On the semi-classical spectrum of an integrable system of dymension 1 around a hyperbolic singularity (Q2380715): Difference between revisions

This paper is a continuation of the papers by \textit{Y. Colin de Verdière} and \textit{B. Parisse} [Commun. Partial Differ. Equations 19, No. 9-10, 1553--1563 (1994; Zbl 0819.35116)] and [Ann. Inst. Henri Poincaré, Phys. Théor. 61, No. 3, 347--367 (1994; Zbl 0845.35076) [CdV-P-2]] in the case of a double well potential for Schrödinger operators in dimension \(1\). In the latter, an implicit Bohr-Sommerfeld rule is obtained for the eigenvalues of the operator. In an energy region where this rule is rather mysterious, namely near the local maximal value of the double well potential, the present paper provides a more precise description of the spectrum. In a pedagogic way, the author recalls the Bohr-Sommerfeld rule of [CdV-P-2] and sketches the proof. Then he presents his result which strongly relies on the structure of the rule. Reviewer's remark: All this would be fine if some unpleasant drawbacks would have been corrected during the editing process. For instance, the year of the reference [CdV-P-2] and of some other reference are false. More importantly, the symbol ``\(\alpha \)'' has at least two different meanings: it is an eigenvalue and also a power of the semiclassical parameter \(h\). The statement of Theorem 3.2 on the Bohr-Sommerfeld rule of [CdV-P-2] is a bit confusing. One should understand that there exist functions \(\epsilon\), \(S^\pm\) of \(E\in[-1;1]\) which admit an asymptotic expansion in \(h\) with smooth coefficients in \(E\) such that the following holds true: \(E\) is an eigenvalue of the Schrödinger operator if and only if formula (3.1) is valid.