Jacobian quotients, an algebraic proof (Q860437)

Consider germs of analytic functions \(f\) at a point \(0\) of a smooth complex surface \(S\) and the curves (or, rather, germs of curves) \(\xi : f=0\) they define. Given two such germs \(f,g\), defining curves \(\xi\) and \(\zeta\) respectively, we have the Jacobian ideal \(\mathcal J=\partial (f,g)/\partial(x,y)\) (relative to local coordinates \(x,y\)) and the curve \(\mathbf J\) it defines. The collection \(Q\) of rational numbers \([\gamma,\xi]/[\gamma, \zeta]\), where \(\gamma\) is a branch of \(\mathbf J\) which is not a branch of either \(\xi\) or \(\zeta\) (and we deal with intersection multiplicity) is called the \textit{set of Jacobian quotients} of \(f\) and \(g\). On the other hand, we have the quotients \(e({\bar {\xi}_p})/e({\bar {\zeta}_p})\), where \(e\) denotes multiplicity and, e.g., \(\bar{\xi_p}\) is the total transform of \(\xi\) to \(p\), a point infinitely near to \(0\). \textit{H. Maugendre} [J. Lond. Math. Soc. 59, 207--226 (1999; Zbl 0941.58027)] proved, by using topological techniques, that if \(K\) is the collection of ratios \(e({\bar {\xi}_p})/e({\bar {\zeta}_p})\) as above, with \(p\) in a suitable set of points infinitely near to \(0\), then \(Q=K\). This is rather surprising, since the set \(K\) depends on the equisingularity (or topological) types of \(\xi\) and \(\zeta\) only, while the Jacobian germ (as well as its most common numerical invariants) does not. In the present paper the author gives a purely algebraic proof of this result. He skilfully uses many of the fundamental tools in the study of plane curve singularities: Puiseux expansions, Newton polygons, free and satellite points, etc. In the inclusion \(K \subseteq Q\) he also uses results on pencils of germs and their special members. All these concepts are carefully explained in [\textit{E. Casas-Alvero}, Singularities of plane curves. London Math. Soc. Lecture Notes Series 276 (2000; Zbl 0967.14018)].