Chapter I.
COLOR NAMES.

Writing from Samoa to Sidney Colvin in London, Stevenson1 says: “Perhaps inthe same way it might amuse you to send us any pattern of wall paperthat might strike you as cheap, pretty, and suitable for a room in a hotand extremely bright climate. It should be borne in mind that ourclimate can be extremely dark, too. Our sitting-room is to be invarnished wood. The room I have particularly in mind is a sort of bedand sitting room, pretty large, lit on three sides, and the colour infavour of its proprietor at present is a topazy yellow. But then withwhat colour to relieve it? For a little work-room of my own at the backI should rather like to see some patterns of unglossy—well, I’llbe hanged if I can describe this red. It’s not Turkish, and it’s notRoman, and it’s not Indian; but it seems to partake of the last two, andyet it can’t be either of them, because it ought to be able to go withvermilion. Ah, what a tangled web we weave! Anyway, with what brains youhave left choose me and send me some—many—patterns of theexact shade.”

(1)Where could be found a more delightful cry for some rational way todescribe color? He wants “a topazy yellow” and a red that is not Turkishnor Roman nor Indian, but that “seems to partake of the last two, andyet it can’t be either of them.” As a cap to the climax comes his demandfor “patterns of the exact shade.” Thus one of the clearest and mostforceful writers of10English finds himself unable to describe the color he wants. And why?Simply because popular language does not clearly state a single one ofthe three qualities united in every color, and which must be knownbefore one may even hope to convey his color conceptions to another.

(2)The incongruous and bizarre nature of our present color names mustappear to any thoughtful person. Baby blue, peacock blue, Nile green,apple green, lemon yellow, straw yellow, rose pink, heliotrope, royalpurple, Magenta, Solferino, plum, and automobile are popular terms,conveying different ideas to different persons and utterly failing todefine colors. The terms used for a single hue, such as pea green, seagreen, olive green, grass green, sage green, evergreen, invisible green,are not to be trusted in ordering a piece of cloth. They invite mistakesand disappointment. Not only are they inaccurate: they areinappropriate. Can we imagine musical tones called lark, canary,cockatoo, crow, cat, dog, or mouse, because they bear some distantresemblance to the cries of those animals? See paragraph131.

Color needs a system.

(3)Music is equipped with a system by which it defines each sound in termsof its pitch, intensify, and duration, without dragging in looseallusions to the endlessly varying sounds of nature. So should color besupplied with an appropriate system, based on the hue, value, andchroma2of our sensations, and not attempting to describe them by the indefiniteand varying colors of natural objects. The system now to be consideredportrays the three dimensions of color, and measures each by anappropriate scale. It does not rest upon the whim of an individual, butupon physical measurements made possible by special color11apparatus. The results may be tested by any one who comes to the problemwith “a clear mind, a good eye, and a fair supply of patience.”

Clear mental images make clear speech. Vague thoughts find vagueutterance.

(4)The child gathers flowers, hoards colored beads, chases butterflies, andbegs for the gaudiest painted toys. At first his strong color sensationsare sufficiently described by the simple terms of red, yellow, green,blue, and purple. But he soon sees that some are light, while others aredark, and later comes to perceive that each hue has many grayer degrees.Now, if he wants to describe a particular red,—such as that of hisfaded cap,—he is not content to merely call it red, since he isaware of other red objects which are very unlike it. So he gropes formeans to define this particular red; and, having no standard ofcomparison,—no scale by which to estimate,—he hesitatinglysays it is a “sort of dull red.”

(5)Thus early is he cramped by the poverty of color language. He has neverbeen given an appropriate word for this color quality, and has to borrowone signifying the opposite of sharp, which belongs to edge tools ratherthan to colors.

Most color terms are borrowed from other senses.

(6)When his older sister refers to the “tone” of her green dress, or speaksof the “key of color” in a picture, he is naturally confused, becausetone and key are terms associated in his mind with music. It may not belong before he will hear that “a color note has been pitched too high,”or that a certain artist “paints in a minor key.” All these terms leadto mixed and indefinite ideas, and leave him unequipped for the clearexpression of color qualities.

(7)Musical art is not so handicapped. It has an established12scale with measured intervals and definite terms. Likewise, coloristicart must establish a scale, measure its intervals, and name itsqualities in unmistakable fashion.

Color has three dimensions.

(8)It may sound strange to say that color has three dimensions, but it iseasily proved by the fact that each of them can be measured. Thus in thecase of the boy’s faded cap its redness orHUE3 is determined by one instrument; the amount of light inthe red, which is itsVALUE,3 is found by another instrument; while stilla third instrument determines the purity orCHROMA3 of thered.

The omission of any one of these three qualities leaves us in doubtas to the character of a color, just as truly as the character of thisstudio would remain undefined if the length were omitted and wedescribed it as 22 feet wide by 14 feet high. The imagination would befree to ascribe any length it chose, from 25 to 100 feet. This lengthmight be differently conceived by every individual who tried to supplythe missing factor.

(9)To illustrate the tri-dimensional nature of colors. Suppose we peel anorange and divide it in five parts, leaving the sections slightlyconnected below (Fig. 4). Then let us say that all the reds we haveever seen are gathered in one of the sections, all yellows in another,all greens in the third, blues in the fourth, and purples in the fifth.Next we will assort theseHUES in eachsection so that the lightest are near the top, and grade regularly tothe darkest near the bottom. A white wafer connects all thesections at the top, and a black wafer may be added beneath. SeePlate I.

(10)The fruit is then filled with assorted colors, graded from white toblack, according to theirVALUES, anddisposed by theirHUES in the fivesections. A slice near the top will uncover light values in allhues, and a slice near the bottom will find dark values in the samehues. A slice across the middle discloses a circuit of hues all ofMIDDLE VALUE; that is, midway betweenthe extremes of white and black.

(11)Two color dimensions are thus shown in the orange, and it remains toexhibit the third, which is calledCHROMA, or strength of color. To do this, we haveonly to take each section in turn, and, without disturbing the valuesalready assorted, shove the grayest in toward the narrow edge, and gradeoutward to the purest on the surface. Each slice across the fruit stillshows the circuit of hues in one uniform value; but the strong chromasare at the outside, while grayer and grayer chromas make a gradationinward to neutral gray at the centre, where all trace of colordisappears. The thin edges of all sections unite in a scale of gray fromblack to white, no matter what hue each contains.

The curved outside of each section shows its particular hue gradedfrom black to white; and, should the section be cut at right angles tothe thin edge, it would show the thirddimension,—chroma,—for the color is graded evenly from thesurface to neutral gray. A pin stuck in at any point traces thethird dimension.

A color sphere can be used to unite the three dimensions of hue,value, and chroma.

(12)Having used the familiar structure of the orange as a help inclassifying colors, let us substitute a geometric solid, like asphere,4and make use of geographical terms. The north pole is white. The southpole is black.14The equator is a circuit of middle reds, yellows, greens, blues, andpurples. Parallels above the equator describe this circuit in lightervalues, and parallels below trace it in darker values. The vertical axisjoining black and white is a neutral scale of gray values, whileperpendiculars to it (like a pin thrust into the orange) are scales ofchroma. Thus our color notions may be brought into an orderly relationby the color sphere. Any color describes its light and strength by itslocation in the solid or on the surface, and is named by its place inthe combined scales of hue, value, and chroma.

Two dimensions fail to describe a color.

(13)Much of the popular misunderstanding of color is caused by ignorance ofthese three dimensions or by an attempt to make two dimensions do thework of three.

(14)Flat diagrams showing hues and values, but omitting to define chromas,are as incomplete as would be a map of Switzerland with the mountainsleft out, or a harbor chart without indications of the depth of water.We find by aid of the measuring instruments that pigments are veryunequal in this third dimension,—chroma,—producing mountainsand valleys on the color sphere, so that, when the color system isworked out in pigments and charted, some colors must be traced well outbeyond the spherical surface (paragraphs125–127). Indeed, a COLOR TREE5 is needed to display by the unequal levels andlengths of its branches the individuality of pigment colors. But,whatever solid or figure is used to illustrate color relations, it mustcombine the three scales of hue, value, and chroma, and these definitescales furnish a name for every color based upon its intrinsicqualities, and free from terms purloined in other sensations, or caughtfrom the fluctuating colors of natural objects.

How this system describes the spectrum.

(15)The solar spectrum and rainbow are the most stimulating colorexperiences with which we are acquainted. Can they be described by thissolid system?

(16)The lightest part of the spectrum is a narrow field of greenish yellow,grading into darker red on one side and into darker green upon theother, followed by still darker blue and purple. Upon the sphere thevalues of these spectral colors trace a path high up on the yellowsection, near white, and slanting downward across the red and greensections, which are traversed near the level of the equator, it goes onto cross the blue and purple well down toward black.

(17)This forms an inclined circuit, crossing the equator at opposite points,and suggests the ecliptic or the rings of Saturn (see outside cover).A pale rainbow would describe a slanting circuit nearer white, anda dimmer one would fall within the sphere, while an intensely brilliantspectrum projects far beyond the surface of the sphere, so greatly isthe chroma of its hues in excess of the common pigments with which wework out our problems.

(18)At the outset it is well to recognize the place of the spectrum in thissystem, not only because it is the established basis of scientificstudy, but especially because the invariable order assumed by its huesis the only stable hint which Nature affords us in her infinite colorplay.

(19)All our color sensations are included in the color solid. None are leftout by its scales of hue, value, and chroma. Indeed, the imagination isled to conceive and locate still purer colors than any we now possess.Such increased degrees of color sensation can be named, and clearlyconveyed by symbols to another person as soon as the system iscomprehended.

Appendix to Chapter I.

Misnomers for Color.

The Century Dictionary helps an intelligent study of color by itsclear definitions and cross-references toHUE,VALUE, andCHROMA,—leaving no excuse forthose who would confuse these three qualities or treat a degree of anyquality as the quality itself.

Obscure statements were frequent in text-books before these newdefinitions appeared. Thus the term “shade” should be applied only todarkened values, and not to hues or chromas. Yet one writer says, “Thisyellow shades into green,” which is certainly a change of hue, and thenspeaks of “a brighter shade” in spite of his evident intention tosuggest a stronger chroma, which is neither a shade nor brighterluminosity.

Children gain wrong notions of “tint and shade” from the so-calledstandard colors shown to them, which present “tints” of red and bluemuch darker than the “shades” of yellow. This is bewildering, and, liketheir elders, they soon drop into the loose habit of calling any degreeof color-strength or color-light a “shade.”Value is a betterterm to describe the light which color reflects to the eye, and allcolor values, light or dark, are measured by thevalue-scale.

“Tone” is used in a confusing way to mean different things. Thus inthe same sentence we see it refers to a single touch of thebrush,—which is not a tone, but a paint spot,—and then we17read that the “tone of the canvas is golden.” This cannot mean that eachpaint spot is the color of gold, but is intended to suggest that thevarious objects depicted seem enveloped in a yellow atmosphere. Tone is,in fact, a musical term appropriate to sound, but out of place incolor. It seems better to call the brush touch acolor-spot: thenthe result of an harmonious relation between all the spots iscolor-envelope, or, as in Rood, “the chromatic composition.”

“Intensity” is a misleading term, if chroma be intended, for itdepends on the relative light of spectral hues. It is a degree ratherthan a quality, as appears in the expressions, intense heat, light,sound,—intensity of stimulus and reaction. Being a degree of manyqualities, it should not be used to describe the quality itself. Theword becomes especially unfit when used to describe two very differentphases of a color,—as its intense illumination, where the chromais greatly weakened, and the strongest chroma which is found in a muchlower value. “Purity” is also to be avoided in speaking of pigments, fornot one of our pigments represents a single pure ray of thespectrum.

Examples are constantly found of the mental blur caused by suchunfortunate terms, and, since misunderstanding becomes impossible withmeasured degrees of hue, value, and chroma, it seems only a question oftime when they will take the place of tint, tone, shade, purity andintensity.

Chapter II.
COLOR QUALITIES.

(20)The three color qualities are hue, value, and chroma.

HUE is the name of a color.

(21)Hue is the quality by which we distinguish one color from another, as ared from a yellow, a green, a blue, or a purple. This namesthe hue, but does not tell whether it is light or dark, weak orstrong,—leaving us in doubt as to its value and its chroma.

Science attributes this quality to difference in the LENGTH of etherwaves impinging on the retina, which causes the sensation of color. Thewave length M. 5269 gives a sensation of green, while M. 6867 gives asensation of red.6

VALUE is the light of a color.

