Stabilizing decompositions of sigma (Q752549): Difference between revisions

Let \(\sigma\) denote the subset of the Hilbert cube Q defined by \(\sigma =\{(x_ i)\in Q:\) \(-1<x_ i<1\) and \(x_ i=0\) for all but finitely many indices \(i\}\). Define \(\sigma\) (n) to be a subset of \(\sigma\) such that a point \((x_ i)\) belongs to \(\sigma\) (n) if \(x_ i=0\) for \(i>n\). \textit{J. P. Henderson} and \textit{J. J. Walsh} [Topology Appl. 16, 143-154 (1983; Zbl 0525.57011)] gave an example of a decomposition G of \(\sigma\) with the finite dimensional closure of the nondegeneracy set \(N_ G\) such that: the decomposition space \(\sigma\) /G is not homeomorphic to \(\sigma\), however, \((\sigma /G)\times {\mathbb{R}}^ 1\) is homeomorphic to \(\sigma\). The author proves a general theorem which will be useful in establishing homeomorphisms between \((\sigma /G)\times {\mathbb{R}}^ 1\) and \(\sigma\) for a large class of decompositions. Here is the statement of the result: Theorem. If G is a cell-like upper semicontinuous decomposition of \(\sigma\) such that \(\sigma\) /G is countable union of finite dimensional compacta and \(\lim_{n\to \infty}n-r_ n=\infty\), \(r_ n=\lim [cl(N_ G)\cap \sigma (n)]\), then \((\sigma /G)\times R^ 1\) is homeomorphic to \(\sigma\). Oberserve that the condition \(\lim_{n\to \infty}n-r_ n=\infty\) is satisfied when the closure of \(N_ G\) is finite-dimensional.