A classification of elementary cycles (Q2535983): Difference between revisions

Let \(\varphi^t\) be a \(C^1\) action of \(\mathbb R\) on an \(n\)-dimensional manifold \(X\). A cycle ( = closed orbit) \(C\) of period \(\tau\) is elementary if, for some \(x\in C\), \(T_x(\varphi^\tau)\colon T_xX\to T_xX\) has an \((n-1)\)-dimensional invariant subspace on which it is hyperbolic (i.e. has no complex eigenvalue of modulus 1). Cycles \(C_1\) and \(C_2\) are flow-equivalent if there is a homeomorphism \(h\) of a neighbourhood of \(C_1\) onto a neighbourhood of \(C_2\) taking \(C_1\) to \(C_2\) and preserving the local action of \(\mathbb R\). The main theorem of the paper classifies elementary cycles up to flow-equivalence. It says that the dimension and orientability (or lack of it) of the stable manifold and of the unstable manifold completely determine the equivalence class of the cycle, thus, in this case, classification by flow-equivalence is the same as classification by the (generally weaker) equivalence relation in which the homeomorphism \(h\) is only required to take orbit components onto orbit components.