CHANGES FROM FGDP

In FGDP I used R(Z) to denote a reflection seismogram, an impedance function, and a spectrum with autocorrelation coefficients r_t. I liked this classic notation, which was used by the mathematicians Wiener and Levinson. It is confusing, however, to use in one equation r_t both for the causal, one-sided, reflection seismogram and for the two-sided autocorrelation. Thus I have introduced S, which is a natural notation for spectrum, although s is admittedly less natural for autocorrelation.