(22)Value is the quality by which we distinguish a light color from a darkone. Color values are loosely called tints and shades, but the terms arefrequently misapplied. A tint should be a light value, and a shadeshould be darker; but the word “shade” has become a general term for anysort of color, so that a shade of yellow may prove to be lighter than atint of blue. A photometric7 scale of value places all colors inrelation to the extremes of white and black, but cannot describe theirhue or their chroma.

Science describes this quality as due to difference in theHEIGHT or amplitude of ether waves impingingon the retina. Small amplitudes of the wave lengths given in paragraph21 produce the sensation of dark green and darkred: larger amplitudes give the sensation of lighter green and lighterred.

CHROMA is the strength of a color.

(23)Chroma is the quality by which we distinguish a strong color from a weakone. To say that a rug is strong in color gives no hint of its hues orvalues, only its chromas. Loss of chroma is loosely called fading, butthis word is frequently used to include changes of value and hue. Taketwo autumn leaves, identical in color, and expose one to the weather,while the other is waxed and pressed in a book. Soon the exposed leaffades into a neutral gray, while the protected one preserves its strongchroma almost intact. If, in fading, the leaf does not change its hue orits value, there is only a loss of chroma, but the fading process ismore likely to induce some change of the other two qualities. Fading,however, cannot define these changes.

Science describes chroma as the purity of one wave length separatedfrom all others. Other wave lengths,INTERMINGLING, make its chroma less pure.A beam of daylight can combine all wave lengths in such balance asto give the sensation of whiteness, because no single wave is inexcess.8

(24)The color sphere (see Fig. 1) is a convenient model to illustrate thesethree qualities,—hue, value, and chroma,—and unite them bymeasured scales.

(25)The north pole of the color sphere is white, and the south pole black.Value or luminosity of colors ranges between these two extremes. This isthe vertical scale, to be memorized asV,20the initial for both value and vertical. Vertical movement through colormay thus be thought of as a change of value, but not as a change of hueor of chroma. Hues of color are spread around the equator of the sphere.This is a horizontal scale, memorized asH, the initial for bothhue and horizontal. Horizontal movement around the color solid is thusthought of as a change of hue, but not of value or of chroma.A line inward from the strong surface hues to the neutral grayaxis, traces the graying of each color, which is loss of chroma, andconversely a line beginning with neutral gray at the vertical axis, andbecoming more and more colored until it passes outside the sphere, is ascale of chroma, which is memorized asC, the initial both forchroma and centre. Thus the sphere lends its three dimensions to colordescription, and a color applied anywhere within, without, or on itssurface is located and named by its degree of hue, of value, and ofchroma.

HUES first appeal to the child, VALUES next, and CHROMAS last.

(26)Color education begins with ability to recognize and name certain hues,such as red, yellow, green, blue, and purple (see paragraphs182 and 183). Nature presents these hues in union withsuch varieties of value and chroma that, unless there be some standardof comparison, it is impossible for one person to describe themintelligently to another.

(27)The solar spectrum forms a basis for scientific color analysis, taughtin technical schools; but it is quite beyond the comprehension of achild. He needs something more tangible and constantly in view to trainhis color notions. He needs to handle colors, place them side by sidefor comparison, imitate them with21crayons, paints, and colored stuffs, so as to test the growth ofperception, and learn by simple yet accurate terms to describe each byits hue, its value, and its chroma.

(28)Pigments, rather than the solar spectrum, are the practical agents ofcolor work. Certain of them, selected and measured by this system (seeChapter V.), will be known asMIDDLE COLORS, because they stand midway in thescales of value and chroma. These middle colors are preserved inimperishable enamels,9 so that the child may handle and fix them in his memory,and thus gain a permanent basis for comparing all degrees of color. Helearns to grade each middle color to its extremes of value andchroma.

(29)Experiments with crayons and paints, and efforts to match middle colors,train his color sense to finer perceptions. Having learned to namecolors, he compares them with the enamels of middle value, and candescribe how light or dark they are. Later he perceives theirdifferences of strength, and, comparing them with the enamels of middlechroma, can describe how weak or strong they are. Thus the fullsignificance of these middle colors as a practical basis for all colorestimates becomes apparent; and, when at a more advanced stage hestudies the best examples of decorative color, he will again encounterthem in the most beautiful products of Oriental art.

Is it possible to define the endless varieties of color?

(30)At first glance it would seem almost hopeless to attempt the naming ofevery kind and degree of color. But, if all these varieties possess thesame three qualities, only in different degrees, and if each quality canbe measured by a scale, then there is a clue to this labyrinth.

A COLOR SPHERE and COLOR TREE to unite hue, value, and chroma.

(31)This clue is found in the union of these three qualities by measuredscales in acolor sphere and color tree.10 The equator of the sphere11 may bedivided into ten parts, and serve as the scale of hue, marked R,YR, Y, GY, G, BG, B, PB, P, and RP. Its vertical axis maybe divided into ten parts to serve as the scale of value, numbered fromblack (0) to white (10). Any perpendicular to the neutral axis is ascale of chroma. On the plane of the equator this scale is numbered 1,2, 3, 4, 5, from the centre to the surface.

(32)This chroma scale may be raised or lowered to any level of value, alwaysremaining perpendicular to the axis, and serving to measure the chromaof every hue at every level of value. The fact that some colors exceedothers to such an extent as to carry them out beyond the sphere isproved by measuring instruments,23but the fact is a new one to many persons. (Figs. 2 and 3.)

(33)For this reason theCOLOR TREE is acompleter model than the sphere, although the simplicity of the lattermakes it best for a child’s comprehension.

(34)The color tree is made by taking the vertical axis of the sphere, whichcarries a scale of value, for the trunk. The branches are at rightangles to the trunk; and, as in the sphere, they carry the scale ofchroma. Colored balls on the branches tell their Hue. In order to showtheMAXIMA of color, each branch isattached to the trunk (or neutral axis) at a level demanded by itsvalue,—the yellow nearest white at the top, then the green, red,blue, and purple branches, approaching black in the order of their lowervalues. It will be remembered that the chroma of the sphere ceased with5 at the equator. The color tree prolongs24this through 6, 7, 8, and 9. The branch ends carry colored balls,representing the most powerful red, yellow, green, blue, and purplepigments which we now possess, and could be lengthened, should strongerchromas be discovered.12

(35)Such models set up a permanent image of color relations. Every point isself-described by its place in the united scales of hue, value, andchroma. These scales fix each new perception of color in the child’smind by its situation in the color solid. The importance of such adefinite image can hardly be overestimated, for without it one colorsensation tends to efface another. When the child looks at a color, andhas no basis of comparison, it soon leaves a vague memory that cannot bedescribed. These models, on the contrary, lead to an intelligentestimate of each color in terms of its hue, its value, and its chroma;while the permanent enamels correct any personal bias by a definitestandard.

(36)Thus defined, a color falls into logical relation with all other colorsin the system, and is easily memorized, so that its image may berecalled at any distance of time or place by the notation.

(37)These solid models help to memorize and assemble colors and the memoryis further strengthened by a simpleNOTATION, which records each color so that it cannotbe mistaken for any other. By these written scales a child gains aninstinctive estimate of relations, so that, when he is delighted with anew color combination, its proportions are noted and understood.

(38)Musical art has long enjoyed the advantages of a definite scale andnotation. Should not the art of coloring gain by similar definition? Themusical scale is not left to personal25whim, nor does it change from day to day; and something as clear andstable would be an advantage in training the color sense.

(39)Perception of color is crude at first. The child sees only the mostobvious distinctions, and prefers the strongest stimulation. Butperception soon becomes refined by exercise, and, when a child tries toimitate the subtle colors of nature with paints, he begins to realizethat the strongest colors are not the most beautiful,—rather thetempered ones, which may be compared to the moderate sounds in music. Todescribe these tempered colors, he must estimate their hue, value, andchroma, and be able to describe in what degree his copy departs from thenatural color. And, with this gain in perception and imitation ofnatural color, he finds a strong desire to invent combinations to pleasehis fancy. Thus the study divides into three related attitudes, whichmay be called recognition, imitation, and invention. Recognition ofcolor is fundamental, but it would be tedious to spend a year or two informal and dry exercises to train recognition of color alone; for eachstep in recognition of color is best tested by exercise in its imitationand arrangement. When perception becomes keener, emphasis can be placedon imitation of the colors found in art and in nature, resting finallyon the selection and grouping of colors for design.13

Every color can be recognized, named, matched, imitated, and writtenby its HUE, VALUE, and CHROMA.

(40)The notation used in this system places Hue (expressed by an initial) atthe left; Value (expressed by a number) at the right and above a line;and Chroma (also expressed by26a number) at the right, below the line. Thus R⁵/₉means

Hue, value, and chroma unite in every color sensation, but the childcannot grasp them all at once.Hue-difference appeals to himfirst, and he gains a permanent idea of five principal hues from theenamels ofMIDDLE COLORS, learning toname, match, imitate, and finally write them by their initials: R (red),Y (yellow), G (green), B (blue), and P (purple). Intermediates formed byuniting successive pairs are also written by the joined initials, YR(yellow-red), GY (green-yellow), BG (blue-green), PB (purple-blue), andRP (red-purple).

(41)Ten differences of hue are as many as a child can render at the outset,yet in matching and imitating them he becomes aware of their light anddark quality, and learns to separate it from hue asvalue-difference. Middle colors, as implied by that name, standmidway between white and black,—that is, on the equator of thesphere,—so that a middle red will be written R⁵/,suggesting the steps 6, 7, 8, and 9 which are above the equator, whilesteps 4, 3, 2, and 1 are below. It is well to show only three values ofa color at first; for instance, the middle value contrasted with a lightand a dark one. These are written R³/, R⁵/,R⁷/. Soon he perceives and can imitate finer differences, andthe red scale may be written entire, as R¹/, R²/,R³/, R⁴/, R⁵/, R⁶/,R⁷/, R⁸/, R⁹/, with black as 0 andwhite as 10.

(42)Chroma-difference is the third and most subtle color quality. Thechild is already unconsciously familiar with the middle chroma of red,having had the enamels ofMIDDLE COLORalways27in view, and the red enamel is to be contrasted with the strongest andweakest red chromas obtainable. These he writes R /₁,R /5, R /₉, seeing that this describes the chromasof red, but leaves out its values. R⁵/₁,R⁵/₅, R⁵/₉, is the completestatement, showing that, while both hue and value are unchanged, thechroma passes from grayish red to middle red (enamel first learned) andout to the strongest red in the chroma scale obtained by vermilion.

(43)It may be long before he can imitate the intervening steps of chroma,many children finding it difficult to express more than five steps ofthe chroma scale, although easily making ten steps of value and fromtwenty to thirty-five steps of hue. This interesting feature is ofpsychologic value, and has been followed in the color tree and colorsphere.

Does such a scientific scheme leave any outlet for feeling andpersonal expression of beauty?

(44)Lest this exact attitude in color study should seem inartistic, comparedwith the free and almost chaotic methods in use, let it be said that thestage thus far outlined is frankly disciplinary. It is somewhat dry andunattractive, just as the early musical training is fatiguing withoutinventive exercises. The child should be encouraged at each step toexercise his fancy.

(45)Instead of cramping his outlook upon nature, it widens his grasp ofcolor, and stores the memory with finer differences, supplying morematerial by which to express his sense of coloristic beauty.

(46)Color harmony, as now treated, is a purely personal affair, difficult torefer to any clear principles or definite laws. The very terms by whichit seeks expression are borrowed from music, and suggest vague analogiesthat fail when put to the test. Color28needs a new set of expressive terms, appropriate to its qualities,before we can make an analysis as to the harmony or discord of our colorsensations.

(47)This need is supplied in the present system by measuredCHARTS, and aNOTATION. Their very construction preserves thebalance of colors, as will be shown in the next chapter, whilethe chapter on harmony (Chapter VII.) shows howharmonious pairs and triads of color may be found byMASKS with measured intervals. In fact, practice inthe use of the charts supplies the imagination with scales and sequencesof color quite as definite and quite as easily written as those soundintervals by which the musician conveys to others his sense of harmony.And, although in neither art can training alone make the artist, yet atechnical grasp of these formal scales gives acquaintance with the fullrange of the instrument, and is indispensable to artistic expression.From these color scales each individual is free to choose combinationsin accord with his feeling for color harmony.

Let us make an outline of the course of color study traced in thepreceding pages.15

PERCEPTION of color.

(48)Hue-difference.

Value-difference.

Chroma-difference.

EXPRESSION of color.

(49)Matching and imitation of hues (using stuffs, crayons, andpaints).

