White lead, smalt and cobalt-blue are not affected by light, but are by foul air. The last two are considered permanent in watercolour painting.

According to Field, the tints of the following pigments are not affected by mixture with lime, consequently they are adapted for use in fresco-painting:

As the effect of light on pigments is a matter of considerable importance to artists, particularly to those working with the thin washes used in water-colour painting, a careful experiment on this

matter was made by the present writer. The washes laid on ordinary drawing-paper were exposed during the summer to sunlight for more than three months and a half, and the effects noted; these were as follows:

WATER-COLOUR PIGMENTS THAT ARE NOT AFFECTED BY LIGHT:

The following pigments were all more or less affected; those that were very little changed head the list, which is arranged so as to indicate the relative amounts of damage suffered, the most fugitive colours being placed at its end:

Chrome-yellow becomes slightly greenish.
Red lead becomes slightly less orange.

Naples yellow becomes slightly greenish brown.
Raw Sienna fades slightly; becomes more yellowish.
Vermilion becomes darker and brownish.

Aureoline fades slightly.

Indian yellow fades slightly.

Antwerp blue fades slightly.

Emerald green fades slightly.

Olive green fades slightly, becomes more brownish.
Rose madder fades slightly, becomes more purplish.
Sepia fades slightly.

Prussian blue fades somewhat.

Hooker's green becomes more bluish.

Gamboge fades and becomes more grey.

Bistre fades and becomes more grey.

Burnt madder fades somewhat.

Neutral tint fades somewhat.

Vandyck brown fades and becomes more grey.

Indigo fades somewhat.

Brown pink fades greatly.

Violet carmine fades greatly, becomes brownish.

Yellow lake fades greatly, becomes brownish.

Crimson lake fades out.

Carmine fades out.

To this we may add that rose madder, burnt or brown madder, and purple madder, all, are a little affected by an exposure to sunlight for seventy hours. Pale washes of the following pigments were completely faded out by a much shorter exposure to sunlight:

Carmine,
Full red,
Dragon's blood,

Yellow lake,
Gall-stone,
Brown pink,

Italian pink,

Violet carmine.

CHAPTER VIII.

ON THE ABNORMAL PERCEPTION OF COLOUR, AND ON COLOUR-BLINDNESS.

We have considered now, with some detail, the various ordinary modes of producing the sensation of colour; but, in order to render our account more complete, it is necessary to touch on some of the unusual or extraordinary methods. In every case examined thus far, the sensation of colour was generated by the action on the eye of coloured light—that is, of waves of light having practically a definite length. As colour, however, is only a sensation, and has no existence apart from the nervous organization of living beings, it may not seem strange to find that it can be produced by white as well as by coloured light, or even that it can be developed in total darkness, without the agency of light of any kind whatever. If the eyes be directed for a few moments toward a sheet of white paper placed on a black background and illuminated by sunlight, on closing them and excluding all light by the hands or otherwise, it will be found that the absence of the light does not at once cause the image of the paper to disappear. After the eyes are closed it will still be plainly visible for several seconds, and will at first be seen quite correctly, as a white object on a black ground; the colour with some observers then changes to blue, green, red, and finally back to blue, the background remaining all the while black. After this first stage the background changes to white, the colour of the sheet of paper appearing blue-green, and finally yellow. Most of these colours are

as distinct and decided as those of natural objects. If the experiment be made for a shorter time, and under a less brilliant illumination, the eyes being first well rested by prolonged closure, the series of colours will be somewhat different. Fechner, Seguin, and Helmholtz observed that the original white colour passed rapidly through a greenish blue (Seguin, green) into a beautiful indigo-blue; this afterward changed into a violet or rose tint. These colours were bright and clear, afterward followed a dirty or grey orange; during the presence of this colour the background changed from black to white, and the orange tint altered often into a dirty yellow-green which completed the series. If, instead of employing white, a coloured object on a grey ground is regarded intently for some time, the eyes will be so affected that, on suddenly removing the coloured object, the grey ground will appear tinged with a complementary

tint; for example, if the object be red, the after-image will be bluish green. It is not necessary to dwell longer on these phenomena at present, as a portion of Chapter XV. will be especially devoted to them. In both the cases men

tioned above, the colour develops itself after the eyes are closed, or at least withdrawn from the illuminated surface. There are, however, cases where very vivid colours can be seen while the eyes are exposed to full daylight. If a disk