120. Before entering into the details of these applications, I think it necessary to offer some considerations which will enable me to establish some propositions or principles, to which I shall have frequent occasion to refer. I propose to give—

121. 1. Definitions of several expressions applicable to colours and their modifications. 2. The means of representing and defining colours and their modifications by the aid of diagrams. 3. A classification of the harmonies of colours. 4. A view of some arrangements of the primary colours with white, black, and grey.

SECTION I.

Definition of the words Tones, Scales, and Hues.

122. The words Tones and Hues recur continually, both in common language and in that of artists; yet they are not so well defined as to be free from ambiguity, or to be well understood.

123. The word TONE of a colour will be employed exclusively to designate the various modifications which that colour, in its greatest intensity, is capable of receiving from white, which lowers its tone, or of black, which heightens it.

124. The word SCALE will be applied to the assemblage of tones of the same colour, thus modified. The pure colour is the normal tone of the scale, if the normal tone does not belong to a broken or reduced scale—i.e., to a scale, of which all the tones are made dull with black (149).

125. The word HUE will be applied exclusively to the modifications which a colour receives from the addition of a small quantity of another. We shall speak, for example, of the tones of the blue scale, the tones of the red scale, &c. We say the hues of blue to designate all the scales whose colours, still remaining blue, yet differ from pure blue; each hue comprehending the tones which constitute a scale more or less allied to the blue scale.

126. I have defined the tones of a colour to be the various modifications, which that colour at its maximum of intensity is capable of receiving from black and white; it must be observed that the condition “maximum of intensity for receiving black,” is absolutely essential to this definition; for if black be added to a tone below the maximum, it would pass into another scale. Artists distinguish colours as pure, broken, reduced, grey, or dull.

127. Pure colours are those termed simple, red, yellow, blue, and those which result from their binary compounds, orange, green, violet, and their hues. (150.) Broken colours are the pure colours mixed with black, from the tone of the lightest to the deepest. According to these definitions, it is evident that in all the scales of simple and binary colours, the tones which are above the pure colour are broken tones.

128. Artists, and especially painters and dyers, admit that the mixture of three primary colours, in a certain proportion, gives black; hence, when these three colours are so mixed that two predominate, black will result, formed from the union of the whole of the colour, which is in small quantity, within suitable proportions of the two predominant colours. For example, if blue be mixed with red and yellow, a little black is produced, which reduces or breaks the orange.

129. We must remember that the primary colours of painters are not those of the prismatic spectrum, but substances employed by them, as red, yellow, and blue colours.

SECTION II.

Of Diagrams designed to Represent and Define
Colours and their Modifications.

130. Various contrivances have been proposed under the titles of Tables, Scales, Colour-Circles, Chromatometers, &c., for representing either by numbers or a rational nomenclature, colours and their modifications. They are generally founded on these three propositions:—1. There are three primary colours. 2. Equal portions of these colours being mixed, produce pure secondary colours. 3. Equal portions of the three primary colours produce black.

131. But we know of no substance which exhibits pure colour; that is, which reflects only one kind of coloured rays, whether pure red, pure yellow, or pure blue. And since it is impossible to procure pure colouring matters, how can it be said that orange, green, and violet are composed of two simple colours mixed in equal proportions? Or that black consists of a mixture of equal parts of three simple colours?

These chromatic tables, &c., point out mixtures which do not produce the results deducible from the principles on which they are said to be based.

132. But most of the blue, red, and yellow colours with which we are acquainted, give, by their binary compounds, violet, green, and orange inferior in brilliancy to the natural violet, green and orange colours of objects. This result would be explained by admitting that colours mixed two by two, reflect at least two kinds of coloured rays; and that where there is any mixture of colours which reflect separately red, yellow, and blue, there is produced a certain amount of black which reduces the brilliancy of the mixture.

133. Conformably with this view, the violet, green, and orange colours which result from a mixture of coloured matters, are most brilliant when the respective colours of these materials approach each other. For example, a mixture of blue and red inclines more to violet than a mixture of blue and yellow inclines to green, and that of red and yellow inclines still more to orange.

