When the eye is fatigued by looking at one of the brilliant primary colours it seems to relieve itself by seeing the secondary colour which is complementary to it: thus if we have looked at red—say at a red light—for some time, and turn the eye away we shall see a green one of the same size and form, being seen in fact by that portion of the eye only that has been fatigued by the strong impression of the red. If we have looked at a yellow light we shall behold a purple image, because this is composed of the other two colours; blue and red is complementary to yellow. Blue, being a cool colour, does not so much fatigue the eye; hence, though by the foregoing rule we ought to see its image in bright orange, in practice we rarely do so. From these facts we learn that, whenever one of the primary colours is used in a picture, the complementary colour formed of the other two ought to be placed not very far from it, so as agreeably to relieve the eye. In most cases the landscape actually being copied will afford sufficient facility for this; sometimes it will not; but these instances are exceptional, and probably will occur in the snow wastes of the north, on the solitude of the ocean, or in the sandy deserts of the tropics, where drear monotony or wild and terrific grandeur constitute the charm of the picture, and fidelity rather than pleasing composition must be the artist’s aim.
We subjoin also a few combinations of colour which will be found useful in landscape painting: Aerial tints for skies, clouds, and distance: for very delicate preparatory wash, cadmium yellow and rose madder; strong ditto, Indian yellow and carmine; neutral ditto, yellow ochre and brown madder; still darker and less aerial, light red, Venetian red, or Indian red. Cobalt for delicate blue skies; French ultramarine for stronger. Cobalt and rose madder for delicate cloud tints. For golden sunsets: aureolin, gamboge, lemon yellow, cadmium yellow, Indian yellow, yellow ochre—to be used according to the brilliancy or depth of colour required, and to be contrasted by cool greys composed of cobalt and rose madder, or French ultramarine. Crimson sunsets: rose madder, carmine, crimson lake, Indian red, purple madder, contrasted with cool grey; and sometimes greenish tones formed by adding a little yellow to the blue and red.
In dark storm clouds, French blue or indigo, with light red, Venetian red, or Indian red, or purple madder. Indigo and Prussian blue require great caution in their use. With any of the beforementioned yellows, they form rich greens for sea tints or foliage; with raw sienna or burnt sienna, they give very deep greens for stormy seas or heavy forest trees. Light red and Prussian blue give a greenish grey. Light red with cobalt or ultramarine give greys somewhat less aerial for middle distance, mountains, &c. Indian red, with the same blues, gives a more opaque grey. Sepia and French blue make a cool grey; raw and burnt sienna are good colours for autumnal foliage, stems of trees where grey is not required, rocks, Dutch galliots, and many other foreground objects; brown madder and vandyke brown afford great depth for foreground shadows. For native complexions, raw sienna with a little of the burnt will give the colour of a Hottentot. A Kafir requires burnt sienna in the half tints, vandyke brown in the shadows, and a cool grey like the reflection of the sky in the lights—this is best produced by a little Chinese white and cobalt laid on thin over the brown. There may be a little blue put into the black of the hair to contrast the better with the brown skin. For a Negro, crimson lake and blue-black; the lights made with Chinese white and blue as before.
CHAPTER XX.
THE ESTIMATION OF DISTANCES AND HINTS ON FIELD OBSERVING.
On measuring the Distance travelled by Wheeled Carriages.
When no instrument for this purpose is obtainable, the best plan we know of is that adopted by the late Dr. Burchell, the eminent South African traveller, and after him by Captain Cornwallis Harris, the explorer and naturalist, in the more distant parts of the same country; and this is, to measure the large wheel carefully, to mark one of its spokes, and count its revolutions during any given time, say a minute, and then convert the result into miles or parts of a mile per hour. Thus, if a wheel be 5yds. in circumference, and it makes six revolutions in a minute, the distance in that time will be 30yds., or 1800yds., i.e., 40yds. more than a mile per hour; twelve revolutions will of course be 80yds. over two miles; and, during former journeys, when our wheel was making eighteen revolutions, we used to reckon the waggon was going, allowing for occasional unavoidable stoppages, two and a half miles per hour. With a watch having a second hand it is easy to note any fraction of time, but with one not so provided less than a minute cannot easily be estimated. After a little practice we became so accustomed to this that we seldom used a watch; but when sitting on the waggon-box would just look over the side, and estimate the rate at which the wheel was going, just as a sailor would in like manner make a very fair estimate of the speed of his ship.
It will generally be found that an African ox-waggon, not overloaded, and on tolerably fair ground, travels about two and a half miles an hour; and we have also found that with pack horses in Australia, if the same rate is assumed, the resulting measurement of the day’s work will be very nearly correct.
We tried once to make a trocheameter, but at the time had never either seen one or read a description of it, and therefore the principle cost us some thinking out. It was perfectly evident that, for motive power, an axle so weighted that it could not revolve in a revolving box would produce the same effect upon the works as an axle made to revolve, by weights or otherwise, in a fixed box would have on those of a clock. We therefore made a box of such a form as to fit between spokes of the hinder wheels of a waggon, and in it fitted an axle with a heavy plummet, so fixed to it as to prevent its turning when the box revolved; on this axle was one tooth fitting into the cogs of a sixty-toothed wheel, which therefore moved one tooth for every revolution, or once round for every sixty; the axle of this had also one tooth acting on another of sixty teeth, so the two were capable of registering sixty times sixty, or three thousand six hundred revolutions, which, supposing the wheel to be only 5yds. in circumference, would measure ten miles and a quarter, the number of revolutions being indicated by a hand fixed upon the axle of each wheel, each moving on its own dial-plate, like those of a patent log. We found that the machine answered quite well enough to convince us that we were right in principle, and to make us regret that we had not the tools and appliances at hand to fit it so perfectly as to insure smoothness and uniformity of action.