On the summability of Fourier series with the method of lacunary arithmetic means (Q1058089)

Let f be \(2\pi\)-periodic and L-integrable over [-\(\pi\),\(\pi\) ] and let \(s_ n(f;x)\) denote the n-th partial sum of the Fourier series of f at x. Suppose \(L_ N\) is the linear operator on the space of continuous functions f defined by \(L_ N: f\to 1/N\sum^{N}_{k=1}s_{n_ k}(f;x).\) Let \(\| f\|\) denote the sup norm of f and \(\| L_ N\| =\sup_{\| f\| \leq 1}\| L_ Nf\|.\) The author proves the following theorem: Suppose that the convex sequence of positive integers \((n_ k)\) satisfies the weak lacunarity condition: \(n_{k+1}/n_ k>1+ck^{- \alpha},\) c is a constant, \(0\leq \alpha \leq 1/2\). Then there exists a positive integer S depending on c and \(\alpha\) such that \[ \| L_ N\| \geq A_ 1\quad N^{-1/2}\quad \log n_{[NS^{-1}]}. \] The above theorem shows that \(L_ n(f)\) does not converge for \(n_ k=2^{k^{\beta}}\) \((1/2<\beta <1)\).