134. In order to represent all the modifications that I have called tones and hues of colours, as well as the relations which exist between those that are complementary to each other, I have devised the following diagram (Plate 7). From a centre, c, I describe two circumferences, y Y. I divide each of these by means of three rays, c a, c b, c d, into arcs of 120 degrees each. I divide the portion of each ray comprised between the two circles y Y into twenty parts, which represent as many tones of the colours red, yellow, and blue.

135. In each of the scales of these three colours there is one tone, which, when pure, represents the colour of the scale to which it relates. I therefore call it the normal tone of that scale. If we represent a unit of surface s, entirely covered by the pigment which reflects the normal colour, and if we suppose that this colouring matter is equally distributed over the surface a 1, we shall represent the tones superior to the normal tone by the unit of surface covered with 1 of the normal colour, plus the quantities of black increasing with the number of tones; and we shall represent the inferior tones by the unit of surface covered with a fraction of the quantity 1, constituting the normal tone, mixed with (——) quantities of black, as the tone has a less elevated number. If the tone 15 of the red scale be the normal tone, the normal tone of the yellow scale will have a lower number, while the normal tone of the blue scale will have a higher number. This depends upon the unequal lightness of the colours.

136. If each arc of 120° be divided into two of 60° and if radii pass through the points of division, beginning at y, there will be represented twenty tones of the orange, green, and violet scales, the colours at the extremities of each diameter being complementary to one another. Each arc of 60° might be divided into arcs of 30°, and thus would be obtained radii representing twenty tones of scales, which I shall call orange-red, orange-yellow, greenish-yellow, greenish-blue, bluish-violet, and violet-red.

137. By dividing each arc into five, for example, by means of five radii, which I divide into twenty parts each, beginning at the circumference y, I shall obtain sixty new scales.

138. Beginning with red, I designate them as follows:—

I attach no importance to this nomenclature; I employ it only as the simplest to distinguish the seventy-two scales just described. The number may be increased indefinitely, by inserting as many as we choose between the above.

142. Let us now represent the gradations of each colour in the scales of the circle by the addition to it of black, progressively increasing till it becomes pure black. Imagine a quadrant whose radius is equal to that of the circle, and arranged so as to turn upon an axis perpendicular to the plane of the circle. Divide this quadrant, 1st, by concentric arcs y yʹ, which coincide with the circles denoted by the same letters; 2nd, by ten radii, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. Divide each of these radii into twenty parts, representing twenty tones, corresponding to the tones of the scales represented on the circle.

143. I suppose that the tenth radius comprises the gradations of normal black, covering the half-circle described by the movement of the quadrant upon its axis; this black mixed in decreasing quantities, with increasing quantities of white, gives the twenty tones of normal grey, and ends by being lost in the white situated above the tone 1. I suppose, further, that the normal tone of each of the scales taken upon each of the radii of the quadrant 1, 2, 3, 4, 5, 6, 7, 8, 9, is formed of the mixture of black with the colour of any of the scales that the circle contains, and in such a proportion that the normal tone 15 of that scale is represented by the unit of surface covered with 1, or ¹⁰/₁₀ of red.

144. The tone 15 of the scale of the

These proportions relate to the effect of the mixtures upon the eye, and not to the material quantity of the red and black substances.

145. We see then—1. That each of these tones, 15, composed of colour and black, reduced by white and deepened by black, gives a scale of twenty tones, so much the more broken as they are nearer the scale of normal black. 2. That the quadrant by its movement upon the axis of the circle, represents the scales of every colour except red, broken by black. These broken scales are equidistant, and are formed of equidistant tones. 3. That all the colours are thus contained in a circle, whose plan comprehends the pure colours; the central space, black; and the intermediate space the pure colours, broken by the various proportions of black.