Matching and imitation of values and hues(using stuffs, crayons, and paints).

Matching and imitation of chromas, values,and hues (using stuffs, crayons, and paints).

Notation of color.

Initial for hue, numeral above for value, numeral below forchroma.

Sequences of color.

Balance of color.

HARMONY of color.

(50)Selection of colors that give pleasure.

Grouping of colors to suit some practical use: wall papers, rugs,book covers, etc.

A definite plan of color study, with freedom as to details ofpresentation.16

(51)Having memorized these broad divisions of the study, a cleverteacher will introduce many a detail, to meet the mood of the class, orcorrelate this subject with other studies, without for a moment losingthe thread of thought or befogging the presentation. But to range atrandom in the immense field of color sensations, without plan ordefinite aim in view, only courts fatigue of the retina and a chaoticstate of mind.

(52)The same broad principles which govern the presentation of other ideasapply with equal force in this study. A little, well apprehended,is better than a mass of undigested facts. If the child is led todiscover, or at least to think he is discovering, new things aboutcolor, the mind will be kept alert and seek out novel illustrations atevery step. Now and then a pupil will be found31who leads both teacher and class byintuitive appreciation ofcolor, and it is a subtle question how far such a nature can be helpedor hurt by formal exercises. But such an exception is rare, and goes toprove that systematic discipline of the color sense is necessary formost children.

(53)Outdoor nature and indoor surroundings offer endless colorillustrations. Birds, flowers, minerals, and the objects in daily usetake on a new interest when their varied colors are brought into aconscious relation, and clearly named. A tri-dimensionalperception, like this sense of color, requires skilful training, andeach lesson must be simplified to the last point practicable. It mustnot be too long, and should lead to some definite result which a childcan grasp and express with tolerable accuracy, while its difficultiesshould be approached by easy stages, so as to avoid failure ordiscouragement. The success of the present effort is the best incentiveto further achievement.

Appendix to Chapter II.
PLATE I.
THE COLOR SPHERE, with Measured Scales of
HUE, VALUE, and CHROMA.

The teacher of elementary grades introduces these scales of temperedcolor as fast as the child’s interest is awakened to their need by theexercises shown in Plates II. and III. Thus the Hue scale is learnedbefore the end of the second year, the Value scale during the next twoyears, and the Chroma scale in the fifth year. By the time a child isten years old these definite color scales have become part of his mentalfurnishing, so that he can name, write, and memorize any colorgroup.

1.The Color Sphere in Skeleton. This diagram shows the middlecolors on the equator, with strong red, yellow, green, blue, and purple,each at its proper level in the value scale, and projecting inaccordance with its scale of chroma. See the complete description ofthese scales inChapter II.

2.The Color Score. Fifteen typical steps taken from the colorsphere are here spread out in a flat field. TheFive Middle Colors form the centre level, with thesame hues in a lighter value above and in a darker value below.Chapter VI. describes the making of this Score, and itsuse in analyzing colors and preserving a written record of theirgroups.

3.The Value Scale and Chroma Scale. Each of the five colormaxima is thus shown at its proper level in the scale of light, andgraded by uniform steps from its strongest chroma inward to neutralityat the axis of the sphere. Pigment inequalities here become veryapparent.

FOR PLATES II. & III.,
SEE APPENDIX TO CHAPTER IV.,
CHILDREN’S COLOR STUDIES.

Chapter III.
COLOR MIXTURE AND BALANCE.

All colors grasped in the hand.

(54)Let us recall the names and order of colors given in the last chapter,with their assemblage in a sphere by the three qualities ofHUE,VALUE, andCHROMA. It will aid the memory to callthe thumb of the left handRED, theforefingerYELLOW, the middle fingerGREEN, the ring fingerBLUE, and the little fingerPURPLE (Fig. 6). When the finger tips are in acircle, they represent a circuit of hues, which has neither beginningnor end, for we can start with any finger and trace a sequence forwardor backward. Now close the tips together for white, and imagine that thefive strong hues have slipped down to the knuckles, where they stand forthe equator of the color Sphere. Still lower down at the wrist isblack.

(55)The hand thus becomes a color holder, with white at the finger tips,black at the wrist, strong colors around the outside, and weaker colorswithin the hollow. Each finger is a scale of its own color, with whiteabove and black below, while the graying of all the hues is traced byimaginary lines which meet in the middle of the hand. Thus a child’shand may be his substitute for the color sphere, and also make himrealize that it is filled with grayer degrees of the outside colors, allof which melt into gray in the centre.

Neighborly and opposite hues; and their mixture.

(56)Let this circle (Fig. 7) stand for the equator of the color sphere withthe five principal hues (written by their initials R, Y, G, B,and P) spaced evenly about it. Some colors are neighbors, as redand yellow, while others are opposites. As soon as a child experimentswith paints, he will notice the different results obtained by mixingthem.

First, the neighbors, that is, any pair which lie next one another,as red and yellow, will unite to make a hue which retains a suggestionof both. It isintermediate between red and yellow, and we callitYELLOW-RED.17

(57)Green and yellow unite to formGREEN-YELLOW, blue and green makeBLUE-GREEN, and so on with each succeeding pair.These intermediates are to be written by their initials, and inserted intheir proper place between the principal hues. It is as if an orange(paragraph 9) were split into ten sectorsinstead of five, with red, yellow, green, blue, and purple as alternatesectors, while half of each adjoining color pair were united to form thesector between them. The original order of five hues is in no wisedisturbed, but linked together by five intermediate steps.

(58)Here is a table of the intermediates made by mixing eachpair:—

Using the left hand again to hold colors, the principal hues remainunchanged on the knuckles, but in the hollows between them are placedintermediate hues, so that the circle now reads: red, yellow-red,yellow, green-yellow, green, blue-green, blue, purple-blue, purple, andred-purple, back to the red with which we started. This circuit iseasilymemorized, so that the child may begin with any colorpoint, and repeat the series clock wise (that is, from left to right) orin reverse order.

(59)Each principal hue has thus made two close neighbors by mixing with thenearest principal hue on either hand. The neighbors of red are ayellow-red on one side and a purple-red on the other. The neighbors ofgreen are a green-yellow on one hand and a blue-green on the other. Itis evident that a still closer neighbor could be made by again mixingeach consecutive pair in this circle of ten hues; and, if the processwere continued long enough, the color steps would become so fine thatthe eye could see only a circuit of hues melting imperceptibly one intoanother.

(60)But it is better for the child to gain a fixed idea of red, yellow,green, blue, and purple, with their intermediates, before attempting tomix pigments, and these ten steps are sufficient for primaryeducation.

(61)Next comes the question of opposites in this circle. A line drawnfrom red, through the centre, finds its opposite, blue-green.18 If thesecolors are mixed, they unite to form gray. Indeed, the centre of thecircle stands for a middle gray, not only because it is the centre ofthe neutral axis between black and white, but also because any pair ofopposites will unite to form gray.

(62)This is a table of five mixtures which make neutral gray:

(63)But if, instead of mixing these opposite hues, we place them side byside, the eye is so stimulated by their difference that each seems togain in strength;i.e., eachenhances the other whenseparate, butdestroys the other when mixed. This is a veryinteresting point to be more fully illustrated by the help of a colorwheel in Chapter V., paragraph106. What weneed to remember is that the mixture of neighborly hues makes them lessstimulating to the eye, because they resemble each other, while amixture of opposite hues extinguishes both in a neutral gray.

Hues once removed, and their mixture.

(64)There remains the question, What will happen if we mix, not twoneighbors, nor two opposites, buta pair of hues once removed in thecircle, such as red and green? A line joining this pair doesnot pass through the neutral centre, but to one side nearer yellow,which shows that this mixture falls between neutral gray and yellow,partaking somewhat of each. In the same way a line joining yellow andblue shows that their mixture contains both green and gray. Indeed,a line joining any two colors in the circuit may be said todescribe their union. A radius crossing this line passes to somehue on the circumference, and describes by its intersection with thefirst line38the chroma of the color made by a mixture of the two originalcolors.

Mixture of white and black: a scale of grays.

(65)So far we have thought only of the plane of the equator, with its circleof middle hues in ten steps, and studied their mixture by drawing linesto join them. Now let us start at the neutral centre, and think upwardto white and downward to black (Fig. 9.)

This vertical line is theneutral axis joining the poles ofwhite and black, which represent the opposites of light and darkness.Middle gray is half-way between. If black is called 0, and white is 10,then the middle point is 5, with 6, 7, 8, and 9 above, while 4, 3, 2,and 1 are below, thus making a vertical scale of grays from black towhite (Chapter II., paragraph 25).

If left to personal preference, an estimate of middle value will varywith each individual who attempts to make it. This appears in theneutral scales already published for schools, and students who dependupon them, discover a variation of over 10 per cent. in the selection ofmiddle gray. Since thisVALUE SCALEunderlies all color work, it needs accurate adjustment by scientificmeans, as in scales of sound, of length, of weight, or oftemperature.

APHOTOMETER (photo, light,andmeter, a measure)19 is shown39on the next page. It measures the relative amount of light which the eyereceives from any source, and so enables us to make a scale with anynumber of regular steps. The principle on which it acts is verysimple.

A rectangular box, divided by a central partition into halves, hassymmetrical openings in the front walls, which permit the light to reachtwo white fields placed upon the back walls. If one looks in through theobservation tube, both halves are seen to be exactly alike, and thewhite fields equally illuminated. A valve is then fitted to one ofthe front openings, so that the light in that half of the photometer maybe gradually diminished. Its white field is thus darkened by measureddegrees, and becomes black when all light is excluded by the closedvalve. While this darkening process goes on in one-half of theinstrument, the white field in the other half does not change, and,looking into the eyepiece, the observer sees each step contrasted withthe original white. One-half is thus said to bevariable becauseof its valve, and the other side is said to befixed. A dialconnected with the valve has a hand moving over it to show how muchlight is admitted to the field in the variable half.

Let us now test one of these personal decisions about middle value.A sample replaces the white field in the fixed half, and by meansof the valve, the white field in the variable half is alternatelydarkened and lightened, until it matches the sample and the eye sees nodifference in the two. The dial then discloses the fact that thissupposedlyMIDDLE VALUE reflects only42 per cent. of the light; that is to say, it is nearly a whole step toolow in a decimal scale. Other samples err nearly as far on the lightside of middle value, and further tests prove not only the varying colorsensitiveness of individuals, but detect a difference between the leftand right eye of the same person.

PHOTOMETER.

Back View.	Front View.

The vagaries of color estimate thus disclosed, lead some to seekshelter in “feeling and inspiration”; but feeling and inspiration aretemperamental, and have nothing to do with the simple facts of vision.A measured and unchanging scale is as necessary and valuable in thetraining of the eye as the musical scale in the discipline of theear.

It will soon be necessary to talk of the values in each color. We maydistinguish the values on the neutral axis from color values by writingthem N¹, N², N³, N⁴,N⁵, N⁶, N⁷, N⁸,N⁹, N¹⁰. Such a scale makes it easy to foresee theresult of mixing light values with dark ones. Any two gray values uniteto form a gray midway between them. Thus N⁴ and N⁶being equally above and below the centre, unite to form N⁵,as will also N⁷41and N³, N⁸ and N², or N⁹ andN¹. But N⁹ and N³ will unite to formN⁶, which is midway between 6 and 9.

Vertical Section through light openings.

PARTS.

(66)When this numbered scale of values is familiar, it serves not only todescribe light and dark grays, but the value of colors which are at thesame level in the scale. Thus R⁷ (popularly called a tint ofred) is neither lighter nor darker than the gray of N⁷.A numeral written above to the right always indicatesvalue,whether of a gray or a color, so that R¹, R²,R³, R⁴, R⁵, R⁶,R⁷, R⁸, R⁹, describes a regular scaleof red values from black to white, while G¹, G²,G³, etc., is a scale of green values.

(67)This matter of a notation for colors will be more fully worked out inChapter VI., but the letters and numerals alreadydescribed greatly simplify what we are about to consider in the mixtureand balance of colors.

Mixture of light hues with dark hues.

(68)Now that we are supplied with a decimal scale of grays, represented bydivisions of the neutral axis (N¹, N², etc.), anda corresponding decimal scale of value for each of the ten hues rangedabout the equator (R¹, R²,—YR¹,YR²,—Y¹,Y²,—GY¹, GY²,—and so on),traced by ten equidistant meridians from black to white, it is notdifficult to foresee what the mixture of any two colors will produce,whether they are of the same level of value, as in the colors of theequator already considered, or whether they are of different levels.

