On Riemann's zeta function (Q1074634)

Riemann's ''second proof'' of the functional equation for \(\zeta\) (s) depends on the inversion formula for a classical theta-function, and consists in proving for \(\pi^{-s/2} \Gamma (s/2) \zeta (s)\) an expression which remains invariant if s is replaced by \(1-s.\) Generalizing this approach, the authors introduce certain polynomials \(p_ j(s)\) related to Hermite polynomials via an integral transformation, and derive for \(p_ j(s) \pi^{-s/2} \Gamma (s/2) \zeta (s)\) an expression which for even j remains invariant and for odd j changes the sign in the replacement \(s\to 1-s.\) Since also \(p_ j(s)=(- 1)^ j p_ j(1-s),\) this formula implies the functional equation again. It is interesting that Riemann's hypothesis holds for \(p_ j(s)\), i.e. its zeros lie on the line \(\sigma =1/2\). Applying the generating function technique, the authors now obtain an identity involving the zeta-function and a free real parameter u. This leads to a new formulation of Riemann's hypothesis.