Annular solutions to the partitioning problem in a ball (Q6906156)

The classical \textit{partitioning problem} in the unit ball \(\mathbb{B}^3\) of the 3-dimensional Euclidean space asks to find, among all surfaces dividing \(\mathbb{B}^3\) into two parts of prescribed volumes, those which are critical points for the area functional. In [Advances in geometric analysis and continuum mechanics. Proceedings of a conference, held at Stanford University, Stanford, CA, USA, August 2-5, 1993 in honor of the seventieth birthday of Robert Finn. Cambridge, MA: International Press. 288--298 (1995; Zbl 0854.53012)], \textit{H. C. Wente} asked whether any annular solution to the partitioning problem in the ball (i.e. any embedded free boundary CMC annulus in \(\mathbb{B}^3\)) should be rotational.\N\NIn the paper under review the authors construct a family of embedded free boundary CMC annuli in \(\mathbb{B}^3\) which are not Delaunay surfaces. More precisely, they prove the following elegant result: For any natural number \(n>1\), there exists a real analytic family \(\{\mathbb{A}(\mu):\mu\in[0,\varepsilon)\}\) of compact CMC annuli in \(\mathbb{R}^3\) such that:\N\begin{itemize}\N\item[1.] \(\mathbb{A}(\mu)\) is contained in the closed unit ball \(\mathbb{B}^3\) with \(\partial\mathbb{A}(\mu)\subset\partial\mathbb{B}^3\).\N\item[2.] \(\mathbb{A}(\mu)\) is embedded.\N\item[3.] \(\mathbb{A}(\mu)\) is free boundary.\N\item[4.] Each \(\mathbb{A}(\mu)\) has a family of spherical curvature lines.\N\item[5.] Each \(\mathbb{A}(\mu)\) is symmetric with respect to the \(x_3=0\) plane, and with respect to \(n\) equiangular vertical planes containing the \(x_3\)-axis.\N\item[6.] \(\mathbb{A}(0)\) is a compact piece of a free boundary nodoid in the unit ball. If \(\mu>0\), \(\mathbb{A}(\mu)\) is not rotational, and has a prismatic symmetry group of order \(4n\).\N\end{itemize}\N\NThis gives a negative answer to the uniqueness problem by Wente, providing the first known annular solutions to the partitioning problem in \(\mathbb{B}^3\) that are not rotational. It also shows that the topological uniqueness for annular solutions to the partitioning problem formulated by \textit{J. C. C. Nitsche} in [Arch. Ration. Mech. Anal. 89, 1--19 (1985; Zbl 0572.52005)] does not hold, even in the embedded case.