Families of diffeomorphisms without periodic curves (Q2571455): Difference between revisions

It is known after the work of \textit{M. Hakim} [Duke Math. J. 92, 403--428 (1998; Zbl 0952.32012)] and \textit{M. Abate} [Duke Math. J. 107, 1, 173--207 (2001; Zbl 1015.37035)] that a germ of holomorphic maps in \(\mathbb C^2\) at the origin, fixing the origin and tangent to the identity has always a certain number of petals (or parabolic curves). Roughly speaking, some of those petals (the ones called ``robust'' in \textit{M. Abate} and \textit{F. Tovena} [Abstr. Appl. Anal., No. 5, 275--294 (2003; Zbl 1018.37027)] ``belong'' to formal separatrices for the formal vector field of which the germ is the time one flow. In the paper under review, the author studies the relations between the germ of a (holomorphic or formal) diffeomorphism (not necessarily tangent to the identity) and the associated formal vector field and foliation. In particular, he proves that if a germ preserves a formal foliation, then there exists a formal curve which is invariant by the foliation and periodic for the germ. Moreover, such a curve can be chosen to be convergent if the foliation does. As a consequence, he shows that all (even formal) germs of diffeomorphisms have at least one irreducible formal periodic curve. The proof of the result is an adaption of the argument from \textit{C. Camacho} and \textit{P. Sad} [Ann. Math. (2) 115, 579--595 (1982; Zbl 0503.32007)]. Also, the author proves that there exist holomorphic germs of diffeomorphisms (which can be chosen either formally linearizable or tangent to the identity) such that they do not have convergent periodic curves.