Legerdemain in Mathematical Physics: Structure, Tricks, and Lacunae in Derivations of the Partition Function of the Two-Dimensional Ising Model and in Proofs of The Stability of Matter

We review various derivations of the partition function of the two-dimensional Ising Model of ferromagnetism and proofs of the stability of matter, paying attention to passages where there would appear to be a lacuna between steps or where the structure of the argument is not so straightforward. Authors cannot include all the intermediate steps, but sometimes most readers and especially students will be mystified by such a transition. Moreover, careful consideration of such lacunae points to interesting physics and not only mathematical technology. Also, when reading the original papers, the structure of the physics argument may be buried by the technical moves. Improvements in the derivations, in subsequent papers by others, may well be clearer and more motivated. But, there is remarkably little written and published about how to read some of the original papers, and the subsequent ones, yet students and their teachers would often benefit from such guidance. I should note that much of the discussion below will benefit from having those papers in front of you.

💡 Research Summary

The paper undertakes a systematic examination of two cornerstone topics in mathematical physics: the exact evaluation of the partition function for the two‑dimensional Ising model and the rigorous proofs of the stability of matter. Although these subjects have been treated in classic works—most notably by Onsager, Kaufman, Lieb, and Thirring—the original publications often contain steps that are either omitted or presented with minimal justification. The authors identify these “lacunae” and trace how later contributions filled the gaps, thereby revealing deeper physical insights that are obscured by technical formalism.

In the first part, the authors revisit the transfer‑matrix approach introduced by Kramers and Wannier and later refined by Onsager. They point out that Onsager’s derivation of the eigenvalues of the row‑to‑row transfer matrix relies on a complex‑plane contour deformation and a normalization constant that are never explicitly constructed. The paper explains why the contour can be deformed without crossing singularities, invoking analyticity of the integrand and the periodic boundary conditions. It then discusses Kaufman’s algebraic method, which sidesteps the contour but leaves the combinatorial interpretation of spin pairings vague. By reviewing Kasteleyn’s orientation of planar graphs and the Pfaffian formulation, the authors demonstrate how the missing sign conventions are resolved through topological arguments involving the Euler characteristic of the lattice. They also show how the critical temperature emerges from a degeneracy of the transfer‑matrix spectrum, linking this to the scale‑invariance observed at the phase transition.

The second part addresses the stability of matter, i.e., the theorem that the total energy of an arbitrarily large collection of electrons and nuclei is bounded below by a constant times the number of particles. The original Dyson–Lenard proof is dissected to expose the step where the electron–electron Coulomb repulsion is bounded by a sum of electron–nucleus terms; the authors clarify the underlying inequality by invoking the Hardy inequality and a clever rearrangement of charges. The paper then moves to the Lieb–Thirring inequality, which provides a more transparent route by exploiting the antisymmetry of fermionic wavefunctions. Detailed derivations are given for the Sobolev embedding that justifies the exponent γ = 3/2 in three dimensions, and the role of the exchange term is highlighted as a direct consequence of Pauli exclusion. Recent refinements—such as Nam’s optimal constants and the Hardy‑Lieb–Thirring inequality—are presented, showing how they tighten the lower bound and why the electron density must belong to specific L^p spaces to make the estimates work.

Throughout both sections, the authors emphasize that the apparent “gaps” are not merely editorial oversights; they reflect deep connections between symmetry (spin reversal, lattice rotations), topology (planar graph orientations), and functional analysis (Sobolev spaces, spectral estimates). By making these connections explicit, the paper provides a pedagogical roadmap for graduate students and a reference for researchers seeking to generalize the classic results to more complex models, such as anisotropic lattices, quantum spin systems, or relativistic many‑body Hamiltonians. The concluding discussion argues that mastering these lacunae equips readers with a richer conceptual toolkit, turning what once seemed like mysterious jumps into opportunities for further theoretical development.