On the value distribution of Hurwitz zeta-functions at the nontrivial zeros of the Riemann zeta-function (Q5956168)

The paper concerns the value distribution of the Hurwitz zeta-function \(\zeta(s,\alpha)=\sum_{n=0}^\infty{1\over (n+\alpha)^s}\) at the nontrivial zeros \(\rho=\beta+i\gamma\) of the Riemann zeta-function. \textit{A. Fujii} [Proc. Japan Acad., Ser. A 65, 139-142 (1989; Zbl 0686.10029)] assuming the Riemann hypothesis proved \[ \sum_{0<\gamma\leq T} \zeta({1\over 2}+i\gamma, \alpha)=-\left(\Lambda({1\over\alpha})+L(1, -\alpha)\right){T\over 2\pi}+O(T^{9\over 10}\log T), \] where \(\Lambda(n)\) is the von Mangoldt-function and \(L(s, \lambda)= \sum_{n=1}^\infty{\exp(2\pi i\lambda n)\over n^s}.\) Avoiding the Riemann hypothesis the author obtains for fixed \(0<\alpha<1\) and \(H\leq T\) that \[ \sum_{T<\gamma\leq T+H} \zeta(\rho, \alpha)=-\left(\Lambda({1\over\alpha})+L(1, -\alpha)\right){H\over 2\pi}+O\left(H\exp\left(-C(\log T)^{1\over 3}\right)+ T^{{1\over 2}+\varepsilon}\right), \] with some absolute constant \(C>0\). For the proof he uses the method of \textit{J. B. Conrey, A. Ghosh} and \textit{S. M. Gonek} [Publ. Math. Orsay 88/02, 77-83 (1988; Zbl 0712.11052)]. By his result the author derives that \({\zeta(s, \alpha)\over \zeta(s)}\) is an entire function if and only if \(\alpha = {1\over 2}\) or \(\alpha = 1\).