On the C Stability Conjecture for Flows
Lan Wen
1996-01-01
Abstract:We study the C stability conjecture of Palis and Smale [PS] for flows in this paper. Let M be a compact Riemannian manifold without boundary. For each r 1, let X(M) denote the set of C vector fields of M, endowed with the C topology. Every S # X(M) generates a C r flow ,=, : M_R R. We say S is C structurally stable if S has a C neighborhood U in X(M) such that every X # U generates a flow , that is topologically equivalent to ,. As usual, two flows are topologically equivalent if there is a homeomorphism h : M M that maps the orbits of one flow onto those of the other flow while preserving the orientation. Thus C structural stability implies C structural stability, when both make sense. In [PS] Palis and Smale conjectured that S is C structurally stable if and only if S satisfies Axiom A plus the strong transversality condition (see below for definitions). The ``if '' part was proved first for r 2 by Robbin [R], and then for r=1 by Robinson [Rs1]. The ``only if '' part was reduced by Robinson [Rs2] to proving that C structural stability implies Axiom A, which has been known as the C stability conjecture. Unlike the ``if '' part, the first target for studying the C stability conjecture is the case of r=1, i.e., the C stability conjecture. The case of r 2 is beyond our knowledge as discussed in [M2], and will not be considered in the present paper. The C stability conjecture has attracted attention for a long time. Many researchers made important contributions to this problem, among whom