scientific article; zbMATH DE number 4010801

From MaRDI portal
Publication:3760754





Answering a very natural question of I. Kaplansky, originated in the finite-dimensional real Lie theory, \textit{S. A. Morris} had shown [Proc. Am. Math. Soc. 31, 625-626 (1972; Zbl 0209.060)] that a quotient group of an NSS-group \((=topo\log ical\) group with no small subgroups) needs not to have the NSS-property. Furthermore, it turned out that every topological group is a quotient group of an NSS-group [the reviewer (1981), an unpublished manuscript, quoted in the present paper]. This last result is an easy corollary of the following: Theorem. The free topological group F(X) is NSS iff the space X admits a continuous metric. This theorem was stated by \textit{S. A. Morris} and \textit{H. B. Thompson} [Proc. Am. Math. Soc. 46, 431-437 (1974; Zbl 0294.22001)]; later another proof was given by \textit{H. B. Thompson} [J. Aust. Math. Soc. 18, 482-484 (1974; Zbl 0297.22001)]. Unfortunately, both their proofs contain serious mistakes, as it is pointed out in the present paper. However, the theorem itself is correct, and the right proof of it forms the rest of the note. It is based on subtle considerations of combinatorial character exploring special bi- invariant metrics on F(X), first introduced by M. I. Graev. The most pleasant corollary for the reviewer is that both his result mentioned above and its proof remain valid.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3760754)