to observe this effect, it is necessary that the grey part reflects to the eye white light, and little or none of the coloured light which the object naturally reflects.

(a.) 4th Modification in a single-coloured stuff.

(287.) For example, if the eye is directed from the back. of a chamber towards a window which admits daylight, and a person clothed in a new blue coat, dyed with indigo or Prussian blue, looks through the window at objects outside (Plate III. fig. 3.), the eye will see the part b of the coat different from the part a, because the nap of the cloth is disposed in a contrary direction to b in a : a appears of a fine blue, while 6 will be of an orange-grey, by the effect of contrast of the blue part with a part that reflects very little white light to the eye, without, or almost without, blue light.

Besides, the pile of the nap loses its regular position; in fact, as it wears, the cloth becomes dull and tarnished, the coloured light is reflected irregularly from all points, and if the effect is not absolutely destroyed, it is at least much weakened.

If the garment be of a deep green, the grey part will appear reddish: if it be of a violet-marron or claret, the grey part will appear yellow.

(288.) The complementary is only developed upon cloths of dark and sombre colours; thus, red, scarlet, orange, yellow, and light-blue garments, do not exhibit it, because they have always too much of the essential colour which is reflected. The modification is limited to that where one of the parts is more strongly illuminated by diffused light than the other (279.).

(289.) It is superfluous to remark that in a drapery where the pile lies all in one direction, but which exhibits folds, these, in changing the direction of the nap, may determine the modification presented by blue and dark-green clothes, as also by violet-marron.

(290.) There is still another circumstance where the fourth modification will appear evident; it is when we look upon a series of light tones, blue, rose, &c. (belonging to the same scale), of a skein of silk or wool placed upon an easel in such

manner that one half of the same skein presents to the eye the threads disposed in a contrary direction to those of the other half. The half of the skein which does not reflect coloured light to the eye appears tinted with the complementary of the other half which does reflect it.

(b.) 4th Modification in a stuff presenting a deep and a light tone belonging to the same scale.

(291.) If we place in juxtaposition a deep tone and a light tone of the same scale well assorted, the light tone will appear of the colour complementary to the scale to which it belongs. This modification is too important in the explanation of certain phenomena often exhibited by the products of the calico-printer to permit of my passing it over hastily.

(292.) When we look for several seconds on a fabric dyed with a coloured ground, and on which we put patterns intended to be white, but which, owing to the imperfection of the process employed, retain a light tone of the colour of the ground, the patterns will appear of the complementary to this latter. Thus, upon a ground of yellow chromate of lead, they will appear violet; upon a ground of orange chrome they will appear blue; rose upon a green ground, &c.: to dispel the illusion, and recognise the true tint of the pattern, it is only necessary to cover the ground with paper perforated with the design of the pattern, which then permits us to see only the pattern coloured like the ground. The influence of the dark tone upon a feeble tone is then such that not only is the latter neutralised, but also the place it occupies upon the cloth appears tinted with its complementary colour.

(293.) From the preceding observations, it may be deduced that we may have a printed cotton, the design of which, although coloured, will appear to most eyes white, and not of the complementary of the ground. For those eyes which see it thus, the perception of the phenomenon of contrast will correct the imperfection of the art of the calicoprinter.

(294.) In the lectures upon Contrast which I gave in 1836 at the Gobelins, I remarked that, in applying paper (cut for the purpose) upon the lights of a blue drapery of the Virgin, in a tapestry representing the Holy Family, after Raphael,

we saw them of light blue, although, when they were seen surrounded with deeper blue tones, they appeared of an orange tint.

(295.) In conclusion, the fourth modification is observed: 1. Every time that a monochromous object of a dark and not very vivid colour is seen in such manner that one portion reflects to the eye its peculiar colour, while the other portion reflects only a feeble light scarcely coloured.

2. Every time that a stuff presents two suitably distant tones of the same colour.

(296.) We can conceive, without difficulty, that if the modification is not manifested with monochromous objects of vivid colours, as yellow, scarlet, &c., it is because that part of the surface of these objects which reflects the least light to the eye reflects, nevertheless, always sufficient of its peculiar colour to neutralise the complementary which the coloured light of the illuminated portion tends to develop. If I am not deceived, I believe that this effect tends to enfeeble the coloured light of the shaded part.

(297.) Although in this chapter I do not propose to treat of the modifications presented by coloured stuffs with white patterns, yet, as it is a case which belongs to the developments in which I am engaged, I cannot avoid mentioning them in this place. If we observe a sky-blue silk with white flowers, the weft of which is in an opposite direction to the weft of the blue ground, we shall see the flowers white, if they are placed in the most favourable manner to receive the white light reflected by them; while, in the contrary position, we shall see these flowers absolutely orange. There is still much white light reflected; but it is not sufficiently vivid to neutralise the development of the complementary of the ground.

(298.) In painting, we recognise two kinds of perspective, the linear and the aerial.

The first is the art of producing, upon a plane surface, the outlines and contours of objects and their various parts in the relations of position and size in which the eye perceives them.

The second is the art of distributing, in a painted imitation, light and shade, as the eye of the painter perceives them in objects placed in different planes, and in each particular object which he wishes to imitate upon a surface.

It is evident that aerial perspective comprehends the observation and reproduction of the principal modifications of colours which I have just examined in succession, and that true and absolute colouring in painting can only be as faithful a reproduction of this as possible.

CHAPTER II.

ON THE DIFFERENCE EXISTING BETWEEN A COLOURED OBJECT AND THE IMITATION OF IT MADE BY A PAINTER, WHEN THE SPECTATOR OBSERVES IT FROM A DIFFERENT POINT OF VIEW FROM HIS.

(299.) THERE is a vast difference between the most perfect imitation of a coloured object and the object itself, upon which we will now dwell a moment, because it is not sufficiently appreciated. Imitation is true, relative to delineation, the distribution of light and shade, and all the resulting modifications of colour, only in the same position where the painter stood in regard to his model: for out of this position, everything relatively to the spectator varies, more or less; while in the imitation he sees the light, shadows, and lines which circumscribe them, and the modifications of colour, constantly in the same manner, whatever be the point of sight.

For example: a spectator who, in a room opposite a window, observes the back of a person in a new blue coat (Pl. III. fig. 3.), which is placed between him and the window, the part a of the coat is blue, and the part 6 an orangegrey. Let the spectator advance in such manner as to see the profile of this person; if, then, he looks at the parts a and 6, they will appear to him different from what they were before he changed his position.*

*There is a position where the spectator will see the part a of an orangegrey, and the part b of a bright blue.

If the painter had painted the coat from the point of view where the spectator was at first, he would have coloured the part a of a bright blue, and the part 6 probably of an orange

grey.

If, now, the spectator regards the imitation of the coat from the second position, when he saw the person in profile, he will still see the part a blue, and the part b of an orangegrey, although in this position the model coat no longer pre

sents these modifications.

Besides, I must insist upon demonstrating a thing which is really very simple.

In fine, because lights, shades, and modifications of colours, and the outlines which circumscribe each part, preserve invariably the same relation in a picture made upon a plane, it evidently follows that this imitation produces the same impression, although we observe it from very different points from that where the painter was placed to represent his model.

(300.) It is for this reason, also, that a person who looks at the painter while he is painting his portrait on the canvas, appears in this portrait to look at the spectator, whatever may be the position of the latter with regard to the picture.