146. The diagram, as just described, thus represents the lowering of pure colours by white, and their gradation by black; their modifications by their mutual mixtures, the modification of hues, and the modification of breaking. We will presently inquire into the possibility of realizing it by means of coloured materials.

147. We have presumed—1. That the normal tone of each of the scales is as pure as possible. 2. That the tones bearing the same number in all the scales,—both those of the pure colours and those of the broken colours,—are, to the sight, of equal depth. 3. That if three tones, of the same number, be taken in three consecutive scales, the tone of the intermediate scale is the mean between the colours of the extreme scales. It is thus easy to explain the modifications of a pure colour commencing with its normal tone.

148. These modifications are so produced that—1. The Pure Colour never leaves its Scale.—The modification is in the direction of the radius of the circle—proceeding from the normal tone towards the centre, it gains white; while proceeding from the normal tone towards the circumference, it gains black.

149. 2. The Pure Colour leaves its Scale by the addition of Black.—In this case the various scales comprised in the quadrant perpendicular to the circle, begin at the normal tone of one of the pure scales of the circle with which the quadrant coincides. This normal tone, resulting from a quantity of colour represented by unity, covering a unit of surface s, the normal tones of the quadrant result from the mixture of black and a fraction of unity of the colour. These mixtures constitute broken colours, each covering a unit of surface s, and are of the same depth as the normal tone of the pure colour. The fraction of the quantity of colour is, in the broken normal tones, so much less, as the scales, to which these tones belong, approximate to the vertical axis of the semicircle.

Besides, each normal tone of the scales of the quadrant is modified, like the normal tones of the scales of the circle, by increasing quantities of white towards the centre, and of increasing quantities of black towards the circumference.

150. 3. A Pure Colour is modified by the addition of another Pure Colour.—In this case hues are formed so much more resembling each other, as the quantities of the second colour are smaller. These modifications are made circularly, so that the tones retain their numbers. Thus admitting, with painters and dyers, that there are only three primary colours, and that by combining these two by two, we obtain all the pure complex colours; and by combining them in threes, all the broken colours; we find that it is possible to represent by this hypothesis, all the modifications of colours.

151. Another advantage of this construction is that of giving to all artists who may make applications of the law of simple contrast, the complementaries of all the pure colours; since the colours of the circular plan which are found at the extremities of the same diameter are complementary to each other. For example, not only are red and green, blue and orange, yellow and violet on the same diameter, but it is so with orange-red and bluish-green, and yellowish-green and violet-red; of red No. 1 and of green No. 1; so that all the colours opposed to each other are mutually complementary.

152. The complementary of a colour contiguous to another being once known, it is easy, according to the principles of combination, to determine the modification that the second must receive from the first; since this modification is the result of the mixture of the complementary with the contiguous colour. In fact, if there is no difficulty when the result is that of the non-complementary mixture with a simple colour, red, yellow, and blue, with a binary colour, orange, green, violet (using the language of painters, 76), there is no greater difficulty when the result is that of the mixture of two binary colours. For, the complementary being much less intense than the colour with which it is mixed, the result will be found by subtracting from the last binary colour the portion of its simple colour, which with the complementary forms white, or in other words, neutralises it.

153. Examples.—1. Orange being added as a complementary to green, neutralises a portion of its blue, and consequently makes it appear less blue or more yellow.

2. Orange being added as complementary to violet, neutralises a portion of its blue, and consequently makes it appear less blue or more red.

3. Green being added as complementary to violet, neutralises a portion of its red, and consequently makes it appear less red or more blue.

154. These three examples are easily explained by subtracting from the binary colour a portion of its simple colour which is identical with that contiguous to it. Thus:—

155. To put the diagram into practice we must adopt invariable types of colour, either in the solar spectrum, or in polarized light, or coloured rings, or colours developed in a constant manner, by any process whatever; then imitate them with the utmost fidelity, by means of colouring matters which should be applied to the circular plan of our chromatic diagram.

