138 On Spectra formed hy the passage of 



of course vanish at the four zero points. Between these points, 

 however, remarkable phenomena occur. A person unacquainted 

 with the true nature of the colours of polarization, and proceeding 

 on the analogy of homogeneous hght, might expect to get a spec- 

 trum consisting only of green rays, seeing that that is the colour of 

 the field when the spectrum arrangement is removed. This how- 

 ever is not the case, and the result very beautifully illustrates to 

 the eye what is well known theoretically to be the true nature of 

 these colours. What we obtain is a continuous spectrum consisting 

 of all the prismatic colours, in greater or less intensity, with the 

 striking pecuharity that there is a well-marked dark band in the 

 red, similar in appearance to the well-known absorption-bands which 

 many substances produce in the spectrum, only blacker and better 

 defined than these are ever seen. 



The following is the mode in which the band makes its appear- 

 ance. As the zero point is passed, the light first makes its appear- 

 ance ia the green of the spectrum, from which point, as the selenite 

 is rotated, the light opens out in both directions. When the hght 

 reaches the red, the black band makes its appearance, and attaias a 

 maximum blackness when the selenite is at 45^, viz. when, without 

 the use of the dispersion prisms, the field would contain green light 

 of maximum brightness. When this point of revolution is passed, 

 the band again fades, the spectrum becomes obscured at each end, 

 the darkness creeping in towards the green, till at 90^ the spectrum 

 has again vanished. The same phenomena recur at each quarter of 

 a revolution. 



Let the Nicol's prisms now be set with their axes parallel, and 

 the same selenite rotated on the stage as before. The result is what 

 we should be led to expect from the last experiment. At the zero 

 points the selenite exercises no influence, and we have a continuous 

 ordinary spectrum. As a zero point is passed, however, a dark 

 band makes its appearance, but this time in the green rays. The 

 band is at first faint and nebulous, but becomes blacker and sharper 

 as the stage is rotated, till at 45"^ it attains its maximum. The 

 spectrum in this experiment never vanishes, but is apparently quite 

 contiauous throughout, save for the appearance of the black band. 



Lastly, let the selenite be fixed at 45° from the zero point, and 

 the analyzer rotated. We have now a combination of the two 

 pre\aous experiments. The band in the red appears alternately 

 with the band in the green at each quarter revolution, the former 

 being at its maximum when the axes of the prisms are perpendicular, 

 the latter when these axes are parallel.* 



* If the axis of tlie selenite makes a isa'eater or less angle than 45° with the 

 plane of polarization, the result is that wliile the same band stiU recurs after 180° 

 of a revolution of the analyzer, tlie complementary band is no longer separated 

 from it by 90°, but by a gix-ater or loss angle. 



