Wiki
Wiki

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

Updated


Statement

Setting as in inequality (1): kk points with integer coordinates 0<xi,yi≤n0<x_i,y_i\le n and all mutual distances distinct.

Inequality (2) (p. 121). There is a positive constant c3c_3 with

k<c3 n (log⁡n)−1/4.k<c_3\,n\,(\log n)^{-1/4}.

The print states no range of nn for (2); it says only that each cic_i is a positive constant.

Conjecture (3) (p. 121). The paper marks with "(?)" the conjecture

k<c4 n2/3(log⁡n)1/6,k<c_4\,n^{2/3}(\log n)^{1/6},

which it says a heuristic argument supports, adding that the argument "lacks conviction since the corresponding argument in one dimension gives a false result". The heuristic argument is not given.

Proof pointer

p. 121. Landau's theorem (Handbuch, 1909) says that the number of integers less than xx that are sums of two squares is asymptotically c1x(log⁡x)−1/2c_1x(\log x)^{-1/2}. Every squared distance in the grid is such an integer below 2n22n^2, so the right side of (1) may be replaced by c2n2(log⁡n)−1/2c_2n^2(\log n)^{-1/2}, and (k2)\binom k2 below that bound gives (2).

Read depth

Claims checked: (2), (3) and the derivation of (2) were read clause by clause on the page image of p. 121. Nothing here is independently reviewed.

Dependencies

Landau's asymptotic for integers that are sums of two squares, cited from E. Landau, Handbuch der Lehre von der Verteilung der Primzahlen (Leipzig, 1909), II, 643.

Source. P. Erdős, R. K. Guy, Distinct distances between lattice points, Elem. Math. 25 (1970), 121--123; the edition read is named on the source card.

Bears on

  • Problem 1208: the N=n2N=n^2 points of the grid are one set of NN points in the plane, so for each nn at which (2) holds every distinct-distance subset of them has fewer than c3n(log⁡n)−1/4c_3n(\log n)^{-1/4} points, which gives F2(n2)<c3n(log⁡n)−1/4F_2(n^2)<c_3n(\log n)^{-1/4}, that is F2(N)F_2(N) is O(N1/2(log⁡N)−1/4)O(N^{1/2}(\log N)^{-1/4}) along the squares. This is an upper bound only; the paper does not state it in terms of F2F_2 and gives no lower bound for F2F_2. Conjecture (3) concerns the grid, not arbitrary sets.