These types must be sufficiently numerous to reproduce the principal colours, in order that a practised eye may without difficulty insert all the tones of the same scale and all the hues of which types are wanting. In fact the diagram thus established, should present terms so near that the various colours of the natural bodies might be referred to them.

156. 1. That it represents all the Modifications resulting from the Mixture of Colours.—Thus any colour lowered by white and deepened with black may, retaining its place in the scale, give rise to an infinite variety of tones; infinite, inasmuch as an unlimited number may be inserted from tone 1 to tone 20.

157. 2. Pure colours, by their mutual modifications, may produce an infinite variety of hues; for between two adjacent hues we may insert as many as we desire.

158. 3. The normal tone of a pure colour represented by a quantity equal to 1, covering the unit of surface, is the commencement of the normal tones and scales proceeding towards black; these normal tones being represented by black and a quantity of colour less than unity, constituting the mixtures which cover a unit of surface s, and colour it of a tone which has the same number as the normal tone of the pure scale to which it relates. It is understood that in proceeding from this tone to the corresponding tone of normal black, we may insert an unlimited number of mixtures of colour and black.

159. The modifications of colours, thus indicated by the diagram, render it extremely easy to understand the definitions given above (123) of the words, scales, tones, hues, pure and broken colours.

160. 2. It affords the means of knowing the complementaries of every colour, since the names written at the two extremities of any one diameter indicate the colours complementary to each other.

161. Examples.—a. Suppose it be required to know the mutual influence of blue and yellow; at one extremity of a diameter we read the word blue, and at its opposite end, the word orange; showing that blue tends to give orange to yellow. Again, at the end of another diameter we read the word yellow, and at its opposite, the word violet; by which we see that yellow tends to give violet to blue.

162. b. Suppose green and blue be contiguous; at one extremity of a diameter we read the word green, and at its opposite end, red; showing that green tending to give red to blue, must render it more violet. Again, at one end of a diameter we read the word blue, and at its opposite end, orange. But what arises from the mixture of green and orange? The orange will tend to neutralize its complementary, blue, in the green; and as it is always too feeble to neutralize all the blue, its influence will be limited to neutralizing a portion of it; whence it results that green, contiguous to blue, will appear more yellow than it really is.

163. c. Let green and yellow be contiguous, we shall see in like manner that the green, by imparting red to the yellow, will render it orange; and that violet, the complementary of yellow, by neutralizing some yellow in the green, will make the green appear bluer, or less yellow.

164. 3. A third advantage of this diagram, which distinguishes it from other chromatic diagrams, is, that it affords the preceding advantages, without being coloured.

165. 4. A fourth advantage is that of its manifesting to all artists who use coloured materials of a definite size, especially the workers of tapestry, carpets and the like, the relation of number which must exist between the tones of the various scales which they work together.

SECTION III.

Harmony of Colours.

166. The eye has an undoubted pleasure in seeing colours, independently of the design and every other quality of the object which displays them. A suitable example to demonstrate this, is the wainscoting of an apartment in one or more flat tints which only attract the eyes and affect them more or less agreeably, as the colours are well or badly chosen.

167. First Case. Agreeable Colour.—Every one, whose eyes are well organized, derives pleasure from looking at the coloured rays transmitted through a coloured glass, whether it be red, orange, yellow, green, blue, or violet.

168. Second Case. Different Tones of the same Scale of Colour.—The simultaneous view of the series of tones of the same scale, which commences with white and ends with dark brown, gives undoubtedly an agreeable sensation, especially if the tones have equal and sufficiently numerous intervals; for example, from eighteen to thirty.

169. Third Case. View of Different Colours, belonging to adjacent Scales, assorted conformably to Contrast.—The simultaneous view of different colours, belonging to scales more or less allied to each other, may be agreeable; but the assortment of scales producing this effect is very difficult to obtain, because the more nearly the scales approach, the more frequently it happens that one of the colours injures that which is adjacent to it, and even both are reciprocally injurious. The painter may, however, take advantage of this harmony, by sacrificing one of the colours, which he subdues, to make the other more brilliant.