(69)For instance, let us mix a light yellow (Y⁷) with a dark red(R³). They are neighbors in hue, but well removed in value.A line joining them centres at YR⁵. This describes theresult of their mixture,—a value intermediate between 7 and 3,with a hue intermediate between R and Y. It is a yellow-red ofmiddle value, popularly called “dark orange.” But, while this term “darkorange” rarely means the same color to three different people, thesemeasured scales give to YR⁵ an unmistakable meaning, just asthe musical scale gives an unmistakable significance to the notes of itsscore.

(70)Evidently, this way of writing colors by their degrees of value and huegives clearness to what would otherwise be hard to express by the colorterms in common use.

(71)If Y⁹ and R⁵ be chosen for mixture, we know atonce that43they unite in YR⁷, which is two steps of the value scaleabove the middle; while Y⁶ and R² makeYR⁴, which is one step below the middle. Charts prepared withthis system show each of these colors and their mixture withexactness.

(72)The foregoing mixtures of dark reds and light yellows are typical of theunion of light and dark values of any neighboring hues, such as yellowand green, green and blue, blue and purple, or purple and red. Next letus think of the result of mixing different values in opposite hues; as,for instance, YR⁷ and B³ (Fig. 11). To thiscombination the color sphere gives a ready answer; for the middle of astraight line through the sphere, and joining them, coincides with theneutral centre, showing that theybalance in neutral gray. Thisis also true of any opposite pair of surface hues where the values areequally removed from the equator.

(73)Suppose we substitute familiar flowers for the notation, thenYR⁷ becomes the buttercup, and B³ is the wildviolet. But, in comparing the two, the eye is more stimulated by thebuttercup than by the violet, not alone because it is lighter, butbecause it is stronger in chroma; that is, farther away from the neutralaxis of the sphere, and in fact out beyond its surface, as shown inFig. 11.

The head of a pin stuck in toward the axis on the 7th level of YR mayrepresent the 9th step in the scale of chroma, such as the buttercup,while the “modest” violet with a chroma of only 4, is shown by itsposition to be nearer the neutral axis than the brilliant buttercup byfive steps of chroma. This is the third dimension of color, and must beincluded in our notation.44So we write the buttercup YR⁷/₉ and the violetB³/₄,—chroma always being written below tothe right of hue, and value always above.

(74)A line joining the head of the pin mentioned above withB³/₄ does not pass through the centre of thesphere, and its middle point is nearer the buttercup than the neutralaxis, showing that the hues of the buttercup and violetdo notbalance in gray.

The neutral centre is a balancing point for colors.

(75)This raises the question, What is balance of color? Artists criticisethe color schemes of paintings as being “too light or too dark”(unbalanced in value), “too weak or too strong” (unbalanced in chroma),and “too hot or too cold” (unbalanced in hue), showing that this is afundamental idea underlying all color arrangements.

(76)Let us assume that the centre of the sphere is the natural balancingpoint for all colors (which will be best shown by Maxwell discs inChapter V., paragraphs106–112), thencolor points equally removed from the centre must balance one another.Thus white balances black. Lighter red balances darker blue-green.Middle red balances middle blue-green. In short, every straight linethrough this centre indicates opposite qualities that balance oneanother. The color points so found are said to be“complementary,” for each supplies what is needed to complementor balance the other in hue, value, and chroma.

(77)The true complement of the buttercup, then, is not the violet, which istoo weak in chroma to balance its strong opposite. We have no blueflower that can equal the chroma of the buttercup. Some other means mustbe found to produce a balance. One way is to use more of the weakercolor. Thus we can make45a bunch of buttercups and violets, using twice as many of the latter, sothat the eye sees anarea of blue twice as great as thearea of yellow-red. Area as a compensation for inequalities ofhue, value, and chroma will be further described under the harmony ofcolor inChapter VII.

(78)But, before leaving this illustration of the buttercup and violet, it iswell to consider another color path connecting them which does not passthrough the sphere,but around it (Fig. 12). Such a path swingingaround from yellow-red to blue slants downward in value, and passesthrough yellow, green-yellow, green, and blue-green, tracing asequence of hue, of which each step is less chromatic than itspredecessor.

This diminishing sequence is easily writtenthus,—YR⁸/₉, Y⁷/₈,GY⁶/₇, G⁵/₆,BG⁴/₅, B³/₄,—and isshown graphically in Fig. 12. Its hue sequence is described by theinitials YR, Y, GY, G, BG, and B. Its value-sequenceappears in the upper numerals, 8, 7, 6, 5, 4, and 3, while thechroma-sequence is included in the lower numerals, 9, 8, 7, 6, 5,and 4. This gives a complete statement of the sequence, definingits peculiarity, that at each change of hue there is a regular decreaseof value and chroma. Nature seems to be partial to this sequence,constantly reiterating it in yellow flowers with their darker greenleaves and underlying shadows. In spring time she may contract itsrange, making the blue more green and the yellow less red, but in autumnshe seems to widen the range, presenting strong contrasts of yellow-redand purple-blue.

(79)Every day she plays upon the values of this sequence,46from the strong contrasts of light and shadow at noon to the hardlyperceptible differences at twilight. The chroma of this sequence expandsduring the summer to strong colors, and contracts in winter to grays.Indeed, Nature, who would seem to be the source of our notions of colorharmony, rarely repeats herself, yet is endlessly balancing inequalitiesof hue, value, and chroma by compensations of quantity.

(80)So subtle is this equilibrium that it is taken for granted andforgotten, except when some violent disturbance disarranges it, such asan earthquake or a thunder-storm.

The triple nature of color balance illustrated.

(81)The simplest idea of balance is the equilibrium of two halves of a sticksupported at its middle point. If one end is heavier than the other, thesupport must be moved nearer to that end.

But, since color unites three qualities, we must seek some type oftriple balance. The game of jackstraws illustrates this, when thedisturbance of one piece involves the displacement of two others. Theaction of three children on a floating plank or the equilibrium of twoacrobats carried on the shoulders of a third may also serve asexamples.

(82)Triple balance may be graphically shown by three discs in contact. Twoof them are suspended by their centres, while they remain in touch witha third supported on a pivot, as in Fig. 14. Let us call the lowest discHue (H), and the lateral discs Value (V) and Chroma (C). Any dip orrotation of the lower disc H will induce sympathetic action in the twolateral discs V and C. When H is inclined, both V47and C change their relations to it. If H is raised vertically, both Vand C dip outward. If H is rotated, both V and C rotate, but in oppositedirections. Indeed, any disturbance of V affects H and C, while H and Vrespond to any movement of C. So we must be prepared to realizethat any change of one color quality involves readjustment of the othertwo.

(83)Color balance soon leads to a study of optics in one direction, toæsthetics in another, and to mathematical proportions in a third, andany attempt at an easy solution of its problems is not likely tosucceed. It is a very complicated question, whose closest counterpart isto be sought in musical rhythms. The fall of musical impulses upon theear can make us gay or sad, and there are color groups which, actingthrough the eye, can convey pleasure or pain to the mind.

(84)A colorist is keenly alive to these feelings of satisfaction orannoyance, and consciously or unconsciously he rejects certaincombinations of color and accepts others. Successful pictures anddecorative schemes are due to some sort of balance uniting “light andshade” (value), “warmth and coolness” (hue), with “brilliancy andgrayness” (chroma); for, when they fail to please, the mind at oncebegins to search for the unbalanced quality, and complains that thecolor is “too hot,” “too dark,” or “too crude.” This effort to establishpleasing proportions may be unconscious in one temperament, while itbecomes a matter of definite analysis in another. Emerson claimed thatthe unconscious only is complete. We gladly permit those whose colorinstinct is unerring—(and how few they are!)—to neglect allrules and set formulas. But education is concerned with the many whohave not this gift.

(85)Any real progress in color education must come not from a blindimitation of past successes, but by a study into the laws48which they exemplify. To exactly copy fine Japanese prints or Persianrugs or Renaissance tapestries, while it cultivates an appreciation oftheir refinements, does not give one the power to create things equallybeautiful. The masterpieces of music correctly rendered do not ofnecessity make a composer. The musician, besides the study ofmasterpieces, absorbs the science of counterpoint, and records by anunmistakable notation the exact character of any new combination ofmusical intervals which he conceives.

(86)So must the art of the colorist be furnished with a scientific basis anda clear form of color notation. This will record the successes andfailures of the past, and aid in a search, by contrast and analysis, forthe fundamentals of color balance. Without a measured and systematicnotation, attempts to describe color harmony only produce hazygeneralities of little value in describing our sensations, and fail toexpress the essential differences between “good” and “bad” color.

Appendix to Chapter III.

False Color Balance. There is awidely accepted error that red, yellow, and blue are “primary,” althoughBrewster’s theory was long ago dropped when the elements of color visionproved to beRED,GREEN, andVIOLET-BLUE. The late Professor Rood calledattention to this in Chapters VIII.–XI. of his book, “ModernChromatics,” which appeared in 1879. Yet we find it very generallytaught in school. Nor does the harm end there, for placing red, yellow,and blue equidistant in a circle, with orange, green, and purple asintermediates, the teacher goes on to state that opposite hues arecomplementary.

Unfortunately, each of these statements is wrong, and, if tested bythe mixture of colored lights or with Maxwell’s rotating discs, theirfalsity is evident.

There can be no doubt that green is not the complement of red, norpurple of yellow, nor orange of blue, for neither one of these pairsunites as it should in a balanced neutrality, and a total test of thecircle gives great excess of orange, showing that red50and yellow usurp too great a portion of the circumference. Starting froma false basis, the Brewster theory can only lead to unbalanced andinharmonious effects of color.

The fundamental color sensations areRED,GREEN, andVIOLET-BLUE.

all of the hues in the right-hand column being compound sensations.The sensation of green is not due to a mixture of yellow and blue, asthe absorptive action of pigments might lead one to think:GREEN IS FUNDAMENTAL, and not made by mixing anyhues of the spectrum, whileYELLOW IS NOTFUNDAMENTAL, but caused by the mingled sensations of red andgreen. This is easily proved by a controlled spectrum, for allyellow-reds, yellows, and green-yellows can be matched by certainproportions of red and green light, all blue-greens, blues, andpurple-blues can be obtained by the union of green and violet light,while purple-blue, purple, and red-purple result from the union ofviolet and red light. But there is no point where a mixture gives red,green, or violet-blue. They are the true primaries, whose mixturesproduce all other hues.

Studio and school-room practice still cling to the discreditedtheory, claiming that, if it fails to describe our color sensations, yetit may be called practically true of pigments, because a red, yellow,and blue pigment suffice to imitate most natural colors. Thisdiscrepancy between pigment mixture and retinal mixture becomes clear assoon as one learns the physical make-up and behavior of paints.

Spectral analysis shows that no pigment is a pure example of thedominant hue which it sends to the eye. Take, for example, the verychromatic pigments representing red and green, such as vermilion andemerald green. If each emitted a single pure hue free from trace of anyother hue, then their mixture would appear yellow, as when spectral redand green unite. But, instead of yellow, their mixture produces a warmgray, called brown or “dull salmon,” and this is to be inferred fromtheir spectra, where it is seen that vermilion emits some green andpurple as well as its dominant color, while the green also sends someblue and red light to the eye.20

Thus stray hues from other parts of the spectrum tend to neutralizethe yellow sensation, which would be strong if each of the pigments werepure in the spectral sense. Pigment absorption affects all palettemixtures, and, failing to obtain a satisfactory yellow by mixture of redand green, painters use original yellow pigments,—such asaureolin, cadmium, and lead chromate,—each of them also impure butgiving a dominant sensation of yellow. Did the eye discriminate, as doesthe ear when it analyzes the separate tones of a chord, then we shouldrecognize that yellow pigments emit both red and green rays.

White light dispersed into a colored band by one prism, may have theprocess reversed by a second prism, so that the eye sees again onlywhite light. But this would not be so, did not the balance of red,green, and violet-blue sensations remain undisturbed. All our ideas ofcolor harmony are based upon this fundamental relation, and, if pigmentsare to render harmonious effects,52we must learn to control their impurities so as to preserve a balance ofred, green, and violet-blue.

Otherwise, the excessive chroma and value of red and yellow pigmentsso overwhelm the lesser degrees of green and blue pigments that nobalance is possible, and the colorist of fine perception must reject notalone the theoretical, but also the practical outcome of a“red-yellow-blue” theory.

Some of the points raised in this discussion are rather subtle forstudents, and may well be left until they arise in a study of optics,but the teacher should grasp them clearly, so as not to be led intofalse statements about primary and complementary hues.

