Ends of non-metrizable manifolds: a generalized bagpipe theorem (Q2670090)

The paper under review is concerned with the study of the end-topology of non-metrizable manifolds, namely of \(n\)-manifolds which are NOT second countable. In particular the authors investigate the Freudenthal compactification of such manifolds and define two distinct classes of ends of them, called short ends and long ends, where the latter actually encode the non-countability of the manifold. The main point here is that the simplest method to obtain a non-metrizable manifold is to glue \(\omega_1\)-copies of a simple space (like an interval or a plane), where \(\omega_1\) is the first uncountable ordinal number. In this way one obtains manifolds that are intuitively too long to be metrizable. This is, roughly speaking, the class of Type I manifolds defined in the paper. The first theorem proved states equivalent conditions under which Type I manifolds are metrizable, whereas the main theorem just characterizes Type I metrizable surfaces. The appendix presents the construction of some special examples, the first one of a Type I manifold with all ends short and whose end space is not second countable, and the second one of a non-metrizable manifold, with all ends short, whose end space is second countable and homeomorphic to the Cantor set.