Conformal welding of annuli (Q1268679): Difference between revisions

If, for the circular rings \(A:a<| z|<1\), \(B:1<| z|<b\) we identify points on \(| z|=1\) by a sense-preserving homeomorphism \(\varphi\) taking \(\exp i\theta\) to \(\exp i\varphi (\theta)\) and we are able to introduce local uniformizing parameters at the identified points we obtain a doubly-connected Riemann domain \(D\) which we can map by \(f\) onto a circular ring \(A(\varphi,f):1<| w|<\exp M(\varphi,b)\) where the identification set corresponds to a Jordan curve \(\varphi\) separating the boundary components. The author calls \(M(\varphi,f)\) the modulus of \(A(\varphi, f)\), and denotes by \(V(\varphi)\) the set of all \(M(\varphi,f)\) for given \(\varphi\). Kôtaro Oikawa raised the problem whether there could exist such a \(\varphi\) for which \(V(\varphi)\) is a point but the conformal structure of \(D\) is not uniquely determined. In studying this the author is led to the following problems. 1. If \(F\) is continuous on the sphere \(\widehat C\) and conformal on \(\widehat C-y\) for a Jordan curve \(y\) but not conformal at a point of \(y\) what can be said about the values of \(F(z)+tz\) for complex \(t\)? 2. If \(E\) is a compact plane set and \(fu\), \(fv\) are the extremal horizontal and vertical slit mappings for its complement, for what complex \(g\) is \(fv+tfu\) univalent?