhile the little drawing assumes a highly lustrous appearance, as though it were made of polished glass; this is quite beautiful, and strikes with some astonishment those who see it for the first time. After some little practice has been gained, the blue and yellow colours will melt into a lustrous blue-grey or pure grey tint now and then for a few seconds, when again the

FIG. 64.-Slide for the Stereoscope, the Right and Left Hand being differently coloured.

contradictory and confusing phantoms just mentioned will make their appearance. Taken altogether, the effect is quite wonderful, and suggestive of something like a new sensation. There has been a good deal of controversy as to whether a true blending or mixture of the two colours actually does take place in the brain. The experiments of De Haldat and Dove, and afterward of Lubeck, Foucault, and Régnault, all point to this result. The results obtained by the author are also favorable to this view. But it must be confessed that the mixture obtained by this method differs in one respect from those previously described. For, when coloured light is mixed with the aid of rotating disks or by Lambert's method, we see only the resultant tint, the

two components disappearing entirely to give place to it. On the other hand, in this binocular mixture of colours, the presence of each of the original colours is all the while to some extent felt, and we are disposed to say that we see a neutral or grey hue which has evidently been made out of blue and yellow. Careful experiments by the author proved that the tint of the true mixture often differed from that obtained by the use of the stereoscope; colours which were pale, however, united more readily than intense ones, and gave less divergent results.* The binocular mixture of colours always produces more or less lustre ; it is not even necessary to employ distinct colours, the same effect being brought about by the mixture of a light and dark shade of the same colour, or simply by the binocular union of white and black, as was shown by Dove. The lustrous appearance of waves, ripples, and broken reflections in water is in each case mainly produced in this way, and hence, strictly speaking, can not be imitated by artists, who are necessarily obliged to present the same colours, the same light or dark shades, impartially to both eyes. It is for reasons similar to the above that a somewhat lustrous appearance is communicated to an oil painting by varnish, or to a watercolour drawing by glass; the eye sees the picture through the light slightly reflected from the glass or varnish, and is enabled apparently to penetrate beneath the mere surface of the pigment, and this slight illusion falls in with and helps the design of the artist.

* "American Journal of Science," May, 1865.

CHAPTER XI.

COMPLEMENTARY COLOURS.

In the previous chapter we found that the mixture of two masses of coloured light in some cases produced white light; this was, for example, true of mixtures of ultramarine-blue and yellow, or of red and greenish-blue. Any two colours which by their union produce white light are called complementary. An accurate knowledge of the nature and appearance of the complementary colours is important for artistic purposes, since these colours furnish the strongest possible contrasts. The best, in fact the only, method of becoming acquainted with the appearance of colours which are complementary is by actually studying them with the aid of suitable apparatus. The results thus obtained should be at the time registered, not in writing, but by imitating as far as possible the actual tints with brush and palette. By the aid of polarized light it is possible to produce with ease and certainty a large series of colours which are truly complementary. There are quite a number of instruments for accomplishing this, but perhaps the simplest and best is that which was contrived by Brücke for this express purpose, and called by him a schistoscope. (See Fig. 65.) This little apparatus is merely a combination of a low-power simple microscope with a polariscope, and can easily be constructed. Starting from below, P is a piece of white cardboard, which is fastened to the stand as indicated, and is consequently capable of being turned so as to reflect upward more or less white light, as may be required. N is

light thus reflected; it At A is a small square C is a crystal of calc

a Nicol's prism, which polarizes the is attached to a blackened stage, S. aperture two millimetres in size. spar; L is a convex lens of a focus such as to cause the two images of the square opening furnished by the calc spar just to touch each other. G and G are polished wedges of glass, the angles being 18°; for rough experiments they may be dispensed with. In order to use this apparatus,

FIG. 65.-The Schistoscope, for the production of Complementary Colours. (Brücke.)

the tube containing the calc spar is to be moved till distinct vision is obtained of the square opening in the stage by the eye placed at L, or rather of the two square openings which will be seen; the tube is then to be revolved till one of these images disappears entirely, and is to be left in this position. Besides the instrument it is necessary to provide a large number of thin slips of selenite or crystalized sulphate of lime. If a clear transparent piece of this substance is procured, it will be easy with a penknife to split off two or three hundred thin slips, and then with the aid of the instrument to select those which are worth pre

serving. To observe the colours it is only necessary to lay one of the slips on the stage between the calc-spar prism and the Nicol's prism, and then to turn the selenite till two brightly coloured squares are seen, as is indicated in Fig. 66. These two squares will always have colours which are

FIG. 66.-Complementary Colours as exhibited by the Schistoscope of Brücke.

complementary. The object of preparing a large number of the slips of selenite is the production of a large series of complementary tints. The thinner slips furnish colours that are more saturated; those which are thick give pale colours, or colours mixed with much white light. It will be found in this way that the following pairs of colours are complementary:

Red.
Orange..

Green-blue.*
Cyan-blue.

* Following Helmholtz, most writers give bluish-green as the complement to red. These observations of Helmholtz were made on the spectrum, the field being small and only a single eye employed. Extended observations with coloured disks, the hue of which can be studied in a more natural way and with both eyes simultaneously, have convinced the present writer that the complement of vermilion is a very green blue, and even the complement of carmine is a very green blue rather than a blue-green.