The parallel plate is different from the F-P etalon in that two different transmission and reflection probabilities are needed at the two interfaces. A priori entering the parallel plate would be given a transmission coefficient of t , while for leaving the plate one would use t '. Similarly there are two different reflection coefficients r and r '. A symmetry argument due to Stokes, called ray reversibility, is explained in the diagram The relationships tell us that there are few degrees of freedom that control the amount of transmission/reflection for the interface, irrespective which way the rays go. > sol:=solve({r*r+t*tp=1,t*rp+r*t=0},{tp,rp}); > assign(sol); > t*tp; The statement r = - r ' is consistent with the phase change that the reflected wave undergoes [however, only if it reflects from an optically denser medium]. For a proper understanding of the latter one needs Fresnel's equations based on a proper wave propagation theory.

 
Hosted by Mywebcities.com