170. Fourth Case. View of very different Colours, belonging to very distant Scales, arranged conformably to Contrast.—The simultaneous view of complementary colours, or of binary assemblages of colours, which, without being complementary, are yet very different, is also an undoubtedly agreeable sensation.

171. Fifth Case. View of various Colours, assorted more or less according to the Law of Contrast, being seen through a glass of a colour not deep enough to allow all the colours peculiar to the glass to be visible, afford a spectacle which is not without its charm, and which is placed between that produced by the tones of the same scale, and that which is produced by various colours; for it is evident, that if the glass were of a deeper colour, it would cause objects to be seen of the colour peculiar to it.

172. Hence we infer that there are six distinct harmonies of colour, comprised in two species.

First Species—Harmonies of Analogy.—1. The harmony of scale, produced by the simultaneous view of different tones of the same scale, more or less approximating. 2. The harmony of hues, produced by the simultaneous view of tones of the same, or nearly of the same depth, belonging to neighbouring scales. 3. The harmony of a dominant coloured light, produced by the simultaneous view of various colours assorted according to the law of contrast, but one of them predominating, as would result from the view of these colours through a slightly-coloured glass.

173. Second Species—Harmonies of Contrast.—1. The harmony of contrast of scale, produced by the simultaneous view of two very distant tones of the same scale. 2. The harmony of contrast of hues, produced by the simultaneous view of tones of different depths, belonging, to neighbouring scales. 3. The harmony of contrast of colours, produced by the simultaneous view of colours, belonging to very distant scales, assorted according to the law of contrast. The difference in the depth of the adjacent tones may further augment the contrast of colours.

SECTION IV.

Assortments of
Red, Orange, Yellow, Green, Blue, Violet
with White, Black, and Grey.

174. It will not be useless to the object of this work to introduce some observations relative to the degree of beauty of certain arrangements of the primitive colours with black, white, and grey. But I cannot too strongly insist upon the fact, that they are not given as a rigorous deduction from scientific rules, for they are only the expression of my particular taste; yet I hope that many classes of artists, especially dressmakers, decorators of all kinds, designers of patterns for woven fabrics, paper-hangings, &c., will find advantage in consulting them.

175. The ground, as well as the interval between the colours, having influence upon their effect, all my observations were made with white, black, grey, and coloured circles, ⁴/₁₀ of an inch in diameter, separated by intervals of ⁴/₁₀ of an inch; thirteen circles arranged in a straight line forming a series.

176. The series designed to show the effect of white were on a ground of normal grey; those to show the effect of black and of grey were upon a white ground, slightly tinged with grey. It is necessary to remark that the coloured circles placed apart, were upon black grounds, which must have exercised some influence.

177. The colours which have been under my notice are red, orange, yellow, green, blue, violet. Their differences in regard to brilliancy are so great as to admit of their being divided into two groups, one comprising red, orange, yellow, and bright green, the other blue and violet, which, with the same depth of tone, have not the brilliancy of the former. I shall call the first group luminous colours, and the second sombre colours. But the deep and broken tones of the luminous scales may in many cases be assimilated to the sombre group, as the light tones of blue and violet may sometimes be employed as luminous colours.

Article I.
Colours with White.

Plate 8.

A. Binary Assortments.

178. All the primary colours gain by their juxtaposition with white, but the binary arrangements which result from them are not equally agreeable; and it is to be remarked that the depth of tone of a colour has a great influence upon the effect of its assortment with white.

The binary assortments in the order of their greatest beauty, are as follows:—light blue and white, rose and white, deep yellow and white, bright green and white, violet and white, orange and white.

Dark blue and dark red produce, with white, too strong a contrast of tone to allow of their assortment being as agreeable as that of their light tones. On the contrary, yellow being a light colour, we must take the normal or deepest tone of yellow to produce its most beautiful effect. Dark green and violet contrast too much in tone with white for their combination to be as agreeable as those which are made with the light tones of these colours. The objection which can be made to the combination of orange and white is that of too much brilliancy; yet I should not be surprised to find that many persons preferred it to that of violet and white.