Chapter IV.
PRISMATIC COLOR.

Pure color is seen in the spectrum of sunlight.

(87)The strongest sensation of color is gained in a darkened room, with aprism used to split a beam of sunlight into its various wave lengths.Through a narrow slit there enters a straight pencil of light which weare accustomed to think of aswhite, although it is a bundle ofvariously colored rays (or waves of ether) whose union and balance is soperfect that no single ray predominates.

(88)Cover the narrow slit, and we are plunged in darkness. Admit the beam,and the eye feels a powerful contrast between the spot of light on thefloor and its surrounding darkness. Place a triangular glass prism nearthe slit to intercept the beam of white light, and suddenly thereappears on the opposite wall a band of brilliant colors. This delightfulexperiment rivets the eye by the beauty and purity of its hues. Allother colors seem weak by comparison.

Their weakness is due to impurity, for all pigments and dyes reflectportions of hues other than their dominant one, which tend to “gray” anddiminish their chroma.

(89)But prismatic color is pure, or very nearly so, because the shape of theglass refracts each hue, and separates it by the length54of its ether wave. These waves have been measured, and science can nameeach hue by its wave length. Thus a certain red is known as M. 6867, anda certain green sensation is M. 5269.21 Without attempting any scientificanalysis of color, let it be said that Sir Isaac Newton made his seriesof experiments in 1687, and was privileged to name this color sequenceby seven steps which he called red, orange, yellow, green, blue, violet,and indigo. Later a scientist named Fraunhofer discovered fine blacklines crossing the solar spectrum, and marked them with letters of thealphabet from a to h. These with the wave length serve to locateevery hue and define every step in the sequence. Since Newton’s time ithas been proved that only three of the spectral hues areprimary;viz., a red, a green, and a violet-blue, while their mixtureproduces all other gradations. By receiving the spectrum on an opaquescreen with fine slits that fit the red and green waves, so that theyalone pass through, these two primary hues can be received on mirrorsinclined at such an angle as to unite on another screen, where, insteadof red or green, the eye sees only yellow.22

(90)A similar arrangement of slits and mirrors for the green and violet-blueproves that they unite to make blue, while a third experiment shows thatthe red and violet-blue can unite to make purple. So yellow, blue-green,and purple are called secondary55hues because they result from the mixture of the three primaries, red,green, and violet-blue.

In comparing these two color lists, we see that the “indigo” and“orange” of Sir Isaac Newton have been discarded. Both are indefinite,and refer to variable products of the vegetable kingdom. Violet is alsoborrowed from the same kingdom; and, in order to describe a violet, wesay it is a purple violet or blue violet, as the case may be, just as wedescribe an orange as a red orange or a yellow orange. Their colordifference is not expressed by the terms “orange” or “violet,” but bythe words “red,” “yellow,” “blue,” or “purple,” all of which are truecolor names and arouse an unmixed color image.

(91)In the nursery a child learns to use the simple color names red, yellow,green, blue, and purple. When familiarity with the color sphere makeshim relate them to each other and place them between black and white bytheir degree of light and strength, there will be no occasion to revertto vegetables, animals, minerals, or the ever-varying hues of sea andsky to express his color sensations.

(92)Another experiment accentuates the difference between spectral andpigment color. When the spectrum is spread on the screen by the use of aprism, and a second prism is placed inverted beyond the first, itregathers the dispersed rays back into their original beam, making awhite spot on the floor. This proves that all the colored rays of lightcombine to balance each other in whiteness. But if pigments which arethe closest possible imitation of these hues are united on a painter’spalette, either by the brush or the knife, theymake gray, and notwhite.

(93)This is another illustration of the behavior of pigments, for, insteadof uniting to form white, they form gray, which is a56darkened or impure form of white; and, lest this should be attributed toa chemical reaction between the various matters that serve as pigments,the experiment can be carried out without allowing one pigment to touchanother by using Maxwell discs, as will be shown in the nextchapter.

(94)Before leaving these prismatic colors, let us study them in the light ofwhat has already been learned of color dimensions. Not only do theypresent different values, but also different chromas. Their values rangefrom darkness at each end, where red and purple become visible, to abrightness in the greenish yellow, which is almost white. So on thecolor tree described in Chapter II., paragraph34, yellow has the highest branch, green is lower, red isbelow the middle, with blue and purple lower down, near black.

(95)Then in chroma they range from the powerful stimulation of the red tothe soothing purple, with green occupying an intermediate step. This isalso given on the color tree by the length of its branches.

(96)In Fig. 15 the vertical curve describes the values of the spectrum asthey grade from red through yellow, green, blue, and purple. Thehorizontal curve describes the chromas of the spectrum in the samesequence; while the third curve leaning outward is obtained by unitingthe first two by two planes at right angles to one another, and sums upthe three qualities by a single descriptive line. Now the red and purpleends are far apart, and science forbids their junction because of theirgreat difference in wave length. But the mind is prone to unite them inorder to produce the red-purples which we see in clouds at sunset, inflowers and57grapes and the amethyst. Indeed, it has been done unhesitatingly in mostcolor schemes in order to supply the opposite of green.

(97)This gives a slanting circuit joining all the branch ends of the colortree, and has been likened to the rings of Saturn in Chapter I.,paragraph 17.

A prismatic color sphere.

(98)With a little effort of the imagination we can picture a prismatic colorsphere, using only the colors of light. In a cylindrical chamber is hunga diaphanous ball similar to a huge soap bubble, which can display coloron its surface without obscuring its interior. Then, at the properpoints of the surrounding wall, three pure beams of colored light areadmitted,—one red, another green, and the third violet-blue.

(99)They fall at proper levels on three sides of the sphere, while theirintermediate gradations encircle the sphere with a complete spectrumplus the needed purple. As they penetrate the sphere, they unite tobalance each other in neutrality. Pure whiteness is at the top, and, bysome imaginary means their light gradually diminishes until theydisappear in darkness below.

(100)This ideal color system is impossible in the present state of ourknowledge and implements. Even were it possible, its immaterial huescould not serve to dye materials or paint pictures. Pigments are, andwill in all probability continue to be, the practical agents ofcoloristic productions, however reluctant the scientist may be to acceptthem as the basis of a color system. It is true that they are chemicallyimpure and imperfectly represent the colors of light. Some of them faderapidly and undergo chemical change, as in the notable case of a greenpigment tested58by this measured system, which in a few weeks lost four steps of chroma,gained two steps of value, and swung into a bluer hue.

(101)But the color sphere to be next described is worked out with a fewreliable pigments, mostly natural earths, whose fading is a matter ofyears and so slight as to be almost imperceptible. Besides, itsprincipal hues are preserved in safe keeping by imperishable enamels,which can be used to correct any tendency of the pigments to distort themeasured intervals of the color sphere.

This meets the most serious objection to a pigment system. Without ita child has nothing tangible which he can keep in constant view toimitate and memorize. With it he builds up a mental image of measuredrelations that describe every color in nature, including the fleetinghues of the rainbow, although they appear but for a moment at rareintervals. Finally, it furnishes a simple notation which records everycolor sensation by a letter and two numerals. With the enlargement ofhis mental power he will unite these in a comprehensive grasp of thelarger relations of color.

Appendix to Chapter IV.

Children’s Color Studies.

These reproductions of children’s work are given as proof that colorcharm and good taste may be cultivated from the start.

Five Middle Hues are first taught bythe use of special crayons, and later with water colors. They representthe equator of the color sphere (seePlateI.),—a circle midway between the extremes of color-light andcolor-strength,—and are known asMiddleRed,Middle Yellow,Middle Green,MiddleBlue, andMiddle Purple.

These are starting-points for training the eye to measure regularscales of Value and Chroma.23 Only with such a trained judgment is it safe toundertake the use of strong colors.24

Beginners should avoid Strong Color. Extreme red, yellow, andblue are discordant. (They “shriek” and “swear.” Mark Twain callsRoxana’s gown “a volcanic eruption of infernal splendors.”) Yet thereare some who claim that the child craves them, and must have them toproduce a thrill. So also does he crave candies, matches, and thecarving-knife. He covets the trumpet, fire-gong, and bass-drum for their“thrill”; but who would think them necessary60to the musical training of the ear? Like the blazing bill-board and thecircus wagon, they may be suffered out-of-doors; but such boisteroussounds and color sprees are unfit for the school-room.

Quiet Color is the Mark of Good Taste. Refinement in dress andthe furnishings of the home is attractive, but we shrink from those whoare “loud” in their speech or their clothing. If we wish our children tobecome well-bred, is it logical to begin by encouraging barbaroustastes? Their young minds are very open to suggestion. They quicklyadopt our standards, and the blame must fall upon us if they acquirecrude color habits. Yellow journalism and rag-time tunes will not helptheir taste in speech or song, nor will violent hues improve their tastein matters of color.

Balance of Color is to be sought. Artists and decorators arewell aware of a fact that slowly dawns upon the student; namely, thatcolor harmony is due to the preservation of a subtle balance andimpossible by the use of extremes. This balance of color resides morewithin the spherical surface of this system than in the excessivechromas which project beyond. It is futile to encourage children inefforts to rival the poppy or buttercup, even with the strongestpigments obtainable. Their sunlit points give pleasure because they aresurrounded and balanced by blue ether and wide green fields. Were theseconditions reversed, so that the flowers appeared as little spots ofblue or green in great fields of blazing red, orange, and yellow, ourpained eyes would be shut in disgust.

The painter knows that pigmentscannot rival the brilliancy ofthe buttercup and poppy, enhanced by their surroundings. What is more,he does not care to attempt it. Nor does the musician wish to imitatethe screech of a siren or the explosion of a gun. These are not subjectsfor art. Harmonious sounds are the study of the musician, and tunedcolors are the materials of the colorist. Corot61in landscape, and Titian, Velasquez, and Whistler in figure painting,show us that Nature’s richest effects and most beautiful color areenveloped in an atmosphere of gray.

Beauty of Color lies in Tempered Relations. Music rarelytouches the extreme range of sound, and harmonious color rarely uses theextremes of color-light or color-strength. Regular scales in the middleregister are first given to train the ear, and so should the eye befirst familiarized with medium degrees of color.

This system provides measured scales, established by specialinstruments, and is able to select the middle points of red, yellow,green, blue, and purple as a basis for comparing and relating allcolors. These five middle colors form a Chromatic Tuning Fork. (See page70.) It is far better that children should first become familiar withthese tuned color intervals which are harmonious in themselves ratherthan begin by blundering among unrelated degrees of harsh and violentcolor. Who would think of teaching the musical scale with a piano out oftune?

The Tuning of Color cannot be left to Personal Whim. The widediscrepancies of red, yellow, and blue, which have been falsely taughtas primary colors, can no more be tuned by a child than the musicalnovice can tune his instrument. Each of these hues has three variablefactors (see page 14, paragraph14), andscientific tests are necessary to measure and relate their unevendegrees of Hue, Value, and Chroma.

Visual estimates of color, without the help of any standard forcomparison, are continually distorted by doubt, guess-work, and thefatigue of the eye. Hardly two persons can agree in the intelligibledescription of color. Not only do individuals differ, but the same eyewill vary in its estimates from day to day. A frequent assumptionthat all strong pigments are equal in chroma, is62far from the truth, and involves beginners in many mishaps. Thus thestrongest blue-green, chromium sesquioxide, is but half the chroma ofits red complement, the sulphuret of mercury. Yet ignorance isconstantly leading to their unbalanced use. Indeed, some are stillunaware that they are the complements of each other.25

It is evident that the fundamental scales of Hue, Value, and Chromamust be established by scientific measures, not by personal bias. Thissystem is unique in the possession of such scales, made possible by thedevising of special instruments for the measurement of color, and cantherefore be trusted as a permanent basis for training the colorsense.

The examples in Plates II. and III. show how successfully the tunedcrayons, cards, and water colors of this system lead a child to fineappreciations of color harmony.

PLATE II.
Color Studies with TUNED CRAYONS
in the Lower Grades.

Children have made every example on this plate, with no othermaterial than the five crayons of middle hue, tempered with gray andblack. A Color Sphere is always kept in the room for reference, andfive color balls to match the five middle hues are placed in the handsof the youngest pupils. Starting with these middle points in the scalesof Value and Chroma, they learn to estimate rightly all lighter anddarker values, all weaker and stronger chromas, and gradually build up adisciplined judgment of color.

Each study can be made the basis of many variations by a simplechange of one color element, as suggested in the text.

For principles involved in these color groups, seeChapter III.

