(519.) I have not yet touched upon an element appertaining to the subject that now occupies us, which seems to me worthy of consideration. I allude to a property possessed in variable degrees by colours,-viz.: that of leaving upon the organ which has perceived them during a certain time the impression of their respective complementaries (116.). It is clear that the more durable this impression is, other things being equal, the less will the organ be disposed to receive distinctly new impressions, for there must necessarily be superpositions of different images, as in the mixed contrast (327.), which, not being coincident, will tend to render the actual effect less marked than it might otherwise be.

CHAPTER III.

ON THE ASSORTMENT OF COLOURS WITH RESPECT TO READING BY
ARTIFICIAL LIGHT.

(520.) I HAVE made but very few observations on the subject of this chapter. Yet, I believe, I am right in affirming that by diffused daylight I read for some minutes black letters printed on yellow paper more easily than black letters printed upon pale yellow-green paper, while with the light of a lamp the contrary took place.

FOURTH DIVISION.

FLAT TINTING.

FIRST SECTION.-MAP COLOURING.

SECOND SECTION.-COLOURING ENGRAVINGS.

SECTION I.

ON COLOURING GEOGRAPHICAL MAPS.

(521.) THE colouring of Geographical Charts, as is well known, gives many advantages in presenting readily to the eyes their different component parts, whether continent, empire, kingdom, or republic; state or country. Until lately, the colouring of maps has always depended upon the caprice of the colourer; yet it appears to me there are some rules which it would not be useless to observe.

(522.) Firstly, the colours should be as pale as possible, especially those which are naturally sombre, as blue and violet, so that the reading of the names may always be easy; but preference must be given to the luminous colours, red, orange, yellow, and light green, and to employing only their bright tones.

(523.) Secondly, all the parts which have some common relation together should receive a single colour, each part being distinguished from the contiguous parts by a difference of tone.

(524.) To attain this end it is not necessary to employ as many different tones as there are parts to be distinguished; it suffices that we can perceive the difference of tone in the contiguous parts without difficulty.

Suppose, for example, there are thirteen divisions in a map, which, although very small, are distinguished one from the other by means of five tones of a single scale. In case we found that certain tones would be too near each other, we could give them an extremely light tint of the colour of the nearest scale.

For instance, if we found that the tone 2, which is near the tone 1, was not sufficiently distinct from it, we could, if the colour was rose, give it a tint of crimson-red.

(525.) If we had a surface contiguous to the preceding, we would have to choose the complementary of the first colour; if we had a second surface, we would have to take a colour distinct from the other two. For example, supposing we had pink and green, we take yellow for the second.

(526.) We could proceed in an analogous manner when we would represent the currents of the ocean; that is, have recourse to the tones of a blue scale, which is the colour usually devoted to water.