“…The main difference is that the Newton polytope of a polynomial does not contain the origin as an interior point and that we have to deal with a non-complete situation. This can be applied for instance to the Mordell-Pommersheim tetrahedron, that is the convex hull of (0, 0, 0), (a, 0, 0), (0, b, 0), (0, 0, c)) where a, b and c are positive integers, which is the Newton polytope of the Brieskorn-Pham polynomial In this context, other questions may arise: for instance, singularity theory predicts that the variance of the spectrum at infinity of a Laurent polynomial in (C * ) n is bounded below by n/12 (this is a global variant of C. Hertling's conjecture [9], see [6] and the references therein). By Proposition 5.1, this would give information about the distribution of the δ-vector of a reflexive polytope.…”