PLATE III.
Color Studies with TUNED WATER COLORS
in the Upper Grades.

Previous work with measured scales, made by the tuned crayons andtested by reference to the color sphere, have so trained the colorjudgment that children may now be trusted with more flexible material.They have memorized the equable degrees of color on the equator of thesphere, and found how lighter colors may balance darker colors, howsmall areas of stronger chroma may be balanced by larger masses ofweaker chroma, and in general gained a disciplined color sense. Definiteimpressions and clear thinking have taken the place of guess-work andblundering.

Thus, before reaching the secondary school, they are put inpossession of the color faculty by a system and notation similar to thatwhich was devised centuries ago for the musical sense. No system,however logical, will produce the artist, but every artist needs somesystematic training at the outset, and this simple method by measuredscales is believed to be the best yet devised.

Each example on this plate may be made the basis of many variants, bysmall changes in the color steps, as suggested in the65text, and further elaborated inChapter VI.Indeed, the studies reproduced on Plates II. and III. are but a handfulamong hundreds of pleasing results produced in a single school.26

For analysis of the groups and means of recording them, seeChapter III.

Chapter V.
A PIGMENT COLOR SPHERE.27

How to make a color sphere with pigments.

(102)The preceding chapters have built up an ideal color solid, in whichevery sensation of color finds its place and is clearly named by itsdegree of hue, value, and chroma.

It has been shown that the neutral centre of the system is abalancing point for all colors, that a line through this centre findsopposite colors which balance and complement each other; and we are nowready to make a practical application, carrying out these idealrelations of color as far as pigments will permit in a color sphere27 (Fig. 16).

(103)The materials are quite simple. First a colorless globe, mounted so asto spin freely on its axis. Then a measured scale of value, speciallydevised for this purpose, obtained by the daylight photometer.28 Next a set ofcarefully chosen pigments, whose reasonable permanence has been testedby long use, and which are prepared so that they will not glisten whenspread on the surface of the globe, but give a uniformly mat surface.A glass palette, palette knife, and some fine brushes complete thelist.

(104)Here is a list of the paints arranged in pairs to represent67the five sets of opposite hues described in Chapter III., paragraphs61–63:—

(105)These paints have various degrees of hue, value, and chroma, but can betempered by additions of the neutrals, zinc white and ivory black, untileach is brought to a middle value and tested on the value scale. Aftereach pair has been thus balanced, they are painted in their appropriatespaces on the globe, forming an equator of balanced hues.

(106)The method of proving this balance has already been suggested in ChapterIV., paragraph93. It consists of an ingeniousimplement devised by Clerk-Maxwell, which gives us a result of mixingcolors without the chemical risks of letting them come in contact, andalso measures accurately the quantity of each which is used(Fig. 17).

(107)This is called a Maxwell disc, and is nothing more than68a circle of firm cardboard, pierced with a central hole to fit thespindle of a rotary motor, and with a radial slit from rim to centre, sothat another disc may be slid over the first to cover any desiredfraction of its surface. Let us paint one of these discs with Venetianred and the other with viridian and cobalt, the first pair in the listof pigments to be used on the globe.

(108)Having dried these two discs, one is combined with the other on themotor shaft so that each color occupies half the circle. As soon as themotor starts, the two colors are no longer distinguished, and rapidrotation melts them so perfectly that the eye sees a new color, due totheir mixture on the retina. This new color is a reddish gray, showingthat the red is more chromatic than the blue-green. But by stopping themotor and sliding the green disc to cover more of the red one, therecomes a point where rotation melts them into a perfectly neutral gray.No hint of either hue remains, and the pair is said to balance.

(109)Since this balance has been obtained byunequal areas of the twopigments, it must compensate for a lack of equal chroma in the hues (seeparagraphs76, 77); and, to measure thisinequality, a slightly larger disc, with decimal divisions on itsrim, is placed back of the two painted ones. If this scale shows the redas occupying 3⅓ parts of the area, while blue-green occupies 6⅔ parts,then the blue-green must be only half as chromatic as the red, since ittakes twice as much to produce the balance.

(110)The red is then grayed (diminished in chroma by additions of a middlegray) until it can occupy half the circle, with blue-green on theremaining half, and still produce neutrality when mixed by rotation.Each disc now reads 5 on the decimal scale. Lest the graying of redshould have disturbed its value, it is again tested on the photometricscale, and reads 4.7, showing it has been69slightly darkened by the graying process. A little white istherefore added until its value is restored to 5.

(111)The two opposites are now completely balanced, for they are equal invalue (5), equal in chroma (5), and have proved their equality ascomplements by uniting in equal areas to form a neutral mixture. It onlyremains to apply them in their proper position on the sphere.

(112)A band is traced around the equator, divided in ten equal spaces, andlettered R, YR, Y, GY, G, BG, B, PB, P, and RP (see Fig. 18). Thisbalanced red and blue-green are applied with the brush to spaces markedR and BG, care being taken to fill, but not to overstep the bounds, andthe color laid absolutely flat, that no unevenness of value or chromamay disturb the balance.

(113)The next pair, represented by Raw Sienna and Ultramarine, is similarlybrought to middle value, balanced by equal areas on the Maxwell discs,and, when correct in each quality, is painted in the spaces Y and PB.Emerald Green and Purple Madder, which form the next pigment pair, aresimilarly tempered, proved, and applied, followed by the two remainingpairs, until the equator of the globe presents its ten equal steps ofmiddle hues.

An equator of ten balanced hues.

(114)Now comes the total test of this circuit of balanced hues by rotation ofthe sphere. As it gains speed, the colors flash less and less, andfinally melt into a middle gray of perfect neutrality. Had it failed toproduce this gray and shown a tinge of any hue still persisting, weshould say that the persistent hue was in excess, or, conversely, thatits opposite hue was deficient in chroma, and failed to preserve itsshare in the balance.

(115)For instance, had rotation discovered the persistence of reddish gray,it would have proved the red too strong, or its opposite, blue-green,too weak, and we should have been forced to retrace our steps, applyinga correction until neutrality was established by the rotation test.

(116)This is the practical demonstration of the assertion (Chapter I.,paragraph 8) that acolor has threedimensions which can be measured. Each of these ten middle hues hasproved its right to a definite place on the color globe by itsmeasurements of value and chroma. Being of equal chroma, all areequidistant from the neutral centre, and, being equal in value, all areequally removed from the poles. If the warm hues (red and yellow) or thecool hues (blue and green) were in excess, the rotation test of thesphere would fail to produce grayness, and so detect its lack ofbalance.29

A chromatic tuning fork.

(117)The five principal steps in this color equator are made in permanentenamel and carefully safeguarded, so that, if the pigments painted onthe globe should change or become soiled, it could be at once detectedand set right. These five are middle red (so called because midwaybetween white and black, as well as midway between our strongest red andthe neutral centre), middle yellow, middle green, middle blue, andmiddle purple. They may be called theCHROMATIC TUNING FORK, for they serve to establishthe pitch of colors, as the musical tuning fork preserves the pitch ofsounds.

Completion of a pigment color sphere.

(118)When the chromatic tuning fork has thus been obtained,71the completion of the globe is only a matter of patience, for the samemethod can be applied at any level in the scale of value, and a newcircuit of balanced hues made to conform with its position between thepoles of white and black.

(119)The surface above and below the equatorial band is set off by parallelsto match the photometric scale, making nine bands or value zones in all,of which the equator is fifth, the black pole being 0 and the whitepole 10.

(120)Ten meridians carry the equatorial hues across all these value zones andtrace the gradation of each hue through a complete scale from black towhite, marked by their values, as shown in paragraph68. Thus the red scale is R¹, R²,R³, R⁴, R⁵ (middle red), R⁶,R⁷, R⁸, and R⁹, and similarly with eachof the other hues. When the circle of hues corresponding to each levelhas been applied and tested, the entire surface of the globe is spreadwith a logical system of color scales, and the eye gratified withregular sequences which move by measured steps in each direction.

(121)Each meridian traces a scale of value for the hue in which it lies. Eachparallel traces a scale of hue for the value at whose level it is drawn.Any oblique path across these scales traces a regular sequence, eachstep combining change of hue with a change of value and chroma. The morethis path approaches the vertical, the less are its changes of hue andthe more its changes of value and chroma; while, the nearer it comes tothe horizontal, the less are its changes of value and chroma, while thegreater become its changes of hue. Of these two oblique paths the firstmay be called that of a Luminist, or painter like Rembrandt, whosecanvases present great contrasts of light and shade, while the second72is that of the Colorist, such as Titian, whose work shows great fulnessof hues without the violent extremes of white and black.

Total balance of the sphere tested by rotation on any desiredaxis.

(122)Not only does the mount of the color sphere permit its rotation on thevertical axis (white-black), but it is so hung that it may be spun onthe ends of any desired axis, as, for instance, that joining our firstcolor pair, red and blue-green. With this pair as poles of rotation,a new equator is traced through all the values of purple on oneside and of green-yellow on the other, which the rotation test melts ina perfect balance of middle gray, proving the correctness of thesevalues. In the same way it may be hung and tested on successive axes,until the total balance of the entire spherical series is proved.

(123)But this color system does not cease with the colors spread on thesurface of a globe.30 The first illustration of an orange filled with colorwas chosen for the purpose of stimulating the imagination to follow asurface color inward to the neutral axis by regular decrease of chroma.A slice at any level of the solid, as at value 8 (Fig. 10), shows each hue of that level passing by evensteps of increasing grayness to the neutral gray N⁸ of theaxis. In the case of red at this level, it is easily described by thenotation R⁸/₃, R⁸/₂,R⁸/₁, of which the initial and upper numerals donot change, but the lower numeral traces loss of chroma by 3, 2, and 1to the neutral axis.

(124)And there are stronger chromas of red outside the surface, which can bewritten R⁸/₄, R⁸/₅,R⁸/₆, etc. Indeed, our color measurements discoversuch differences of chroma in the various pigments used, that the colortree referred to in paragraphs34, 35, isnecessary73to bring before the eye their maximum chromas, most of which are welloutside the spherical shell and at various levels of value. One way todescribe the color sphere is to suggest that a color tree, the intervalsbetween whose irregular branches are filled with appropriate color, canbe placed in a turning lathe and turned down until the color maxima areremoved, thus producing a color solid no larger than the chroma of itsweakest pigment (Fig. 2).

Charts of the color solid.

(125)Thus it becomes evident that, while the color sphere is a valuable helpto the child in conceiving color relations, in uniting the three scalesof color measure, and in furnishing with its mount an excellent test ofthe theory of color balance, yet it is always restricted to the chromaof its weakest color, the surplus chromas of all other colors beingthought of as enormous mountains built out at various levels to reachthe maxima of our pigments.

(126)The complete color solid is, therefore, of irregular shape, withmountains and valleys, corresponding to the inequalities of pigments. Todisplay these inequalities to the eye, we must prepare cross sections orcharts of the solid, some horizontal, some vertical, and othersoblique.

(127)Such a set of charts forms an atlas of the color solid, enabling one tosee any color in its relation to all other colors, and name it by itsdegree of hue, value, and chroma. Fig. 20 is a horizontal chart of allcolors which present middle value (5), and describes by an unevencontour the chroma of every hue at this level. The dotted fifth circleis the equator of the color sphere, whose principal hues,R⁵/₅. Y⁵/₅,G⁵/₅, B⁵/₅, andP⁵/₅, form the chromatic tuning fork, paragraph117.

(128)In this single chart the eye readily distinguishes some three hundreddifferent colors, each of which may be written by its hue, value, andchroma. And even the slightest variation of one of them can be defined.Thus, if the principal red were to fade slightly, so that it was atrifle lighter and a trifle weaker than the enamel, it would be writtenR^5.1/_4.9, showing it had lightened by 1 per cent.and weakened by 1 per cent. The discrimination made possible by thisdecimal notation is much finer than our present visual limit. Its usewill stimulate finer perception of color.

(129)Such a very elementary sketch of the Color Solid and Color Atlas, whichis all that can be given in the confines of this small book, will beelsewhere presented on a larger and more complete scale. It should becontrasted with the ideal form composed of prismatic colors, suggestedin the last chapter, paragraphs7598, 99, which was shown to be impracticable, butwhose ideal conditions it follows as far as the limitations of pigmentspermit.

(130)Besides its value in education as setting all our color notions inorder, and supplying a simple method for their clear expression, itpromises to do away with much of the misunderstanding that accompaniesthe every-day use of color.

