LIGHT: DISPERSION 283 



With regard to the width of the slit, as the spectrum may be 

 regarded as an infinite number of images of it, differently de- 

 flected according to their colours, thenarroiver the slit the purer 

 the spectrum, especially at a small screen distance. But with 

 the simpler experiments a wide slit gives much more brilliance, 

 without any obvious confusion at a good screen distance (which 

 spreads out the colours more), and as much as 6 mm. wide 

 may sometimes be employed with advantage. 



161. Minimum Deviation. It will be found that there is 

 one position of the prism which refracts the rays least, called 

 the position of minimum deviation. This is the proper position 

 for the prism. Now and then, however, it may be desirable 

 to turn the prism more round in order to get the greater 

 length of spectrum thus produced. 



162. Different Colours Differently Refracted. Newton's 

 two experiments to prove this are both striking, and both 

 easy. For the first, we place a short slit say a square aper- 

 ture 3 mm. long each side in the stage, and project its 

 spectrum in the ordinary way, with a prism bottle. Behind 

 this, we adjust at the proper height the glass prism with its 

 refracting edge horizontal, screening stray rays if necessary 

 by a black card pierced with an aperture. The rays are now 

 again deflected either up or down, according to the position of 

 the glass prism ; and as this is moved along in the rays of the 

 spectrum from the first prism, from the red end, it will be seen 

 that the rays are refracted more and more as we get towards 

 the blue end. 



Newton's other experiment is particularly elegant, depend- 

 ing on the fact that if each colour has its own degree of 

 refrangibility, it must also have its own angle of total re- 

 flexion ; violet rays being (because more refrangible) totally 

 reflected at an angle which allows red rays to pass. To per- 

 form it with the lantern we arrange as in fig. 150, removing 

 the objective, and placing a perpendicular slit at N, on the 

 front of the nozzle. Through this we send a sharpened 



