Mathematicians overturn 150 year geometry rule using torus surfaces

Saturday 14 March 2026 - 07:00

By: Dakir Madiha

Mathematicians overturn 150 year geometry rule using torus surfaces

A team of mathematicians has demonstrated that a long-standing principle in differential geometry proposed more than 150 years ago is not universally valid. By constructing two distinct torus-shaped surfaces that share identical geometric measurements and curvature properties, the researchers showed that these local data do not always determine a unique global shape.

The original principle was formulated in 1867 by French mathematician Pierre Ossian Bonnet. It stated that the geometry of a compact surface could be uniquely determined by two pieces of information: its metric, which describes distances between points on the surface, and its mean curvature, which measures how strongly the surface bends in space.

Researchers from several universities have now identified a counterexample to this rule. Alexander Bobenko of the Technical University of Berlin, Tim Hoffmann of the Technical University of Munich, and Andrew O. Sageman-Furnas of North Carolina State University produced two compact torus surfaces embedded in three-dimensional space that share the same metric and the same mean curvature function while remaining geometrically distinct.

Their findings were published in Publications Mathématiques de l’IHÉS. The study introduces the first confirmed examples of what mathematicians call “compact Bonnet pairs,” meaning two closed surfaces that match in both metric and curvature data but are not congruent.

Previous exceptions to Bonnet’s rule were already known, but they appeared only in noncompact surfaces, which either extend infinitely or include boundaries. For compact surfaces such as spheres, mathematicians had demonstrated that the rule held true. In the case of torus surfaces, earlier theoretical work had shown that at most two distinct surfaces could share the same metric and mean curvature, yet no explicit pair had ever been constructed.

According to Hoffmann, the discovery resolves a problem that had remained open for decades in the study of surface geometry. The work shows that even for closed shapes resembling a doughnut, local geometric measurements are not always sufficient to determine a single global form.

The breakthrough emerged from an unconventional research approach that combined discrete geometry and computational exploration. As reported by Quanta Magazine, Sageman-Furnas initially searched for compact Bonnet pairs among discrete surfaces, simplified approximations of smooth shapes that can be analyzed through computer calculations.

A computational search conducted in 2018 identified a promising discrete torus configuration. The team then used that structure as a blueprint to develop a smooth analytic version of the surface.

One key idea was to constrain the lines of curvature so they lie either within planes or on spheres. This strategy builds on classical work by the nineteenth century mathematician Gaston Darboux. After years of combining theoretical analysis with computer-assisted experimentation, the researchers succeeded in producing smooth tori whose curvature lines close correctly.

The construction also resolves a related open question posed in 1929 by Wilhelm Cohn-Vossen and later highlighted by mathematician Marcel Berger in 2010. That problem asked whether a compact surface immersion could be uniquely determined by the real analytic properties of its metric alone.

Despite the advance, questions remain. The two torus surfaces identified by the team intersect themselves in space, leaving open the challenge of finding Bonnet pairs that do not self-intersect. Bobenko hopes future work will produce such examples.

Robert Bryant of Duke University noted that the absence of concrete examples for so long had led many mathematicians to believe compact Bonnet pairs might not exist. As Bryant told Quanta Magazine, the assumption persisted largely because no one had succeeded in constructing a clear example.