(131)Popular color names are incongruous, irrational, and often ludicrous.One must smile in reading the list of 25 steps in a scale of blue, madeby Schiffer-Muller in 1772:—

The advantage of spacing these 25 colors in 13 groups, some withthree and others with but one example, is not apparent; nor whyultramarine should be several steps above turquoise, for the reverse isgenerally true. Besides which the hue of turquoise is greenish, whilethat of ultramarine is purplish, but the list cannot show this; and theremarkable statement that one kind of blue is “hellish,” while anotheris “celestial,” should rest upon an experience that few can claim.Failing to define color-value and color-hue, the list gives no hint ofcolor-strength, except at76C and D, where one kind of blue is “dying” when the next is “beingborn,” which not inaptly describes the color memory of many a person.Finally, it assures us that Queen blue is “fine” and King blue is“deep.”

This year the fashionable shades are “burnt onion” and “freshspinach.” The florists talk of a “pink violet” and a “green pink.”A maker of inks describes the red as a “true crimson scarlet,”which is a contradiction in terms. These and a host of other namesborrowed from the most heterogeneous sources, become outlawed as soon asthe simple color terms and measures of this system are adopted.

Color anarchy is replaced by systematic color description.

Appendix to Chapter V.

Color schemes based on Brewster’s mistaken theory.

Runge, of Hamburg (1810), suggested that red, yellow, and blue beplaced equidistant around the equator of a sphere, with white and blackat opposite poles. As the yellow was very light and the blue very dark,any coherency in the value scales of red, yellow, and blue wasimpossible.

Chevreul, of Paris (1861), seeking uniform color scales for hisworkmen at the Gobelins, devised a hollow cylinder built up of ten colorcircles. The upper circle had red, yellow, and blue spaced equidistant,and, as in Runge’s solid, yellow was very light and blue very dark. Eachcircle was then made “one-tenth” darker than the next above, until blackwas reached at the base. Although each circle was supposed to liehorizontally, only the black lowest circle presents a level of uniformvalues.

Yellow values increase their luminosity thrice as fast as purplevalues, so that each circle should tilt at an increasing angle, and theupper circle of strongest colors be inclined at 60° to the black base.Besides this fault shared with Runge’s sphere, it falls into another bynot diminishing the size of the lower circles where added blackdiminishes the chroma.

Desire to make colors fit a chosen contour, and the absence of78measuring instruments, cause these schemes to ignore the facts of colorrelation. Like ancient maps made to satisfy a conqueror, they amuse bytheir distortion.

Brewster’s mistaken theory underlies these schemes, as is also thecase with Froebel’s gifts, whose color balls continue to give wrongnotions at the very threshold of color education. As pointed out in theAppendix to Chapter III., the “red-yellow-blue”theory inevitably spreads the warm field of yellow-red too far, andcontracts the blue field, so that balance of color is renderedimpossible, as illustrated in the gaudy chromo and flamingbill-board.

These schemes are criticised by Rood as “not only in the mainarbitrary, but also vague”; and, although Chevreul’s charts werepublished by the government in most elaborate form, their usefulness issmall. Interest in the growth of the present system, because of itsmeasured character, led Professor Rood to give assistance in the tests,and at his request a color sphere was made for the Physical Cabinet atColumbia.

Chapter VI.
COLOR NOTATION.

Suggestion of a chromatic score.

(132)The last chapter traced a series of steps leading to the construction ofa practical color sphere. Each color was tested by appropriateinstruments to assure its degree of hue, value, and chroma, before beingplaced in position. Then the total sphere was tested to detect any lackof balance.

(133)Each color was alsowritten by a letter and two numerals, showingits place in the three scales of hue, value, and chroma. This naturallysuggests, not only a record of each separate color sensation, but also aunion of these records in series and groups to form acolorscore, similar to the musical score by which the measured relationsof sound are recorded.

(134)A very simple form of color score may be easily imagined as atransparent envelope wrapped around the equator of the sphere, andforming a vertical cylinder (Fig. 21). On the envelope the equatortraces a horizontal centre line, which is at 5 of thevaluescale, with zones 6, 7, 8, and 9 as parallels above, and the zones4, 3, 2, and 1 below. Vertical lines are drawn through ten equidistantpoints on this centre line, corresponding with the divisions of thehue scale, and marked R, YR, Y, GY, G, BG, B, PB, P,and RP.

(135)The transparent envelope is thus divided into one hundred compartments,which provide for ten steps of value in each of the ten middle colors.Now, if we cut open this envelope along one of the verticals,—as,for instance, red-purple (RP), it may be spread out, making a flat chartof the color sphere (Fig. 22).

Why green is given the centre of the score.

(136)A cylindrical envelope might be opened on any desired meridian, but itis an advantage to have green (G) at the centre of the chart, and it istherefore opened at the opposite point, red-purple (RP). To the right ofthe green centre are the meridians of green-yellow (GY), yellow (Y),yellow-red (YR), and red (R), all of which are known aswarmcolors, because they contain yellow and red. To the left are themeridians of blue-green (BG), blue (B), purple-blue (PB), and purple(P), all of which are calledcool colors, because they containblue. Green, being neither warm nor cold of itself, and becoming so onlyby additions of yellow or of blue, thus serves as a balancing point orcentre in the hue-scale.31

(137)The color score presents four large divisions or color fields made bythe intersection of the equator with the meridian of green. Above thecentre are all light colors, and below it are all dark colors. To theright of the centre are all warm colors, and to the left are all coolcolors. Middle green (5G⁵/₅) is the centre ofbalance for these contrasted qualities, recognized by all81practical color workers. The chart forms a rectangle whose length equalsthe equator of the color sphere and its height equals the axis(a proportion of 3.14:1), representing a union and balance of thescales of hue and of value. This provides for two color dimensions; but,to be complete, the chart must provide for the third dimension,chroma.

(138)Replacing the chart around the sphere and joining its ends, so that itre-forms the transparent envelope, we may thrust a pin through at anypoint until it pierces the surface of the sphere. Indeed, the pin can bethrust deeper until it reaches the neutral axis, thus forming a scale ofchroma for the color point where it enters (see paragraph12). In the same way any colors on the sphere, within thesphere, or without it, can have pins thrust into the chart to mark theirplace, and the length by which each pin projects can be taken as ameasure of chroma. If the chart is now unrolled, it retains the pins,which by their place describe the hue and value of a color, while theirlength describes its chroma.

Pins stuck into the score represent chroma.

(139)With this idea of the third color dimension incorporated in the score wecan discard the pin, and record its length by a numeral. Any dot placedon the score marks a certain degree of hue and value, while a numeralbeside it marks the degree of chroma which it carries, uniting with thehue and value of that point to give us a certain color. Glancing over aseries of such color points, the eye easily grasps their individualcharacter, and connects them into an intelligible series.

(140)Thus a flat chart becomes the projection of the color solid, and anycolor in that solid is transferred to the surface of the chart,retaining its degrees of hue, value, and chroma. So far the scales havebeen spoken of as divided into ten steps, but82they may be subdivided much finer, if desired, by use of the decimalpoint. It is a question of convenience whether to make a small scorewith only the large divisions, or a much larger score with a hundredtimes as many steps. In the latter case each hue has ten steps, themiddle step of green being distinguished as 5G⁵/₅to suggest the four steps 1G, 2G, 3G, 4G, which precede it, and 6G, 7G,8G, and 9G, which follow it toward blue-green.

The score preserves color records in a convenient shape.

Such a color score, or notation diagram, to be made small or large asthe case demands, offers a very convenient means for recording colorcombinations, when pigments are not at hand.

(141)To display its three dimensions, a little model can be made with threevisiting cards, so placed as to present their mutual intersection atright angles (Fig. 24).

5G⁵/₅ is their centre of mutual balance.A central plane separates all colors into two contrasted fields. Tothe right are all warm colors, to the left are all cool colors. Each ofthese83fields is again divided by the plane of the equator into lighter colorsabove and darker colors below. These four color fields are againsubdivided by a transverse plane through 5G⁵/₅into strong colors in front and weak colors beyond orbehind it.

(142)Any color group, whose record must all be written to the right of thecentre, is warm, because red and yellow are dominant. One to the left ofthe centre must be cool, because it is dominated by blue. A groupwritten all above the centre must have light in excess, while onewritten entirely below is dark to excess. Finally, a score writtenall in front of the centre represents only strong chromas, while onewritten behind it contains only weak chromas. From this we gather that abalanced composition of color preserves some sort of equilibrium,uniting degrees of warm and cool, of light and dark, and of weak andstrong, which is made at once apparent by the dots on the score.

(143)A single color, like that of a violet, a rose, or a buttercup,appears as a dot on the score, with a numeral added for its chroma.A parti-colored flower, such as a nasturtium, is shown by two dotswith their chromas, and a bunch of red and yellow flowers will give bytheir dots a color passage, or “silhouette,” whose warmth and lightnessis unmistakable.

The chroma of each flower written with the silhouette completes therecord. The hues of a beautiful Persian rug, with dark redpredominating, or a verdure tapestry, in which green is dominant, or aJapanese print, with blue dominant, will trace upon the score a patterndescriptive of its color qualities. These records, with practice, becomeas significant to the eye as the musical score. The general character ofa color combination is apparent at a glance, while its degrees of chromaare readily joined to fill out the mental image.

(144)Such a plan of color notation grows naturally from the spherical systemof measured colors. It is hardly to be hoped, in devising a color score,that it should not seem crude at first. But the measures forming thebasis of this record can be verified by impartial instruments, and havea permanent value in the general study of color. They also afford somedefinite data as to personal bias in color estimates.

(145)This makes it possible to collect in a convenient form two contrastingand valuable records, one preserving such effects of color as aregenerally called pleasing, and another of such groups as are foundunpleasant to the eye. Out of such material something may be gained,more reliable than the shifting, personal, and contradictory statementsabout color harmony now prevalent.

31.To put this in terms of the spectrum wave lengths, long waves at the redend of the spectrum give the sensation of warmth, while short waves atthe violet end cause the sensation of coolness. Midway between theseextremes is the wave length of green.

Chapter VII.
COLOR HARMONY.

Colors may be grouped to please or to give annoyance.

(146)Attempts to define the laws of harmonious color have not attained markedsuccess, and the cause is not far to seek. The very sensationsunderlying these effects of concord or of discord are themselvesundefined. The misleading formula of my student days—that threeparts of yellow, five parts of red, and eight parts of blue wouldcombine harmoniously—was unable to define thekind of red,yellow, and blue intended; that is, the hue, value, and chroma of eachof these colors was unknown, and the formula meant a different thing toeach person who tried to use it.

(147)It is true that a certain red, green, and blue can be united in suchproportions on Maxwell discs as to balance in a neutral gray; but theslightest change in either the hue, value, or chroma, of any one ofthem, upsets the balance. A new proportion is then needed to regainthe neutral mixture. This has already been shown in the discussion oftriple balance (paragraph 82).

(148)Harmony of color has been still further complicated by the use of termsthat belong to musical harmony. Now music is ameasured art, andhas found a set of intervals which are defined scientifically. The twoarts have many points of similarity; and the impulses of sound waves onthe ear, like those of light waves on the eye, are measured vibrations.But they are far apart in their scales, and differ so much in important86particulars that no practical relationship can be set up. The intervalsof color sensation require fit names and measures, ere their infinitevariety can be organized into a fixed system.

(149)Any effort to compare certain sounds to certain colors soon leads to thewildest vagaries.

Harmony of sound is unlike harmony of color.

(150)The poverty of color language tempts to a borrowing from the richerterminology of music. Musical terms, such as “pitch, key, note, tone,chord, modulation, nocturne, and symphony,” are frequently used in thedescription of color, serving by association to convey certain vagueideas.

(151)In the same way the termcolor harmony, from association withmusical harmony, presents to the mind an image of colorarrangement,—varied, yet well proportioned, grouped in orderlyfashion, and agreeable to the eye. But any attempt to define this imagein terms of color is disappointing. Here is a beautiful Persian rug: whydo we call it beautiful? One says “because its colors arerich.”Why are they rich? “Because they aredeep in tone.” What doesthat mean? The double-bass and the fog-horn aredeep in tone, butnot necessarily beautiful on that account. “Oh, no,” says another, “itis all inone harmonious key.” But what is a key of color? Is itmade by all the values of one color, such as red, or by all the hues ofequal value, such as the middle hues in our color solid?

(152)Certainly it is neither, for the rug has both light and dark colors;and, of the reds, yellows, greens, and blues, some are stronger andothers weaker. Then what do we mean by a key of color? One must eithercontinue to flounder about or frankly confess ignorance.

