On a Tauberian theorem of Hardy and Littlewood (Q5947475): Difference between revisions

It is the aim of this paper to give a modified proof of a partial result in the proof of the classical high-indices-theorem of Hardy-Littlewood. The latter theorem says: If for some fixed sequence \(0\leq (\lambda_n) \uparrow\infty\) we have \(\sum_0^\infty a_n\exp (\lambda_nx)\to s\) \((x\to 1-)\) then \(\sum^n_0 a_k\to s\) \((n\to\infty)\). The author proves this result under the additional Tauberian condition \(a_n=O(1)\). The proof is somewhat simpler than the original one but follows similar arguments. The main theorem in this paper follows also from different approaches as e.g. from results on Laplace transforms of the reviewer and \textit{R. Trautner} [Ratio Tauberian theorems for Laplace transforms without monotonicity assumptions, Q. J. Math., Oxf. II. Ser. 36, 363-381 (1985; Zbl 0581.44002)].