“…The main idea is now that, for the question of boundedness, each Bloch function u n (ξ, x) can be replaced by the corresponding plane wave exp(ix • ξ). Arguments of this sort have been carried out in a similar context in [37,3,2,8], for instance. When ∇ ξ λ n (ξ) = 0 for all ξ such that λ n (ξ) = µ, the Fermi surface is smooth and can be locally replaced by a sphere.…”