108 A CHAPTER ON LIGHT AND COLOUR. 



illumination where homogeneous light would produce contrasts 

 of light and shade. 



15. — To study the effect of the mixture of light from two 

 exactly similar sources, the light from a single source, as nearly as 

 possible a point, may be either reflected from two mirrors incHned 

 very slightly to one another, or refracted through a double prism 

 (called a biprism) of very small angle. The two images in either 

 case serve as the two similar sources of light. 



If the Hght so reflected or refracted be received on a suitable 

 screen, the illumination is found not to be uniform. The centre 

 of the screen is brightly illuminated from each source .of light, but 

 in passing to one side, we reach a point where the light from one 

 source is half a wave-length behind that from the other, and con- 

 sequently in a condition to extinguish the effect of the first. 

 Hence a dark line is found at this place. Further on the differ- 

 ence of path is a whole wave-length, and a bright band results, 

 followed by other dark and bright bands alternately. 



If the source of light be monochromatic, the alternations of 

 light and darkness can be well observed for some distance. By 

 measuring the breadth of the intervals between the bands, and 

 the distance of the screen from the source of light, the wave-length 

 of the light employed can be calculated. The scale of the effect 

 is larger for red than for blue or violet light, the breadth of the 

 bands gradually diminishing with the wave-length of the light 

 as we pass from red to violet. If white light be used, these 

 differences soon cause the contrasts to lose their sharpness, as the 

 bright bands of some colours fall on the dark bands of other 

 colours at a very slight distance from the centre of the screen. 



16. — We see thus that the addition of Hght to light may pro- 

 duce darkness, and we may hence expect that the subtraction of 

 some part of the illumination due to a point of light may increase 

 the intensity of the brightness at some points in space. This 

 expectation can be proved to be well-founded. 



By interposing an obstacle with a straight edge in the path of 

 the light falling on a screen from a luminous point, the effects 

 within the boundary of the geometrical shadow shew this strange 

 effect. The brightness outside the shadow does not abruptly cease 

 at its edge, but is succeeded by a series of alternations of bright- 



