Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated


Source. Conjectures 2 and 3 and the passage between them, p. 560; the reformulations on pp. 564 and 565; of P. Erdős, R. L. Graham, P. Montgomery, B. L. Rothschild, J. Spencer and E. G. Straus, Euclidean Ramsey Theorems, III, Infinite and Finite Sets (Keszthely 1973), Colloq. Math. Soc. János Bolyai 10, North-Holland (1975), 559--583, as identified on the source card.

Statement

Conjecture 2 (p. 560), as posed: "If E2E^2 is 2-colored so that there is no equilateral triangle of side dd, then there is a monochromatic equilateral triangle of side d′d', for d′≠dd'\ne d." The hypothesis is read with "monochromatic" understood, as the surrounding text and the reformulation on p. 565 show: the paper restates Conjecture 2 as saying that Tf={(a,a,a)}T_f=\{(a,a,a)\} for some a>0a>0, or Tf=∅T_f=\emptyset, so the conclusion is for every d′≠dd'\ne d.

Conjecture 3 (p. 560), as posed: "If KK is a triangle which is not equilateral, then R(K)R(K) is true."

The notation R(K)R(K), Rf(K)R_f(K) and TfT_f is that of Theorem 1. The paper derives Conjecture 3 from Conjecture 2 through Theorem 1 and states that the two are equivalent in view of Theorem 1 (p. 560); it restates Conjecture 3 as Tf⊂{(a,a,a)∣a>0}T_f\subset\{(a,a,a)\mid a>0\} for every two-coloring ff (p. 564). It notes (p. 582) that it would suffice to prove Conjecture 3 for the non-equilateral isosceles triangles, and that it has no (a,a,b)(a,a,b)-triangle with R(a,a,b)R(a,a,b) and a/ba/b transcendental.

Status in the paper

Posed, not proved. The paper proves R(K)R(K) for several families of triangles (Theorem 9, Corollary 10, Corollary 15, Theorem 17), shows that the exceptional set TfT_f is totally disconnected (Theorem 5), and shows that minimal finite witnesses for the (1,1,x)(1,1,x)-triangle grow without bound as x→1x\to1 (Theorem 28).

Bears on

  • Problem 173: Conjecture 2 is the problem's statement in the paper's terms, that a two-coloring misses at most one triangle (which must then be equilateral), and Conjecture 3 is equivalent to it by Theorem 1. The page poses the question and settles it for no coloring.