364 OPTICAL PROJECTION 



211. Quarter- wave Plates. The light is in fact circularly 

 polarised by the compounding of the two vibrations as they 

 leave the film; and therefore, although the analyser again 

 analyses the circular motion into two rectangular ones, it is 

 quite indifferent at what part of the circular orbit it does so. 

 A mica of this thickness, known as a ' quarter- wave plate,' 

 is therefore in constant use for many experiments in polari- 

 sation. 



It has been pointed out l that the colour cannot be quite 

 the same all round the revolution of the analyser, except for 

 the homogeneous light to which the thickness is adjusted; 

 and when this is yellow (as usual), in white light a very slight 

 residuum of orange appears in one position, and of bluish-grey 

 in the other, owing to the longer and shorter wave-lengths not 

 being in precisely quarter-wave relations. 



The direction of the principal axis (that which joins the 

 centres of the two systems of ' brushes ' in mica presently 

 described) should always be marked upon the plate in some 

 way. One way is to make diamond scratches on the edges of 

 the glass discs containing the film, at intervals of 45 ; or the 

 principal axis may be scratched across the diameter of the 

 mica itself before mounting. I prefer the latter, which is 

 quite unnoticeable unless carefully looked for. Such a film, 

 in unmounted discs, can then be used in any position in the 

 rotating frame (fig. 195). But as two definite positions are 

 constantly required in this class of experiments, I prefer to 

 mount two separate plates, each as large as can be inserted 

 in a standard frame, permanently, one with its axis perpen- 

 dicular, and the other at 45. They can be put into thinner 

 frames this way, and are always ready adjusted, which is 

 convenient when only one stage is provided. 



The effect of a quarter- wave plate obviously depends upon 

 its axes being at an angle of 45 with the preceding planes of 

 polarisation. It is readily seen that the plate, if placed with 



1 See note to p. 856. 



