1885.] Phenomenon of Crystalline Reflection. 185 



incident wave and its fellow. In these cases, therefore, an incident 

 wave would produce hardly more than one refracted wave, namely, 

 that one which nearly agrees with the incident wave in direction of 

 polarisation. In these cases the colours are not produced. It appears, 

 therefore, that their production demands that the incident wave shall 

 be very determinately divided into two refracted waves, accompanied 

 of course by reflected waves. 



It seems evident that the thickness of the stratum affects the result 

 through the difference of phase which it entails in the two refracted 

 waves on arriving at the second twin plane. But whereas in the 

 ordinary case of the production of colour by the interposition of a 

 crystalline plate between a polariser and an analyser, we are concerned 

 only with the difference of retardation of the differently polarised 

 pencils which are transmitted across the plate, and not with the 

 absolute retardation, it is possible that in this case we must take into 

 account not only the difference of retardation for the differently 

 polarised pencils which traverse the stratum, but also the absolute 

 retardation ; that is, the retardation of the light reflected from the 

 second relatively to that reflected from the first twin plane. 



21. I have not up to the present seen my way to going further. It 

 is certainly very extraordinary and paradoxical that light should suffer 

 total or all but total reflection at a transparent stratum of the very 

 same substance, merely differing in orientation, in which the light 

 had been travelling, and that, independently of its polarisation. It 

 can have nothing to do with ordinary total internal reflection, since it 

 is observed at quite moderate incidences, and only within very narrow 

 limits of the angle of incidence. 



VOL. XXXVIII. 



o 



