SPECTRUM. Suppose a small hole of any shape made in the window shutter of a darkened room, and that sunshine be ad mitted through it, and the light received upon a white screen, placed perpendicularly to the line joining it and the hole ; then, a round image of the sun will be formed upon the screen, the diameter of the image depending upon the distance of the screen from the hole. Observe that whatever may be the shape of the hole, the image of the sun will be round. A triangular hole would not give a triangular image, nor a square hole a square image ; the sun being round, its image is round; for the image is not produced by a ray of sunshine which enters through the hole and falls upon the screen, producing a spot of light the same shape and size as the hole, but by pencils of light which diverge from every part of the sun, and after crossing each other in passing through the hole, proceed till they reach the screen, where they form a round image of the sun, the size of which increases as the distance of the screen from the sole increases. Thus, the sun being about half a degree in angular meter, if the screen be placed 10 feet from the hole, the sun's ige will be about 1 inch in diameter; if 20 feet from the hole, aches in diameter, and so on.
This being understood, let a glass prism be placed with its edge ' • behind the hole. Then, since white light is not homo geneous, the pencils will all be decomposed by refraction through the prism into pencils of the different coloured lights of which white light is composed ; so that the screen, instead of receiving a single round image of the sun in white light, will receive upon a different part of it as many different coloured round images of the sun as there are different kinds of light in white light separable by refraction. These images will partly overlap one upon the other, and produce a long image of the sun, having belts across it of differ ent colours, armnged in the following order, (if the refracting angle of the prism be suitably taken,) viz., red, omnge, yellow, green, blue, indigo, violet ; which colours are called the " prismatic colours," and the entire coloured image the " prismatic spectrum." Should, however, the refracting angle of the prism be too small, there will be a space of white light in the centre of the spectrum, produced by the coincidence of a portion of each of the coloured images.
The reader will perhaps find this account of the way in which the prismatic spectrum is generally produced somewhat different from the accounts given of it in popular treatises on Optics ; these popular explanations generally proceeding on the assumption that the light which is admitted through the hole is a single beam of light,—which is manifestly incorrect. Every photographer knows that an image is formed of external objects by light admitted through a small hole in the front of a dark box, and received upon a focussing screen ; and that this is true, however small the hole may be and whatever its shape. We have, therefore, to deal with an image of the sun, and not with a single my of white light.
It appears, then, that when the spectrum is formed by admitting the light through a hole, however small, the bands of different colours contain an admixture of lights of different refrangibilities. In order to obviate this evil, M. Fraunhofer admitted the light through a long and extremely fine slit, instead of a hole, and placed the prism with its edge parallel to the slit, and at a con.siderable &dance from it. But even this arrangement is not sufficiently exact, for the additional precaution must be taken of covering the prism with an opaque diaphragm, having an extremely fine slit parallel to the edge of the prism and therefore parallel to the other slit, so that the light from the first slit may also pass through the second slit. In this way the spectrum is rendered nearly pure, and the different parts of it free from the a.dmixtu/ e of other colours. And here it is important to observe that in the experiments of Sir David Brewster, in which he detected white light in every part of the prism in a state of admixture with his supposed simple colours, red, blue, and yellow, allowance had not been made by him for the impurity of his spectrum, so that his conclusions that Newton's theory was wrong and that the seven colours of the spectrum may be reduced to three was founded on experiments conducted in ignorance apparently of an elementary principle in geometrical optics.