prism, the angles of the two prisms may be made such, that when the prisms are placed close together with their edges turned opposite ways, they will exactly oppose each other's action, and will refract the coloured rays equally but in contrary directions, so that an exact compensation will be effected, and the light will be refracted without colour.1 The achromatic telescope is constructed on this principle. It consists of a tube with an object-glass or lens at one end to bring the rays to a focus and form an image of the distant object, and a magnifying glass at the other end to view the image thus formed. Now it is found that the object-glass, instead of making the rays converge to one point, disperses them, and gives a confused and coloured image: but by constructing it of two lenses in contact, one of flint and the other of crown glass of certain forms and proportions, the dispersion is counteracted, and a perfectly well-defined and colourless image of the object is formed.2 It was thought to be impossible to produce refraction without colour, till Mr. Hall, a gentleman of Worcestershire, constructed a telescope on this principle in the year 1733; and twenty-five years afterwards, the achromatic telescope was brought to perfection by Mr. Dollond, a celebrated optician in London.

A perfectly homogeneous colour is very rarely to be found, but the tints of all substances are most brilliant when viewed in light of their own colour. The red of a

wafer is much more vivid in red than in white light; whereas if placed in homogeneous yellow light, it can no longer appear red, because there is not a ray of red in the yellow light. Were it not that the wafer, like all other bodies, whether coloured or not, reflects white light at its outer surface, it would appear absolutely black when placed in yellow light.

After looking steadily for a short time at a coloured ob

ject, such as a red wafer, on turning the eyes to a white substance, a green image of the wafer appears, which is called the accidental colour of red. All tints have their accidental colours: thus the accidental colour of orange

is blue; that of yellow is indigo; of green, reddish-white; of blue, orange-red; of violet, yellow; and of white, black; and vice versa. When the direct and accidental colours are of the same intensity, the accidental is then called the complementary colour, because any two colours are said to be complementary to one another which produce white when combined.

From recent experiments by M. Plateau of Brussels, it appears that two complementary colours from direct impression, which would produce white when combined, produce black, or extinguish one another by their union when accidental; and also that the combination of all the tints of the solar spectrum produces white light if they be from a direct impression on the eye, whereas blackness results from a union of the same tints if they be accidental; and in every case where the real colours produce white by their combination, the accidental colours of the same tints produce black. When the image of an object is impressed on the retina only for a few moments, the picture left is exactly of the same colour with the object, but in an extremely short time the picture is succeeded by the accidental image. M. Plateau attributes this phenomenon to a re-action of the retina after being excited by direct vision, so that the accidental impression is of an opposite nature to the corresponding direct impression. He conceives, that when the eye is excited by being fixed for a time on a coloured object, and then withdrawn from the excitement, that it endeavours to return to its state of repose, but in so doing that it passes this point and spontaneously assumes an opposite condition, like a spring which, bent in one direction, in returning to its state of

rest bends as much the contrary way. The accidental image thus results from a particular modification of the organ of sight, in virtue of which it spontaneously gives us a new sensation after it has been excited by direct vision. If the prevailing impression be a very strong white light, its accidental image is not black, but a variety of colours in succession. According to M. Plateau, the retina offers a resistance to the action of light, which increases with the duration of this action; whence after looking intently at an object for a long time it appears to decrease in brilliancy. The imagination has a powerful influence on our optical impressions, and has been known to revive the images of highly luminous objects months, and even years, afterwards.

SECTION XX.

INTERFERENCE OF LIGHT. UNDULATORY THEORY OF LIGHT. PROPAGATION OF LIGHT. -NEWTON'S RINGS. MEASUREMENT OF THE LENGTH OF THE WAVES OF LIGHT, AND OF THE FREQUENCY OF THE VIBRATIONS OF ETHER FOR EACH COLOUR. NEWTON'S SCALE OF COLOURS. DIFFRACTION OF LIGHT. SIR JOHN HERSCHEL'S THEORY OF THE ABSORPTION OF LIGHT.-REFRACTION AND REFLECTION OF LIGHT.

NEWTON and most of his immediate successors imagined light to be a material substance, emitted by all self-luminous bodies in extremely minute particles, moving in straight lines with prodigious velocity, which, by impinging upon the optic nerves, produce the sensation of light. Many of the observed phenomena have been explained by this theory; it is, however, totally inadequate to account for the following circumstances.

When two equal rays of red light, proceeding from two luminous points, fall upon a sheet of white paper in a dark room, they produce a red spot on it, which will be twice as bright as either ray would produce singly, provided the difference in the lengths of the two beams, from the luminous points to the red spot on the paper, be exactly the 0.0000258th part of an inch. The same effect will take place if the difference in the lengths be twice, three times, four times, &c. that quantity. But if the difference in the lengths of the two rays be equal to one half of the 0.0000258th part of an inch, or to its 1, 2, 31, &c. part, the one light will entirely extinguish the other, and will produce absolute darkness on the paper where the united beams fall. If the difference in the lengths of their paths be equal to the 14, 24, 31, &c. of the 0·0000258th

part of an inch, the red spot arising from the combined beams will be of the same intensity which one alone would produce. If violet light be employed, the difference in the lengths of the two beams must be equal to the 0·0000157th part of an inch, in order to produce the same phenomena; and for the other colours, the difference must be intermediate between the 0·0000258th and the 0·0000157th part of an inch. Similar phenomena may be seen by viewing the flame of a candle through two very fine slits in a card extremely near to one another1; or by admitting the sun's light into a dark room through a pin-hole about the fortieth of an inch in diameter, receiving the image on a sheet of white paper, and holding a slender wire in the light. Its shadow will be found to consist of a bright white bar or stripe in the middle, with a series of alternate black and brightly coloured stripes on each side. The rays which bend round the wire in two streams are of equal lengths in the middle stripe; it is consequently doubly bright from their combined effect; but the rays which fall on the paper on each side of the bright stripe, being of such unequal lengths as to destroy one another, form black lines. On each side of these black lines the rays are again of such lengths as to combine to form bright stripes, and so on alternately, till the light is too faint to be visible. When any homogeneous light is used, such as red, the alternations are only black and red; but on account of the heterogeneous nature of white light, the black lines alternate with vivid stripes or fringes of prismatic colours, arising from the superposition of systems of alternate black lines and lines of each homoThat the alternation of black lines and geneous colour. coloured fringes actually does arise from the mixture of the two streams of light which flow round the wire, is proved by their vanishing the instant one of the streams is interrupted. It may therefore be concluded, as often as