B. Tertiary Assortments of Colours complementary to each other with White.

179. It is to me impossible to establish an order of beauty among binary combinations of primary complementary colours. I shall therefore only describe the effect of white interposed between the binary complementary assortments, or between each of the complementary colours.

180. 1. Red and Green are of all complementary colours the most equal in depth; for red, as regards its brilliancy, is midway between yellow and blue; and in green these two extremes are united. 2. The arrangement, white, red, green, white, &c., is not decidedly superior to the preceding, at least when the colours are not deep. 3. The arrangement white, red, white, green, white, &c., seems to me inferior to the preceding.

181. 1. Blue and Orange are more opposed to each other than red and green, because the least brilliant colour blue, is separated, while the most brilliant are combined in orange. 2. The arrangement, white, orange, blue, white, &c., is agreeable. 3. The arrangement, white, orange, white, blue, white &c., is also agreeable.

182. 1. Yellow and Violet form an arrangement which, as regards depth of tone, is most distinct, since the least intense or lightest colour, the yellow, is separated from the others. Because of this great contrast of tone, the deep, but pure, greenish-yellow combines better with light violet, than light yellow and deep violet. 2. The arrangement, white, yellow, violet, white, &c., appears to me inferior to the preceding arrangement (1). 3. The arrangement, white, yellow, white, violet, white, &c., seems to me inferior to 2.

C. Ternary Assortments of Colours not complementary with White.

183. 1. Red and orange do not accord well. 2. The arrangement, white, red, orange, white, &c., is scarcely preferable. 3. The arrangement, white, red, white, orange, white, &c., is not so bad as the preceding, because white being favourable to all the colours, its interposition between the colours which injure each other, can only produce an advantageous effect.

184. Red and Yellow accord pretty well, especially if the red is purple-red rather than scarlet, and the yellow rather greenish than orange. 2, The arrangement, white, red, yellow, white, is preferable to the preceding. 3. The arrangement white, red, white, yellow, white, is still better.

185. 1. Red and Blue accord passably, especially if the red incline rather to scarlet than to crimson. Deep tones are preferable to light ones. 2. The arrangement white, red, blue, white, &c., is preferable to 1. 3. The arrangement white, red, white, blue, white, is preferable to the second.

186. 1. Red and Violet do not accord well, yet they are found in some natural productions, as the sweet pea. 2. The arrangement white, red, violet, white, is not so bad as the preceding. 3. The arrangement white, red, white, violet, white, is preferable.

187. 1. Orange and Yellow accord incomparably better than red and orange. 2. The arrangement white, orange, yellow, white, is agreeable. 3. The arrangement white, orange, white, yellow, white, is not so good as 2, and perhaps 1, because there is too much white.

188. 1. Orange and Green do not accord well. 2. The arrangement white, orange, green, white, is preferable to 1. 3. The arrangement white, orange, white, green, white, is perhaps preferable to 2.

189. 1. Orange and Violet accord passably, yet not so well as orange and green; the contrast, in the latter case, is greater than in the arrangement orange and violet. 2. The arrangement white, orange, violet, white, &c., is preferable to the preceding. 3. The arrangement white, orange, white, violet, white, &c., is preferable to 2.

190. 1. Yellow and Green form an agreeable combination. 2. The arrangement white, yellow, green, white, &c., is still more agreeable. 3. The arrangement white, yellow, white, green, white, &c., is inferior to the preceding, and perhaps to the first. The inferiority of 3 seems to me to be caused by too much light for the green.

191. 1. Yellow and Blue. The arrangement of yellow and blue is more agreeable than that of yellow and green, but it is less lively. 2. The arrangement white, yellow, blue, white, &c., is perhaps preferable to the preceding. 3. The arrangement white, yellow, white, blue, &c., is perhaps inferior to the preceding.

