Generalized axial maps and Euclidean immersions of lens spaces (Q1432033)

Let \(L^{2n+1}(2^e)\) be the \((2n+1)\)-dimensional \(2^e\)-torsion lens space, for positive integers \(n\) and \(e\). The purpose of this work is to study the immersion of these spaces in Euclidean spaces. More specifically, the authors introduce the concept of generalized \(e\)-axial maps, based in ideas first suggested by \textit{J. Adem}, \textit{S. Gitler} and \textit{I. M. James} [Bol. Soc. Mat. Mex., II. Ser. 17, 59--62 (1972; Zbl 0311.57016)]. The main result of the paper is: Theorem (1.2). If \(L^{2n+1}(2^e)\) immerses in Euclidean codimension \(k\), then there is an \(e\)-axial map of the form \(a:S^{2n+1}\times_{e}P^{2n+1}\to P^{2n+1}.\) The converse holds except perhaps for \(n=2,3\) or \(5\). Also they settle the immersion problem for \(L^{2n+1}(2^e)\) when \(n<9\), i.e. they find the optimal Euclidean dimension in which \(L^{2n+1}(2^e)\) can be immersed, independent of the fact they can not settle completely the converse in the main Theorem. To do this they consider a few particular cases where, as they point out, the most difficult ones are that \(L^{11}(4)\) immerses in \(\mathbb{R}^{17}\) but \(L^{11}(8)\) does not. Here they use the obstruction-theoretic method of modified Postnikov towers. At the end they discuss the relation between generalized \(e\)-axial maps and certain equivariant maps into Stiefel manifolds. As a result of a misprint, the references [1] and [2] which are the reference given above and \textit{M. F. Atiyah} [K-Theory, W. A. Benjamin, Inc. (1967; Zbl 0159.53302)], respectively, do not appear in the bibliography.