Some \(\sigma\)-algebras (Q1065187)

A given group G of transformations acting on a fundamental space E of the same cardinality as E is considered. Suppose that the cardinality is \(\omega_ 1\), where \(\omega_ 1\) is the first uncountable ordinal. Consider such measures on E which vanish on singletons. Then one of the results of the paper states: Under the continuum hypothesis there exists a collection \(\{E_{mn}\}\) where m,n are positive integers and \(E_{mn}\subset E\), such that \(E_{mn}\) are almost G-invariant in E. Moreover if \(\mu\) is a non- trivial \(\sigma\)-finite measure on E, at least one of the sets \(E_{mn}\) is such that it does not belong to the domain of \(\mu\). Here an almost G-invariant set is defined in the following way: If \(X\subset E\) then X is almost G-invariant in E if for any \(g\in G\) we have card(g(X)\(\Delta\) X)\(\leq \omega_ 0\) where \(\omega_ 0\) is the first infinite cardinal.