Boundary behavior of proper holomorphic correspondences (Q1068998)

Let \(\Omega\) and D be bounded domains in \({\mathbb{C}}^ n\) and let \(\Gamma\) be a subvariety in \(\Omega\) \(\times D\) with \(\pi_{\Omega}\) and \(\pi_ D\) the projections of \(\Gamma\) into \(\Omega\) and D respectively. The set valued correspondence \(f\equiv \pi_ D\circ \pi^{-1}_{\Omega}\) (written \(f: \Omega -\circ D\)) (when \(\Gamma\) is irreducible or when \(\pi_{\Omega}\) and \(\pi_ D\) are proper f is called irreducible or proper respectively) is studied. Proper holomorphic mappings give examples of proper correspondences, but there are others. If \(f: \Omega - \circ D\) is a proper holomorphic correspondence then there is a positive integer p and a subvariety \(V\subset \Omega\) such that for \(z_ 0\in \Omega \setminus V\) there is an open neighborhood U of \(z_ 0\) in \(\Omega\) \(\setminus V\) and proper holomorphic mappings \(f_ i: U\to D\) \((i=1,2,...,p)\) such that the set f(z) is \(\{f_ 1(z),...,f_ p(z)\}\). The authors prove four theorems. These results extend and improve known (and important) results on proper mappings. We quote two of their four theorems. Theorem 1. If \(\Omega\) is a bounded, simply connected, strictly pseudoconvex domain (in \({\mathbb{C}}^ n)\) with \(C^{\infty}\) boundary, then every irreducible proper holomorphic self correspondence \(f: \Omega -\circ \Omega\) of dimension n is a biholomorphic mapping. Theorem 3. Suppose \(\Omega_ j\) are bounded \(C^{\infty}\) domains in \({\mathbb{C}}^ n\) and \(f: \Omega\) \({}_ 1-\circ \Omega_ 2\) is an irreducible proper holomorphic correspondence. If \(\Omega_ 1\) is pseudoconvex of finite type and \(\Omega_ 2\) is strictly pseudoconvex then f splits (smoothly to \(\partial \Omega_ 1)\) at each point of \({\bar \Omega}{}_ 1.\) To say that f splits at \(z_ 0\in \Omega\) means there is U open \((z_ 0\in U)\) and holomorphic mappings \(f_ j: \Omega_ 1\cap U\to \Omega_ 2\), \(j=1,2,...,p\), that represent f.