Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Source. Theorem 1, p. 447, of Kurt Mahler, On the lattice points on curves of genus 1, Proc. London Math. Soc. (2) 39 (1935), 431--466, the edition named on the source card. The form is the one fixed in Section 1, p. 432.
Read depth. Claims checked: the statement and the hypotheses on were read clause by clause on pp. 432 and 447, the proof for its structure. Nothing here is independently reviewed.
Statement
Let
be a cubic binary form with integer coefficients and with only simple linear factors (Section 1, p. 432).
Theorem 1 (p. 447). For every there is a positive number such that for every integer there is an integer with
which represents in at least different ways, (), with integers .
So the number of integer solutions of , finite for each by Thue's theorem when is irreducible (p. 431), is not bounded. The introduction (p. 431) states that there are infinitely many with . For the forms and with positive variables, the paper proves bounds of this shape as Theorem 6 and Theorem 7, through the angle version Theorem 5. The introduction also says that the paper cannot prove similar results for primitive solutions ( coprime), and that whether their number is bounded is still open (pp. 431--432).
Proof pointer
Sections 1--14, pp. 432--447. The curve has genus 1 and is uniformized by elliptic functions; from a point with elliptic argument the chord-and-tangent construction gives the points with arguments and , . Sections 4--7 choose a lattice point with off finitely many lines through the origin, so that on the similar curve , , the points built from it are distinct and finite; then (Section 14). Sections 8--13 bound the denominators of their coordinates by a recursion; with the least common multiple of the denominators, the points scaled by are lattice points on , and the bounds give for . Put .
Dependencies
None outside the paper's own Sections 1--14.
Bears on
- Problem 829: only through Theorem 6, on sums of two cubes of positive integers, which the paper derives from the angle version Theorem 5; that page states the relation. Theorem 1 itself counts solutions in integers of either sign and proves no upper bound.