Surfaces of constant mean curvature bounded by two planar curves (Q1361041)

As indicated by the title, the author considers surfaces of constant mean curvature (CMC) bounded by two planar curves. The best known examples of such surfaces are given by portions of Delaunay surfaces (the rotationally symmetric CMC surfaces) bounded by circles in parallel planes. Among these, the most interesting class for this paper are the nodoids. Nodoids have a periodic meridian curve with no inflections. The meridian curve thus consists of loops and has periodic self-intersections. If one keeps the various portions of a nodoid in mind, then the statements of the results in the paper are quite natural. A typical result (Theorem 3.1) is as follows. Let \(M\) be an embedded CMC surface bounded by Jordan curves in parallel planes \(P_1\) and \(P_2\). If \(M\) does not intersect either exterior domain bounded by the curves and the distance between the planes is greater than \(1/| H| \) where \(H\) is the mean curvature, then \(M\) lies in the slab between the two planes. This can be checked (explicitly) for the outer loops of nodoids (with the separation distance becoming sharp for the hemispheres). The author extends the property to a fairly general class of surfaces (as one can get some idea of by imagining the two planes not quite parallel to the axis of the nodoid). The proofs use Alexandrov reflection techniques and a balancing formula. Similar techniques were used in [\textit{R. Sa Earp, F. Brito, W. H. Meeks III} and \textit{H. Rosenberg}, Indiana Univ. Math. J. 40, 333-343 (1991; Zbl 0759.53003)]. The basic use of reflection is illustrated in the second Section where two symmetry results (one for vertical and one for horizontal symmetry) are proved. In the following sections, the hypothesis of non-intersection with exterior domains (mentioned in Theorem 3.1 above) is introduced and the balancing formula is employed. Though the paper has no direct bearing on the conjecture of Shiffman and Meeks as discussed in [\textit{R. Schoen}, J. Differ. Geom. 18, 791-809 (1983; Zbl 0575.53037)], the arguments are of interest in this regard. As a final note, the reader should beware of references to the example of Figure 1. The author first refers to the example as an embedded surface near the top of page 202. The indicated surface, however, clearly has self-intersections. This can be remedied by lowering the plane \(P_2\) to a level below the self-intersection (and sliding the point that generates the boundary curve \(\Gamma_2\) along the meridian accordingly). In the reference to Figure 1 near the bottom of page 205 the author presumably intends to express that there are embedded CMC surfaces which do not intersect ext(\(\Omega_i\)) \textit{and} are not included in \(S\). The reader should have no trouble visualizing such an example on his own.