Wiki
Wiki

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

Updated


Statement

Setting: bands and Mmax⁡(k)=k(n−k)+(k2)+1M_{\max}(k)=k(n-k)+\binom k2+1 as in Lemma 1; [x][x] is the integer part of xx.

Lemma 4 (p. 133). For nn sufficiently large, every configuration of nn points in a band with k>[n+2]k>[\sqrt{n+2}] determines more than Mmax⁡([n+2]−1)M_{\max}([\sqrt{n+2}]-1) lines.

No explicit threshold for nn is given.

Proof pointer

P. 133. The Kelly--Moser lower bounds Mmin⁡(k)M_{\min}(k) of Lemma 1 increase up to k=[(n+0.5)/3]k=[(n+0.5)/3] and, for k>[n+2]k>[\sqrt{n+2}], exceed Mmax⁡([n+2]−1)M_{\max}([\sqrt{n+2}]-1), so only bands with k>n/3k>n/3 remain. For those, the theorem of Beck that the paper cites (a configuration with k≥xk\ge x determines more than c x(n−x)c\,x(n-x) lines, cc absolute; p. 131) gives at least c(n/3)(2n/3)c(n/3)(2n/3) lines, which exceeds Mmax⁡([n+2]−1)M_{\max}([\sqrt{n+2}]-1), of order n3/2n^{3/2}, once nn is large.

Read depth

Claims checked: the statement was read clause by clause on the page image of the print, and the proof on p. 133 was followed. Beck's theorem is cited, not proved, in the paper and was not read. Nothing here is independently reviewed.

Dependencies

Lemma 1 for the lower bound Mmin⁡(k)M_{\min}(k). External input named by the paper: J. Beck, On the lattice property of the plane and some problems of Dirac, Motzkin and Erdős in combinatorial geometry, Combinatorica 3 (1983), 281--297, with Szemerédi and Trotter, Combinatorica 3 (1983), 381--392.

Source. P. Salamon and P. Erdős, The solution to a problem of Grünbaum, Canad. Math. Bull. 31 (1988), no. 2, 129--138, DOI 10.4153/CMB-1988-020-2; the edition read is named on the source card.

Bears on

  • Problem 606: the lemma is the reason the bands beyond k=[n+2]k=[\sqrt{n+2}] contribute nothing below the continuum in the paper's answer for large nn, described on the main result page. The lemma gives no threshold for nn, and that answer is stated only for n≥n∗n\ge n^*, with n∗n^* unknown (p. 137).