and upon the wings of certain butterflies; for, according to the preceding distinction, it is evident that in these natural associations, the effect arising from the addition of the complementaries to each of the two colours which would injure the mat colours is entirely insensible in injuring colours which acquire metallic lustre from the organic structure of the feathers and scales upon which they are found. Finally, I shall add, that it would be necessary, before raising the objection, to demonstrate that the same red associated with green, the same violet associated with yellow, and, lastly, the same blue associated with orange equally glossy, will be less effective than the natural assortments I now take for examples.

Influence of Form taken into Consideration in the Effect of Contrast of two Colours.

(856.) If gloss has so much influence upon the effects of contrast of two juxtaposed colours, the form of the coloured parts presenting them has undoubtedly an influence also: thus the elegance of the form, the arrangements of the parts, their symmetry, the effects of light and shade upon the surfaces independent of all colour, finally the association of ideas which may connect this form with an agreeable recollection, will prevent the perception of the ill effect of two associated colours which are not at all glossy; it is thus, for example, that flowers impress us with associations which, without shocking us, yet, nevertheless, would not produce a good effect if we saw them upon two plain surfaces deprived of gloss. We may instance anew, for example, the flower of the sweet pea, which offers the alliance of red and violet. It cannot be doubted that the red and the violet of this flower, being juxtaposed-the red to green and the violet to yellow -do not produce a finer effect than that resulting from their association in the flower mentioned above.

(857.) The relation of the preceding facts to criticism is easily perceived when we proceed to judge of the association

of two colours in themselves, or to compare together different binary associations.

In the first case we can inquire if the association an artist has made of two colours has attained the end he proposed,— that of improving both, or that of improving one and sacrificing the other.

In the second case we can compare together the effects of one colour of the same class-reds, for example—each making part of a binary association, we can compare together the effects of binary associations of different colours, always under the optical point of view, and with the intention, if these associations are the product of art, of judging the artist who has made them. Then the critic must be directed by the considerations which are summed up as follows:

1o. The kind of association: the greater the difference between the colours, the more they mutually beautify each other; and inversely, the less difference there is, the more they will tend to injure one another (845.-851.).

2o. The equality in height of tone (852.).

3°. The difference of tone, the one being deep, the other light (853.).

4°. The glossiness of the surfaces which sends them to the eye (855.).

5°. The form of the body of which these surfaces limit the extent (856.).

CHAPTER II.

OF THE COMPLEX ASSOCIATIONS OF COLOURS REVIEWED CRITICALLY.

(858.) Ir is evident that the rules prescribed for judging a colour and the associations of two colours in an absolute manner, must serve for judging under the same relation the colours of an association, however complex it may be, when found in a picture, tapestry, carpet, stained-glass

window, or in the decoration of a theatre, apartment, &c.; but to view the ensemble, we must proceed conformably to the distinctions we have established of the harmonies of analogy and the harmonies of contrast; for, otherwise, it would not be possible to express clearly a sound judgment concerning the specialities of the constituent assortments and the general effect of the ensemble of all these assortments; besides, before expressing this judgment, we must know how to see colours independently of all form or pattern; in a word, independently of all which is not colour, even when it is necessary in a picture.

(859.) In this manner we can see a complex association of colours, such as are presented to us by a picture, for instance, thus taking the most complicated case. We shall consider the masses of colours which are upon the same plane, the extent which each occupies, and the harmony that unites them together. On submitting to a similar examination the colours on the other planes, we then can look at the colours in passing from those of the first plane to the colours of the latter. The critic who is well satisfied with seeing clearly at the same time only a very small number of the objects that a picture presents to him (748., 483.), and who is also accustomed to examine a coloured composition in the manner I have described, is, relatively to things upon which he successively concentrates his attention, in the position of a person who reads in succession three kinds of writing traced on the same side of a sheet of paper; one of them composed of lines across the width of the paper, the other composed of lines crossing the first at right angles, and the third composed of lines running diagonally across the paper. After giving himself up to this examination, the critic must renew the ensemble of the picture under the relation of its colours, and then, having fixed upon their particular and general associations, he will be in a condition to penetrate the thought of the painter, and of seeing if he has employed the most suitable harmonies to express it; but this is not the place to treat this subject: it belongs to the following Chapter (865., &c.). I will only remark, that if it is easier to form with opposed colours than with neighbouring colours, binary assortments favourable to the

associated colours in a composition where a great number of pure and brilliant colours are employed, it is more difficult to harmonise these latter in allying them to one another, than if we proceeded with a small number of colours, which would add only to the harmonies of analogy or of contrast of scale or of hues.

CHAPTER III.

OF THE TWOFOLD INFLUENCE PRESENTED UNDER THE CRITICAL POINT OF VIEW WHICH THE PHYSICAL CONDITION OF THE COLOURED MATERIALS EMPLOYED IN VARIOUS ARTS, AND THE SPECIALITY OF THESE ARTS, EXERCISE UPON THE PARTICULAR PRODUCTS OF EACH OF THEM.

(860.) I HAVE so deep a conviction that the greatest artists cannot free themselves from certain rules without compromising the art itself, that I believe it useful to insist upon everything which can substantiate my opinion. Such is the motive with which I refer to the results that naturally flow from the physical state of the coloured materials employed in various arts, and of the special object of these arts; results of which I have already shown the importance in the case where we judge if certain innovations have the merit or advantages which their authors promised for them. It is under this connection that I now consider the arts of painting which employ coloured materials in a state of infinite division (so to speak), and the arts which employ them of a certain size, as the threads of the tapestry-weaver, the pieces of the mosaic worker, &c.

§ 1.

OF THE ARTS OF PAINTING WITH COLOURED MATERIALS IN A STATE OF SO CALLED INFINITE DIVISION, CONSIDERED RELATIVELY TO THE PHYSICAL STATE OF THESE MATERIALS AND THE SPECIALITY OF THE ART EMPLOYING THEM.

(861.) In the first place, I must explain the meaning of the expression in a state of infinite division, applied to coloured materials employed by painters: in reality, the divi

sion of these materials is not infinite, it is not even carried to the point attainable by mechanical means. If it were possible to perceive them in a painting by means of magnifying glasses, we should then see that a coloured surface, which appears of a uniform colour to the naked eye, is composed of distinct coloured particles, disposed in parallel or concentric lines, or in spots, according to the handling of the pencil. By these means we could distinguish in oil-paintings some parts which would resemble an enamel, because they would contain so many opaque particles which the drying oil does not make transparent, while the other parts would resemble a coloured glass, because the oily vehicle does not contain sufficient opaque particles to be entirely deprived of its transparency.

I shall now consider, successively, painting in chiar'oscuro

and in flat tints.

ARTICLE 1.

Painting in Chiar'oscuro.

(862.) In commencing with the fact that the pigments of the painter appear to be in a state of infinite division, we come to see clearly the possibility of tracing lines as fine as it is possible to make them, by means of a pencil filled with a fluid charged more or less with colouring materials, of intimately mixing these pigments together, so as to blend them with each other. From this state of things we deduce the possibility of making a perfect delineation of the different parts of objects of which the painter wishes to reproduce the image, and represent exactly all the modifications of light which exist in the model.

(863.) I now recal how satisfactory the study has been which we have made of the modifications under which bodies appear, when they are rendered visible by the direct light of the sun, or by diffused daylight, in enabling us to judge if these modifications have been faithfully reproduced by the painter in a given work. I must refer the reader of this subject to the division of this book concerning painting.

(864.) From the perfection of the drawing, and the gradation of white and coloured light, result the perfection of