192. 1. Green and Blue produce an indifferent effect, but better when the colours are deep. 2. The arrangement white, green, blue, white, &c., is preferable. 3. The arrangement white, green, white, blue, white, &c., has a still better effect, because the light is more equally distributed.

193. 1. Green and Violet, especially when light, form a combination preferable to the preceding, green and blue. 2. The arrangement white, green, violet, white, &c., is not decidedly superior to the preceding. 3. The arrangement, white, green, white, violet, white, &c., is not decidedly superior to it.

194. 1. Blue and Violet accord badly. 2. The arrangement, white, blue, violet, white, &c. is scarcely preferable to the preceding (1). 3. The arrangement white, blue, white, violet, white, &c., is not so bad as the preceding (2).

Article II.
Colours and Black.

Plate 9. (Frontispiece.)

195. I do not know whether the use of black for mourning, prevents the use of it, in numberless cases, where it would produce excellent effects; it may be combined most advantageously, not only with sombre colours to produce the harmony of analogy, but also with light and brilliant colours to produce the harmony of contrast.

196. Chinese artists appear to have made excellent use of it, for I have often seen furniture, painting, ornaments, &c., where it has been most judiciously employed. I recommend those artists for whom this paragraph is particularly designed, to attend to the following observations, not doubting that many will be profitable to them.

A. Binary Combinations.

197. No combination of primary colours with black is disagreeable, but there exists among these a generic difference of harmony, which is not shown, at least to nearly the same degree, in the binary combinations of white with the same colours. In fact the brilliancy of white is so predominant, that whatever may be the difference of lightness or brilliancy observed between the various associated colours, there will always be the harmony of contrast, according to what has been said (44-52) of the influence of white in raising the tone and augmenting the intensity of the colour adjacent to it.

198. If the binary combinations of black be examined in this point of view, it will be seen that the deep tones of all the scales, and even of the blue and violet scales (which are not, properly speaking, deep), form with it harmonies of analogy and not of contrast. So likewise do the unbroken tones of the red, orange, yellow-green scales, and the very light tones of the violet and blue scale. We may add, according to what has been said (55), that the combinations of black with sombre colours, such as blue and violet, whose complementaries, orange and greenish-yellow, are luminous, may diminish the contrast of tone if the colours be juxtaposed with black or one not far from it, and in this case the black loses much of its vigour.

199. Blue and Black, Violet and Black, make combinations which may be employed successfully when only dark colours are required. The first is superior to the second.

200. Light combinations which exhibit the harmonies of contrast, appear to me in the order of beauty thus:—

Red or Rose and Black, Orange and Black, Yellow and Black, lastly, Bright Green and Black. As to yellow, I repeat that it must be brilliant and intense, inasmuch as black tends to impoverish its tone.

B.—Ternary Combinations of Colours mutually Complementary with Black.

201. 1. Red, Green, &c. 2. Black, Red, Green, Black, &c. This arrangement being quite different from the former, it is difficult to decide respecting their comparative beauty. 3. Black, Red, Black, Green, Black, &c., appear to me inferior to the preceding, because there is too much black.

202. 1. Blue, Orange, &c. 2. Black, Blue, Orange, Black. I prefer the first to the second; the proportion of dark colour being too strong relatively to the orange. 3. Black, Blue, Black, Orange, Black, &c. This pleases me less than the first.

The effect of black with blue and orange is inferior to that of white.

203. 1. Yellow, Violet, &c. 2. Black, Yellow, Violet, &c. 3. Black, Yellow, Black, Violet, Black, &c. The second is superior to the third, because the proportion of sombre colours with the yellow is too strong in the latter. The first appears to me superior to the second.

C.—Ternary Combinations of Colours not Complementary with Black.

204. 1. Red, Orange, &c. 2. Black, Red, Orange, Black, &c. 3. Black, Red, Black, Orange, Black, &c. As orange and red injure each other, there is an advantage in separating them by black. The third arrangement is preferable to the second; and both are preferable to those in which black is replaced by white.

