assistant (observer K), the result would have been a red mixed with less blue, consequently a colour much more like the red of the spectrum. From experiments of this kind Maxwell was able to construct the curves of intensity of the two fundamental colours which are perceived by those who are colour-blind to red; these curves are shown in Fig. 30. The letters A, B, C, D, etc., mark the positions of the fixed lines in the solar spectrum; the curved line marked GR exhibits the intensity of the green element, the line marked BL that of the blue or violet. It will be noticed that the green sensation attains its maximum about half way between the lines D and E, that is, in the yellowish green; while the highest point of the other curve is about half way between F and G, that is, in the blue space. Maxwell also constructed similar intensity curves for a normal eye; they are represented in Fig. 31, the

curve for red being indicated by a heavy line, the others as above. The green and blue curves have about the same disposition as with the colour-blind person, while the red attains its maximum between C and D, but nearer D—that is, in the red-orange space.

A set of observations was also made by Maxwell on the same colour-blind gentleman, with the aid of coloured disks in rapid rotation; and, from the colour equations thus obtained, the positions of the colours perceived by him were laid down in Newton's diagram, in a manner similar to that explained in the appendix to Chapter XIV. In Fig. 32, V shows the position assumed for red or vermilion; U, that of ultramarine-blue; and G, that of emerald

green. They are placed according to Maxwell's method, at the three angles of an equilateral triangle. W would be the position of white for a normal eye, and Y that of chrome-yellow. D is the position of the defective colour, which Maxwell was able to imitate by mingling, by the method of rotating disks, 86 parts of vermilion and 14 of ultramarine-blue. A line drawn from D through W contains along its length the various shades of grey and the white of the colour-blind. The grey which they perceive when green and

blue are mixed lies at w; the white of white paper, i. e., a more luminous grey, was on the same line but considerably farther along outside.

It may perhaps be as well to add to the above one or two remarks concerning the construction of Newton's diagram for the colour-blind. Let us suppose that the pure colours of the spectrum are employed, and that the missing colour is the fundamental red; we then place the fundamental green at G, Fig. 33, the fundamental blue or violet at U, and the missing red at D. Then along the line UG will lie mixtures of blue and green, and at w will be the white of the colour-blind person. Along the line DG will be situated various shades of green, from dark green to bright green, the latter colour predominating as we approach G. Along the line D U we shall have various shades of blue, from bright blue to dark blue, the colour being very dark near D and very bright near U. A line like the dotted one (Fig. 33) will contain various shades of green, from light green to dark green, but none of them so intense as

those situated along the line D G; in other words, they all will be mingled with what the colour-blind call white.

If the defective colour-sensation is supposed still red, but to be only partially absent, the diagram takes the form indicated in Fig. 34; that is, red, instead of occupying the position at one of the angles of an equilateral triangle, will be moved up toward the centre to R'. White will also be shifted from W to w, and the white of a person thus affected would appear to the normal eye of a somewhat greenish-blue hue. Between D and R' lie, so to speak, mixtures of red with darkness, and along the line R'G will be various mixtures of red and green, in which, according to a normal eye, the green element quite predominates; that is to say, their orange is more

like our yellow, their yellow like our greenish yellow, etc. Along the line R' U will be their mixtures of red and blue, or a series of purples, which will be more bluish than ours.

The condition of the normal eye by lamp-light is shown in Fig. 35. The blue or violet is moved from U, its position by daylight, up to u; white is moved from W to w-that is, into a region that would be called by daylight yellow. Yellow itself, Y, is not far from this new representative of white, and consequently by candlelight appears always whitish. In the purples, along the line Ru, the red element predominates; and in the mixtures of green and blue, along the line G u, the green constituent has the upper hand.

If we were colour-blind to every kind of light except red, then the colour diagram would assume a form similar to that shown in Fig. 36, D representing the darkest red perceptible to eyes so constituted. This sensation would be brought about by pure feeble red

light, or by a mixture of intense green and blue light, or by either of the latter. As we advance from D toward R, the red light gains in brightness, and out at w becomes very bright and stands for white. When a red glass is held before the eyes, something approximating to this kind of vision is produced.

FIG. 86.-Newton's Diagram for Persons Colour-blind to Green and Violet.

CHAPTER IX.

THE COLOUR THEORY OF YOUNG AND HELMHOLTZ.

It is well known to painters that approximate representations of all colours can be produced by the use of very few pigments. Three pigments or coloured powders will suffice, a red, yellow, and a blue; for example, crimsonlake, gamboge, and Prussian blue. The red and yellow mingled in various proportions will furnish different shades of orange and orange-yellow; the blue and yellow will give a great variety of greens; the red and blue all the purple and violet hues. There have been instances of painters in water-colours who used only these three pigments, adding lampblack for the purpose of darkening them and obtaining the browns and greys. Now, though it is not possible in this way to obtain as brilliant representatives of the hues of nature as with a less economical palette, yet substitutes of a more or less satisfactory character can actually be produced in this manner. These facts have been known to painters from the earliest ages, and furnished the foundation for the so-called theory of three primary colours, red, yellow, and blue. The most distinguished defender in modern times of this theory was Sir David Brewster, so justly celebrated for his many and brilliant optical discoveries. He maintained that there were three original or fundamental kinds of light, red, yellow, and blue, and that by their mixture in various proportions all other kinds of coloured light were produced, in the manner just described for pigments. Brewster in fact thought he had demonstrated