POLARISED LIGHT 353 



ciably separated, owing to the thinness of the crystal. When 

 these beams, polarised in rectangular planes, are again re- 

 solved by the analyser, we have two similar sets of waves 

 brought into interference, since one set has been retarded 

 more than the other in the slice of crystal. If the retardation 

 does not exceed a few wave-lengths, therefore, we get colour, 

 and this is complementary in positions of the analyser differ- 

 ing by 90 . 1 



It is usual to show this by a selenite giving red and green 

 images, and yellow and blue. The two may be mounted in 

 one slide. That no interference is produced whilst the vibra- 

 tions are in rectangular planes, is shown by removing the 

 analyser. On adding this and rotating, the complementary 

 colours appear. But when the analyser is in an intermediate 

 position, no colour appears, one ray from the selenite being 

 then suppressed by the analyser, and the other transmitted 

 without modification. 



That the phenomena of interference are essentially similar 

 to those of a thin film (only that in a selenite the retardation 

 is that of the difference of two paths through the film, while 

 in a ' thin film ' one of the rays is retarded by twice the whole 

 thickness, so that the selenite or mica is much thicker than 

 a ' thin film ' producing the same retardation), is best demon- 

 strated by spectrum analysis, just as the Newton's rings and 

 soap film were analysed. A plate of crystal ground concave 

 or convex exhibits Newton's rings, which are readily analysed 

 by placing a slit in a thin plate of metal across the slide in 

 the stage, focussing, and interposing a prism (the compound 

 direct prism is much the handiest for such experiments). The 

 slit across a Newton's ring slide shows precisely the same 



1 It must be understood that this work is solely concerned with practical 

 experimental conditions. No attempt is made or intended to explain pheno- 

 mena, or even to show their relations, further than the order of experiments may 

 do so. For some elementary explanation the reader is referred to my treatise 

 on Light : a Course of Experimental Optics, or other systematic treatises on 

 the subject may be consulted. 



A A 



