The push-out space of transnormal immersions (Q707539)

The notion of a transnormal manifold was first introduced by \textit{S. A. Robertson} [Mich. Math. J. 11, 97--105 (1964; Zbl 0123.29501)] as a generalization of constant width for submanifolds of Euclidean space. Its extension to immersions into Euclidean space essentially reduces up to coverings to embeddings into Euclidean space. Transnormality requires that the normal space meet the embedding as normal space only. For a transnormal embedding the number \(r\) of intersection points of a normal space is constant throughout the embedding if the manifold is complete with respect to the induced metric. For such an embedding the normal bundle has a parallel (flat) direct summand. Here it is assumed that the normal bundle is flat itself. Under this assumption the author obtains results on the components of the pushout space of a transnormal immersion, where these spaces have been defined by \textit{S. Carter} and \textit{Z. Sentürk} in [J. Lond. Math. Soc., II. Ser. 50, No. 2, 404--416 (1994; Zbl 0807.53042)]. They extend previous results on the pushout spaces obtained by \textit{S. Carter} and the author in [Beitr. Algebra Geom. 41, No. 2, 359--370 (2000; Zbl 0980.53066)], providing a deeper insight into the behaviour of the normal bundle of a transnormal embedding as presented in a first version in the paper by \textit{B. Wegner} in [J. Differ. Geom. 16, 93--100 (1981; Zbl 0406.53005)].