A better bound for the excentricities not admitting the equichordal body (Q912389)

The equichordal problem of Fujivara (1916) and Blaschke, H. Rothe and Weitzenböck (1917) asks for the existence of a convex curve (called equichordal curve) containing two e-points \(0,0'\) (an e-point is interior to the curve and the chords through it are of constant length). \textit{G. A. Dirac} [J. Lond. Math. Soc. 27, 429-437 (1952; Zbl 0047.158)] gave necessary conditions for the existence of such a curve. The assumption of convexity was dropped by \textit{E. Wirsing} [Arch. Math. 9, 300-307 (1958; Zbl 0083.384)]. If the length of the chords is normalized to 1, the distance \(a=| 00'|\), \(0<a<1\), is the excentricity of the problem. \textit{E. Ehrhart} [Enseign. Math., II. Ser. 13, 119-124 (1967; Zbl 0153.518)] proved that for \(a\geq 0.5\) there is no equichordal curve. The first author [Boll. Unione Mat. Ital., VII. Ser. A 2, No.2, 203-211 (1988; Zbl 0652.52001)] proved non-existence of the equichordal curve for \(a\geq 0.33.\) In this paper the authors improve a necessary condition due to Dirac and show non-existence of the curve for \(a\geq 0.235\).