A Counterexample to Corzatt's Four-Direction Conjecture
Year: 2026
Type: preprint
Abstract: Corzatt conjectured that if the lattice points of a convex body in the plane can be covered by n lines, then they can be covered by n lines having at most four different directions. We show that this is false. The 35 lattice points of the hexagon with vertices (0, 0), (1, 0), (8, 2), (8, 3), (6, 7) and (5, 8) are covered by the eight lines y = 2, y = 3, x − y = 0, x − y = 2, x + y = 11, 2x − y = 2, 2x − y = 3 and 3x − y = 11, which have five directions. No seven lines cover these points and no other eight lines do, so every cover by eight lines needs five directions; with four directions, nine lines are needed. The proof is a finite computation that three elementary lemmas make small, and a standalone exact verifier repeats it in under half a minute. The example was found by growing lattice polygons one point at a time from a minimal polygon of lattice width eight, while tracking how a cover of the final set restricts to the starting polygon. The decisive step was to follow restrictions that cannot be shortened but are not optimal. A second counterexample, inequivalent to the first under lattice-preserving affine maps, shares seven of its eight lines.
Author: Seamons, John
Language: English
Subfield: Discrete Mathematics and Combinatorics
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225912864
