46 PROCESSES FOR PURIFYING THE SPECTRUM. 



due to the different refrangibility of those different coloured rays. Let, therefore, in 

 fig. 116, r o y g b~i v be a series of such circular images, all conjointly composing the 

 spectrum E D, arranged in their proper order, r being the red image, and v the violet ; 

 if we take any one of these images, such as y, and consider its constitution, we shall 

 find that, by reason of the overlapping of the adjacent images, there are rays in it 

 which belong to the images on either side, and instead of the circle y consisting of 

 yellow light only, it includes orange light derived from the circle o, and green light 

 derived from the circle g, as is obvious from the overlapping of those circles o and g, 

 upon the circle under consideration, y. 



155. Farther, an ordinary spectrum without a penumbra differs in constitution in its 

 various parts. If a line, E D, be drawn through its centre longitudinally, the over- 

 lapping of the successive circles along that line takes place to the greatest extent ; but 

 if another line, m n, be drawn along its edge, on that line the successive circles do not 

 overlap, and, therefore, while along the axis of the spectrum, E D, the intermixture of 

 the rays and of the colours is at a maximum, along the edges of the spectrum there is 

 no overlapping, no intermixture, and each one of the coloured' rays is in a state of 

 purity. The light of such a spectrum, therefore, becomes more and more homogene- 

 ous as we pass from the axis and go to the outer edge. 



156. But suppose the total length of the spectrum remaining the same, those circu- 

 lar images are diminished in diameter, as in Jig. 117. Here, through that diminution 

 of diameter, overlapping is prevented, and each one of the rays is separated out for itself, 

 there is no superposition, and the colours are equally homogeneous in the axis of the 

 spectrum and on its edges. 



157. This diminution in the magnitude of the circular images may be produced in 

 practice in several different ways. The one which NEWTON recommends, and which 

 is generally adopted, is to use the aperture in the shutter as the radiant source, instead 

 of the sun himself. The diameter of the sun subtends an angle of about half a degree. 

 But over the diameter of the aperture in the shutter we have, of course, a complete 

 control ; we can make it half an inch, or the hundredth of an inch, as experiments re- 

 quire. Instead, therefore, of adjusting the screen, E D,fig. 115, at such a distance as 

 to receive the image of the sun himself, we remove it farther off, until we have depicted 

 upon it an image of the aperture in the shutter well denned, and with sharp edges 

 On interposing the prism, we find upon the screen, when adjusted to the proper focal 

 point, a spectrum of the aperture, in which the separation of the rays may be carried to 

 any extent by diminishing its diameter. 



158. When this diminution is carried too far, the quantity of light admitted becomes 

 very small, and under these circumstances NEWTON shows that great advantages arise, 

 by employing, instead of a circular hole, a longitudinal slit or oblong aperture; then the 

 separation of colour is equally perfect, and more light being admitted, the different ex- 

 periments may be made in a satisfactory way. 



159. Under certain circumstances, it is desirable to possess a spectrum in which the 

 intermingling of the colours by overlapping takes place in a regular manner. The plan 

 adopted by NEWTON was to employ a triangular hole. The image of this, when form- 