205. 1. Red, Yellow, &c. 2. Black, Red, Yellow, Black. 3. Black, Red, Black, Yellow, Black, &c. The two last arrangements appear to me superior to the first, and many persons would prefer them to the arrangement in which white replaces black. I cannot too strongly recommend the arrangement 2 and 3 to artists, for whom these observations are especially designed.

206. 1. Red, Blue, &c. 2. Black, Red, Blue, Black. 3. Black, Red, Black, Blue, Black, &c. No. 2 is preferable to No. 3, because there are too many sombre colours in the latter, and because these differ too much from the red. The effect of black upon the binary arrangement, red and blue, is inferior to that of white.

207. 1. Red, Violet, &c. 2. Black, Red, Violet, Black, &c. 3. Black, Red, Black, Violet, Black, &c. The red and violet injure each other, it is therefore advantageous to separate them by black; but the latter does not produce so good an effect as white. It is difficult to say whether No. 3 is preferable to 2; because, if there be in the latter red near violet, this defect is more than compensated in 3, by the predominance of sombre colours over the red.

208. 1. Orange-Yellow, &c. 2. Black, Orange-Yellow, &c. 3. Black, Orange, Black, Yellow, Black, &c. The orange and yellow being very luminous, the black allies itself to them very well in the arrangement 2 and 3; and if the arrangement, white, orange, yellow, white, be preferred to No. 2, I think that in the arrangement 3, the black produces a superior effect to the white.

209. 1. Orange, Green, &c. 2. Black, Orange, Green, &c. 3. Black, Orange, Black, Green, Black, &c. Black combines well with orange, and with bright green. In like manner it combines well with orange and yellow. If, in the arrangement 2, white be preferred to black, I think it cannot be in 3. I recommend to artists the combination of black with the binary arrangements orange and yellow, orange and green.

210. 1. Orange, Violet, &c. 2. Black, Orange, Violet, Black, &c. 3. Black, Orange, Black, Violet, Black, &c. Black does not combine so well as white with orange and violet, because the proportion of dark colours relatively to orange, a very vivid colour, is too great.

211. 1. Yellow, Bright Green, &c. 2. Black, Yellow, Green, Black, &c. 3. Black, Yellow, Black, Green, Black, &c. Yellow and light green being luminous colours, black combines very well with them; and if in the arrangement 2 the effect of white be preferred to that of black, I think that it cannot be in the arrangement 3.

212. 1. Yellow, Blue, &c. 2. Black, Yellow, Blue, Black, &c. 3. Black, Yellow, Black, Blue, Black, &c. If the arrangement 2 be preferable to 3, I think it inferior to 1. The black does not appear to combine so well as the white in the group yellow and blue.

213. 1. Green, Blue, &c. 2. Black, Green, Blue, Black, &c. 3. Black, Green, Black, Blue, Black, &c. Although green and blue do not well accord, yet the combination with black is not decidedly advantageous. Because of the augmentation of the proportion of sombre colours, the white has a superior effect to the black.

214. 1. Green, Violet, &c. 2. Black, Green, Violet, Black. 3. Black, Green, Black, Violet, Black, &c. If the black unite better with green and violet than with green and blue, yet these ternary combinations are inferior to the binary combinations; and inferior to the ternary combination where it is replaced by white.

215. 1. Blue, Violet, &c. 2. Black, Blue, Violet, Black, &c. 3. Black, Blue, Black, Violet, Black, &c.

216. Although blue and violet are colours which do not accord well, and although there is an advantage in separating them, yet it must be remembered that black in isolating them, does not relieve their sombre colour, but, on the other hand, the harmony of the arrangements of 2 and 3 is more agreeable as a harmony of analogy than the harmony of contrast presented by white with the same colours. Thus there are cases in which the assemblage of black, blue, and violet, may be advantageous, when it is desired to produce diversified, but not striking effects.

Article III.
Colours with Grey