(153)Musical harmony explains itself in clear language. It87is illustrated by fixed and definite sound intervals, whose measuredrelations form the basis of musical composition. Each key has anunmistakable character, and the written score presents a statement thatmeans practically the same thing to every person of musicalintelligence. But the adequate terms of color harmony are yet to beworked out.

Let us leave these musical analogies, retaining only the clue thata measured and orderly relation underlies the idea of harmony.The color solid which has been the subject of these pages is built uponmeasured color relations. It unites measured scales of hue, value, andchroma, and gives a definite color name to every sensation from themaxima of color-light and color-strength to their disappearance indarkness.

(154)Must not this theoretical color solid, therefore, locate all theelements which combine to produce color harmony or color discord?32

(155)Instead of theorizing, let us experiment. As a child at the piano, whofirst strikes random and widely separated notes, but soon seeks for theintervals of a familiar air, so let us, after roaming over the colorglobe and its charts, select one familiar color, and study what otherswill combine with it to please the eye.

(156)Here is a grayish green stuff for a dress, and the little girl who is towear it asks what other colors she may use with it. First let us find iton our instrument, so as to realize its relation to other degrees ofcolor. Its value is 6,—one step above the equator of middle value.Its hue is green, G, and its chroma 5. It is writtenG⁶/₅.

(157)Color paths lead out from this point in every direction.88Where shall we find harmonious colors, where discordant, where thosepaths most frequently travelled? Are there new ones still to beexplored?

(158)There are three typical paths: one vertical, with rapid change ofvalue;another lateral, with rapid change of hue; and athirdinward, through the neutral centre to seek the opposite color field.All other paths are combinations of two or three of these typicaldirections in the color solid.

Three typical color paths.

(159)1. The vertical path finds only lighter and darker values ofgray-green,—“self-colors or shades,” they are generallycalled,—and offers a safe path, even for those deficient in colorsensation, avoiding all complications of hue, and leaving the eye freeto estimate different degrees of a singlequality,—color-light.

(160)2. The lateral path passes through neighboring hues on either side. Inthis case it is a sequence from blue, through green into yellow. This issimply change of hue, without change of value or chroma if the path belevel, but, by inclining it, one end of the sequence becomes lighter,while the other end darkens. It thus becomes an intermediate between thefirst and second typical paths, combining, at each step, a changeof hue with a change of value. This is more complicated, but also moreinteresting, showing how the character of the gray-green dress will beset off by alighter hat of Leghorn straw, and further improvedby a trimming ofdarker blue-green. The sequence can be madestill more subtle and attractive by choosing a straw whose yellow isstronger than the green of the dress, while aweaker89chroma of blue-green is used in the trimming. This is clearly expressedby the notation thus: Y⁸/₇,G⁶/₅, BG⁴/₃, and written onthe score by three dots and their chromas,—7, 5, and 3 (seeFig. 23).

(161)3. The inward path which leads by increase of gray to the neutralcentre, and on to the opposite hue red-purple,RP⁴/₅, is full of pitfalls for the inexpert. Itcombines great change of hue and chroma, with small change of value.

(162)If any other color point be chosen in place of gray-green, the sametypical paths are just as easily traced, written by the notation, andrecorded on the color score.

These paths trace sequences from any point in the color solid.

(163)In the construction of the color solid we saw that its scales were madeof equal steps in hue, value, and chroma, and tested by balance on thecentre of neutral gray. Any step will serve as a point of departure totrace regular sequences of the three types. The vertical type is asequence of value only. It is somewhat tame, lacking the change of hueand chroma, but giving a monotonous harmony of regular values. Thehorizontal type traces a sequence of neighboring hues, less tame thanthe vertical type, but monotonous in value and chroma. The inward typeconnects opposite hues by a sequence of chroma balanced on middle gray,and is more stimulating to the eyes.

(164)These paths have so far been treated as made up of equal steps in eachdirection, with the accompanying idea of equal quantities of color ateach step. But by usingunequal quantities of color, the balancemay be preserved by compensations to the intervals that separate thecolors (see paragraphs109, 110).

Unequal color quantities compensated by relations of hue, value, andchroma.

(165)Small bits of powerful color can be used to balance large90fields of weak chroma. For instance, a spot of strong reddishpurple is balanced and enhanced by a field of gray-green. So an amethystpin at the neck of the girl’s dress will appear to advantage with thegown, and also with the Leghorn straw. But a large field of strongcolor, such as a cloth jacket of reddish purple, would be fatal to themeasured harmony we seek.

(166)This use of a small point of strong chroma, if repeated at intervals,sets up a notion of rhythm; but, in order to be rhythmic, there must berecurrent emphasis, “a succession of similar units, combiningunlike elements.” This quality must not be confused with the unaccentedsuccession, seen in a measured scale of hue, value, or chroma.

Paper masks to isolate color intervals.

(167)A sheet of paper large enough to hide the color sphere may be perforatedwith three or more openings in a straight line, and applied against thesurface, so as to isolate the steps of any sequence which we wish tostudy. Thus the sequence given in paragraph160—Y⁸/₇,G⁶/₅, BG⁴/₃—may bechanged to bring it on the surface of the sphere, when it readsY⁸/₃, G⁶/₅,BG⁵/₅. A mask with round holes, spaced so asto uncover these three spots, relieves the eye from the distraction ofother colors. Keeping the centre spot on green, the mask may be moved soas to study the effect of changing hue or value of the other two stepsin the sequence.

(168)The sequence is lightened by sliding the whole mask upward, and darkenedby dropping it lower. Then the result of using the same intervals inanother field is easily studied by moving the mask to another part ofthe solid.

(169)Change of interval immediately modifies the character of a colorsequence. This is readily shown by having an under-mask, with a long,continuous slit, and an over-mask whose perforations91are arranged in several rows, each row giving different spaces betweenthe perforations. In the case of the girl’s clothing, the same sequenceproduces quite a different effect, if two perforations of the over-maskare brought nearer to select a lighter yellow-green dress, while theends of the sequence remain unchanged. To move the middle perforationnear the other end, selects a darker bluish green dress, on which thetrimming will be less contrasted, while the hat appears brighter thanbefore, because of greater contrast.

(170)The variations of color sequence which can thus be studied out by simplemasks are almost endless; yet upon a measured system the character ofeach effect is easily described, and, if need be, preserved by a writtenrecord.

Invention of color groups.

(171)Experiments with variable masks for the selection of color intervals,such as have been described, soon stimulate the imagination, so that itconceives sequences through any part of the color solid. The color imagebecomes a permanent mental adjunct. Five middle colors, tempered withwhite and black, permit us to devise the greatest variety of sequences,some light, others dark, some combining small difference of chroma withlarge difference of hue, others uniting large intervals of chroma withsmall intervals of hue, and so on through a well-nigh inexhaustibleseries.

(172)As this constructive imagination gains power, the solid and its chartsmay be laid aside.We can now think color consecutively. Eachcolor suggests its place in the system, and may be taken as a point ofdeparture for the invention of groups to carry out a desiredrelation.

(173)This selective mental process is helped by the score described92in the last chapter; and the quantity of each color chosen for the groupis easily indicated by a variable circle, drawn round the various pointson the diagram. Thus, in the case of the child’s clothes, a largecircle around G⁶/₅ gives the area of that color ascompared with smaller circles around Y⁸/₇ andBG⁴/₃, representing the area of the straw and thetrimming.

(174)When the plotting of color groups has become instinctive from longpractice, it opens a wide field of color study. Take as illustration thewings of butterflies or the many varieties of pansies. These fascinatingcolor schemes can be written with indications of area that record theirdifferences by a simple diagram. In the same way, rugs, tapestries,mosaics,—whatever attracts by its beauty and harmony ofcolor,—can be recorded and studied in measured terms; and themental process of estimating hues, values, chromas, and areas byestablished scales must lead the color sense to finer and finerperceptions.

The same process serves as well to record the most annoying andinharmonious color groups. When sufficient of these records have beenobtained, they furnish definite material for a contrast of the colorcombinations which please, with those that cause disgust. Such acontrast should discover some broad law of color harmony. It will thenbe in measured terms which can be clearly given; not a vague personalstatement, conveying different meanings to each one whohears it.

Constant exercise needed to train the color sense.

(175)Appreciation of beautiful color grows by exercise and discrimination,just as naturally as fine perception of music or literature. Each is anoutlet for the expression of taste,—a language which may be usedclumsily or with skill.

(176)As color perception becomes finer, it discards the more93crude and violent contrasts. A child revels in strong chromas, butthe mark of a colorist is ability to employ low chroma withoutimpoverishing the color effect. As a boy’s shrieks and groans can betempered to musical utterance, so his debauches in violent red, green,and purple must be replaced by tempered hues.

(177)Raphael, Titian, Velasquez, Corot, Chavannes, and Whistler are mastersin the use of gray. Personal bias may lead one colorist a little moretoward warm colors, and another slightly toward the cool field, in eachcase attaining a sense of harmonious balance by tempered degrees ofvalue and chroma.33

(178)It is not claimed that discipline in the use of subtle colors will makeanother Corot or Velasquez, but it will make for comprehension of theirskill. It is grotesque to watch gaudily dressed persons going intoecstasies over the delicate coloring of a Botticelli, when the internalas well as the external evidence is against them.

(179)The colors which we choose, not only in personal apparel, but in ourrooms and decorations, are mute witnesses to a stage of colorperception.

If that perception is trained to finer distinctions, the mind can nolonger be content with coarse expression. It begins to feel anincongruity between the “loud” color of the wall paper, bought becauseit was fashionable, and the quiet hues of the rug, which was a gift fromsome artistic friend. It sees that, although the furniture is coveredwith durable and costly materials, their color “swears” at that of thecurtains and wood-work. In short, the94room has been jumbled together at various periods, without any plan orsense of color design.

(180)Good taste demands that a room be furnished, not alone for convenienceand comfort, but also with an eye to the beauty of the various objects,so that, instead of confusing and destroying the colors, each mayenhance the other. And, when this sense of color harmony is aroused, itselects and arranges the books, the rugs, the lamp shade, the souvenirsof travel and friendship, the wall paper, pictures, and hangings, sothat they fit into a color scheme, not only charming to the eye at firstglance, but which continues to please the mind as it traces out anintelligent plan, bringing all into general harmony.

(181)Nor will this cease when one room has been put to rights. Such acoloristic attitude is not satisfied until the vista into the nextapartment is made attractive. Or should there be a suite of rooms, itdemands that, with variety in each one, they all be brought intoharmonious sequence. Thus the study of color finds immediate andpractical use in daily life. It is a needed discipline of color vision,in the sense that geometry is a discipline of the mind, and it alsoenters into the pleasure and refinement of life at every step. Skill orawkwardness in its use exerts as positive an influence upon us as do theharmonies and discords of sound, and a far more continuous one. It isthought a defect to be unmusical. Should it not be considered a mark ofdefective cultivation to be insensitive to color?

(182)In this slight sketch of color education it has been assumed that we areto deal with those who have normal perceptions. But there are some whoinherit or develop various degrees of color-blindness; and a word intheir behalf may be opportune.

(183)A case of total color-blindness is very rare, but a few95are on record. When a child shows deficient color perception,34 a littlecare may save him much discomfort, and patient training may correct it.If he mismatches some hues, confuses their names, seems incapable of thefiner distinctions of color, study to find the hues which he estimateswell, and then help him to venture a little into that field where hisperception is at fault. Improvement is pretty sure to follow when thisis sympathetically done. One student, who never outgrew the habit ofgiving a purplish hue to all his work, despite many expedients and theuse of various lights and colored objects to correct it, is the singleexception among hundreds whom it has been my privilege to watch as theyimproved their first crude estimates, and gained skill in expressingtheir sense of Nature’s subtle color.

(184)To sum up, the first chapter suggests a measured color system in placeof guess-work. The next describes the three color qualities, andsketches a child’s growth in color perception. The third tells howcolors may be mingled in such proportions as to balance. After theimpracticability of using spectral color has been shown in the fourthchapter, the fifth proceeds to build a practical color solid. The sixthprovides for a written record of color, and the last applies all thathas preceded to suggestions for the study of color harmony.

(185)Wide gaps appear in this outline. There is much that deserves fullertreatment. But, if the search for refined color and a clearer outlookupon its relations are stimulated by this fragmentary sketch, some ofits faults may be overlooked.

 
REPRODUCTION OF FLOWER STUDIES,
PAINTED WITH MUNSELL WATER COLOR
Published by
Wadsworth, Howland & Co., Incorporated, Boston, Mass.