io8 ELEMENTS OF SCIENCE 



and must therefore undergo correspondingly different 

 amounts of refraction. Therefore it is impossible they 

 should be parallel when they issue forth from a body 

 after having entered it through a curved surface. Every 

 convex surface causes convergence of the rays, and every 

 concave surface scatters. 



When speaking about heat, we said* that there 

 appeared to be some reason for supposing that its rays 

 differ qualitatively as well as in mere differences of 

 intensity. What was thus suggested as to heat, is 

 certain and evident as regards light. 



No reader can have failed to see, now and again, 

 manifestations of colour in the neighbourhood of some 

 piece of glass, or of crystal, and he may have remarked 

 a resemblance between the colours thus appearing and 

 the hues of the rainbow. In the rainbow, the lowest of 

 the series of tints it exhibits is violet, and to that succeed 

 indigo, blue, green, yellow, orange, and red. There are, 

 however, so many intermediate shades, that it is impossible 

 to see where one ends and the next begins. It is gene- 

 rally believed that light, which is apparently colourless, 

 somehow consists of different coloured rays. This belief 

 certainly appears to be confirmed by a very simple 

 experiment. If a circular piece of cardboard be painted 

 with the colours of the rainbow, each patch of colour 

 narrowing to a point at the centre of the card ; then if 

 a rod be passed through that centre, and the card be 

 turned rapidly round it, the separate colours will 

 disappear and the card will assume a grey, or nearly 

 white, appearance. 



Colour has been before referred to,t but its con- 

 sideration was postponed until the reflexion of light 



* See ante, pp. 98 and 101. t See ante, p. 104. 



