Convexity theorems for the circle methods of summability (Q1196853)

Let \(s=\{s_ n\}\) be a given sequence of complex numbers and let \(m\) be the set of all bounded complex sequences. The symbols \(B\), \(E_ \alpha\), \(T_ \delta\), \(S_ \delta\) denote Borel, Euler-Knopp, Taylor and Meyer- König methods respectively. It is known that the family of summability methods consisting of Abel methods and all Cesàro methods \(C_ \alpha\), \(\alpha>-1\), is convex. The result is due to \textit{J. E. Littlewood} [Proc. Lond. Math. Soc., II. Ser. 9, 434-448 (1911; Jahrb. F.d.M. 42, 276)] and \textit{A. F. Anderson} [Studier over Cesàro's summabilitets metode, Dissertation, Kœbenhavn (1921; Jahrb. F.d.M. 48, 225)]. In this paper the authors prove that this theorem remains true if we replace the Abel method and Cesàro methods by the Borel method and Euler-Knopp methods of positive order, respectively. More precisely, they prove that if \(\alpha>0\), then \(B-\lim s_ n=\sigma\wedge E_ \alpha s\in m\) implies \(E_ \beta-\lim s_ n=\sigma\) for each \(\beta\) with \(0<\beta<\alpha\). They give two different proofs of this theorem. They further demonstrate that if \(\gamma>0\), \(0<\delta<1\), then in the above theorem one may replace \(B\) by \(E_ \gamma\), \(BE_ \gamma\), \(T_ \delta\), \(T_ \delta E_ \gamma\), \(S_ \delta\) or \(S_ \delta E_ \gamma\).