104 THE ART OF PROJECTING. 



appear, as before. The prism may be an ordinary one, 

 but the bow will not then be perfect for a semicircle. 

 The size of this bow will depend upon the size of the 

 prism, for the ring of light can be indefinitely enlarged 

 by varying the angle of incidence of the beam. 



With a curved slit cut about the size of a 'common 

 transparency and projected with the lantern, holding 

 a common prism in front of the objective, a very good 

 bow is seen. As the refraction bends the rays down- 

 ward, it will be necessary to tip the front of the lantern 

 up considerable in order to get the bow upon the 

 screen. (Fig. 79.) 



Lastly, let a beam of parallel rays, about an inch in 

 diameter, fall upon a glass sphere filled with water, 

 an ordinary small glass flask answers well, but the 

 larger the flask the greater should be the size of the 

 beam. If, now, a small white screen be placed between 

 the sphere and the aperture in the window, a bow will 

 be seen concentric with the aperture and arranged so 

 that the red is outside and the violet inside. At a 

 greater distance from the aperture another bow will be 

 formed, much fainter than the first, and with the colors 

 in the inverse order. If different colored glasses are 

 interposed in the path of the beam of white light, the 

 bow will be seen to consist mainly of the tint of the 

 glass. 



CHROMATIC ABERRATION. 



Caustic curves, due to chromatic aberration in the 

 lenses, may be projected by taking two rather large 

 lenses of short focus, such, for instance, as those made 

 for lantern condensers. 



Place the first one as if for common projections. 

 The second may be held in the hand and brought near 

 to the focus of the first and then inclined, as shown in 



