A note on \(\tau \)-convergence, \(\tau \)-convergent algebra and applications (Q664704)

If on a bounded sequence \(x\in l^{\infty}\) all Banach limits are the same then \(x\) is said to be almost convergent. Such an \(x\) forms a closed linear (but not closed for pointwise multiplication) subspace of \(l^{\infty}\). Further, a bounded sequence \(x\) is called multiplier if the product \(xy\) almost converges whenever \(y\) does. In [Ill. J. Math. 16, 687--694 (1972; Zbl 0241.43002)], \textit{C. Chou} defined \(\tau\)-convergence and proved that a bounded sequence is a multiplier iff it \(\tau\)-converges. The paper under review continues and extends these ideas. First the notion of multiplier is strengthened by adding multiplicativity in the sense that a multiplier \(x\) is supposed to additionally satisfy that the almost limit of the product \(xy\) is the product of the almost limits of \(x\) and \(y\) for all almost convergent \(y\). Then, parallel to Chou, the author proves (with a shorter proof valid also for complex scalars) that this stronger notion of multiplier, too, is equivalent to being \(\tau\)-convergent; hence both notions of multiplier coincide (which is not stated explicitly in the paper). In the second part of the paper the author shows that the set \(M_\tau\) of all \(\tau\)-convergent sequences forms a unital \(C^*\)-subalgebra of \(l^{\infty}\) and considers the \(\tau\)-compactification of the natural numbers via the Gelfand transform of \(M_\tau\); this compactification differs from the Bohr and the Stone-Čech compactifications. It is shown that \(M_\tau\) is not \(C^*\)-reflexive.