Extension of a theorem of Fejér to double Fourier-Stieltjes series (Q1602282)

Suppose \(F\) is a function of bounded variation on \([0,2\pi]\). A classical theorem of Fejér says \[ \lim_{n\to\infty}{1\over n} s_n(dF,x)=\begin{cases} {1\over \pi}\Big(F(x+)-F(x-)\Big) & {\text{ if }}0<x<2 \cr \pi, {1\over \pi}\Big(F(0+)-F(0)+F(2\pi)-F(2\pi-)\Big) & {\text{ if }} x=0 {\text{ or }} x=2\pi \end{cases}\tag{1} \] where \(s_n(dF,x) \) denotes the \(n-\)th partial sum of the Fourier-Stieltjes series of \(F\) and \(F(x+)=\lim_{\varepsilon\to 0+}F(x+\varepsilon), F(x-)=\lim_{\varepsilon\to 0+}F(x-\varepsilon)\). The author gives a proof of (1) for the case \(x=0\) or \(x=2\pi\) which is not considered in \textit{A. Zygmund}'s book [``Trigonometric series'', Vol. 1, 2nd ed. (1959; Zbl 0085.05601)]. Also, the author gives a representation of (1) in terminology of measure. To do so, he extends the definition of \(F\) by the condition \(F(x+2\pi)-F(x)=F(2\pi)-F(0)\) for all \(x\in {\mathbb R}\). Then an interval function is defined by \[ \mu(F, I)=F(b)-F(a) \quad \text{ for the finite open interval } I=(a,b). \] Denote by \(\mu_F\) the Borel measure induced by the interval function \(\mu(F,\cdot)\). Then (1) is rewritten as \[ \lim_{n\to\infty}{1\over n} s_n(dF,x)= {1\over \pi}\mu_F(\{x\})\tag{2} \] where \(\{x\}\) denotes the set consisting of the single point \(x\). The main purpose of this paper is to extend (2) to the double Fourier-Stieltjes series. The result is that for a function \(F\) of bounded variation on \([0,2\pi]\times [0,2\pi]\) in the sense of Hardy and Krause, \[ \lim_{m,n\to\infty}{1\over mn} s_{mn}(dF,x)= {1\over {\pi}^2}\mu_F(\{(x,y)\})\tag{3} \] where \(s_{mn}(dF,x,y)\) denotes the \((m,n)-\)th rectangular partial sum of the Fourier-Stieltjes series of \(F\) and \(\mu_F(\{(x,y)\})\) denotes the value of the corresponding measure \(\mu_F\) at the single point set \(\{(x,y)\}\).