LIGHT AND SOUND. 



divergent before they reach the eye, and are brought 

 to a focus, where they should be, on the retina. 



We have said above, that in short-sighted per- 

 sons the rays do not reach the retina unless the 

 object is held close to the eyes. The effect pro- 

 duced by this is similar to that of employing con- 

 cave spectacles ; because the nearer we hold an 

 object to our sight, the angle of the rays from it is 

 the wider ; the rays are more expanded before they 

 enter the eye that is, more divergent. Thus 



the extreme rays 

 from a point to 

 =-< l the pupil of the 

 eye make a 

 greater angle at 

 o than those from 



Fig. 19. 



a point of a more distant object make at a; that is, 

 the rays from o are more divergent on entering the 

 eye than the rays from a, and thus nearness of an 

 object is equivalent to seeing it at a greater dis- 

 tance through a concave lens. So when the object 

 a is further distant than <?, the rays from a have a 

 less divergence, which is equivalent to viewing it 

 at a nearer distance with a convex lens. Long- 

 sighted persons must therefore diminish the diver- 

 gency of the rays falling on the crystalline lens, 

 and this they can do either by holding the object 

 at which they are looking at a greater distance, or 

 by using convex spectacles. Short-sighted per- 

 sons can, to some extent, remedy their defective 

 vision by partially closing the eyes. The effect of 

 this partial closure is to cut off those rays which 

 would pass through the marginal parts of the lens, 

 and which principally cause the confusion and 

 indistinctness of their vision. 



We think it unnecessary to discuss the questions 

 why, having two eyes, we should see objects single, 

 and why we should see objects erect, when their 

 image on the retina is inverted, as these questions 

 belong more to physiology and the science of mind 

 than to optics. 



DISPERSION OF LIGHT BY REFRACTION. 



If the ethereal waves whose motion constitutes 

 light were all of the same length in other words, 

 if light were homogeneous refraction by prisms 

 and lenses would be as we have described it. But 

 when a small beam of sunlight falls on a prism, 

 the refracted beam is not white, but is split up 



Fig. 20. 



into a number of coloured beams. The experi- 

 ment which was devised by Newton may be thus 

 performed. In the shutter, AB, of a darkened 

 room, a small hole, C, is made, to give entrance 



to a small pencil of sunlight, CD, which, if un 

 interrupted, would proceed in a straight line to 

 E, and form a white spot, or image of the sun. 

 A prism, FGH, is interposed in the path of the 

 beam, and a screen is placed to receive the pencil 

 emerging at K. The sun's image formed on the 

 screen, instead of being a circular spot, is now a 

 narrow oblong, whose width is the same as the 

 diameter of the spot, but length many times 

 greater. Also this stripe is coloured throughout 

 its whole length, being red at the lowest part, R, 

 and the colours shading imperceptibly through 

 orange, yellow, green, blue, and indigo, to 

 violet at V. This experiment proves that sunlight 

 is a mixture of light of different colours and of 

 different refrangibilities. If we make a hole in the 

 screen at R, so as to allow only the red rays to 

 pass through, and then place a second prism 

 in their path, we find that they are incapable of 

 further subdivision, and so for all the other 

 colours. Also, if we look at the coloured stripe, 

 or spectrum, as it is called, through another prism, 

 like FGH, and placed similarly to it, all the colours 

 are re-combined into white, and the circular image 

 of the sun is again seen. In order that the colours 

 of the spectrum may be pure, it is necessary, first, 

 that the deviation should be the least possible ; 

 and secondly, that the aperture in the shutter 

 should be small. The first condition is satisfied 

 by turning the prism till the beam leaves the 

 second face at the same angle as it strikes the 

 first ; or, more simply, till the spectrum occupies 

 the lowest position on the screen possible. The 

 necessity of the second condition will be obvious, 

 if we consider that in a wide aperture, the highest 

 and lowest rays will each depict its own spectrum 

 on the screen, and that the red or orange of the 

 lowest spectrum may overlap the orange or yellow 

 of the highest. 



To explain the production of coloured fringes 

 when an object is looked at through a prism or 

 lens, let us suppose 

 that each point of 

 the object sends out 

 rays of all colours. 

 As the red rays are 

 least refracted, there 

 will be a red image 

 of the object AB 

 formed as at ab, and 

 a violet one at ef t 

 the other coloured 

 images being inter- 

 mediate in position. 

 Between e and b, all 

 the colours will be 

 mixed in the proper 



Fig. 21. 



proportion to form white light ; and an eye at E 

 will see the image colourless from e to 6, but red 

 predominating in ae, and blue in bf. If we vary 

 the material of the prism, by taking, for instance, a 

 prism of dense flint-glass, instead of one of crown- 

 glass, we find the length of the spectrum con- 

 siderably altered, and the proportions of its differ- 

 ent parts to be no longer the same as before. Not 

 only has flint-glass a higher dispersive power than 

 crown-glass, but it disperses the violet end of the 

 spectrum in a higher degree than the red end. 

 All spectra produced by refraction are therefore 

 said to be irrational. If the dispersion were 

 always proportional to the refraction, any bending 



247 



