scientific article; zbMATH DE number 3877551

From MaRDI portal
Publication:3344477





The Rayleigh functions \(\sigma_ n(\nu)\), defined as the sums of the (2n)th powers of the reciprocals of the non-zero roots of the Bessel function \(J_{\nu}(z)\), or alternatively by the formulas \[ \sigma_ 1(\nu)=1/4(\nu +1),\quad (\nu +n)\sigma_ n(\nu)=\sum^{n- 1}_{k=1}\sigma_ k(\nu)\sigma_{n-k}(\nu),\quad n=2,3,..., \] were considered extensively by N. Kishore; for references see the review of \textit{E. C. Obi}, J. Math. Anal. Appl. 52, 648-659 (1975) in Zbl 0321.33015. The author shows that \[ [J_ 0(z)]^{-2}=1- \sum^{\infty}_{n=1}2\sigma '_ n(0)z^{2n},\quad | z| <j_{0,1} \] and that \[ [1-J^ 2_ 0(z)]/J^ 2_ 1(z)=2- \sum^{\infty}_{n=1}2\sigma '_ n(1)z^{2n},\quad | z| <j_{1,1}, \] where \(j_{\nu,1}\) is the smallest positive zero of \(J_{\nu}(z)\). In this way he obtains formulas for the recursive calculation of \(\sigma'_ n(0)\) and \(\sigma'_ n(1)\). He shows that a certain sequence considered by \textit{L. Carlitz} [Proc. Am. Math. Soc. 14, 1-9 (1963; Zbl 0113.034)] is related simply to \(\{\sigma'_ n(0)\}\). \(\{\) Reviewer's remark: In part (ii) of the Proposition, \(J_ 1\) should be \(J^ 2_ 1.\}\)











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3344477)