Bounds for the Newton number in the plane (Q1601401)

The Newton number \(N(D)\) of a compact body \(D\) in \(E^d\) is the maximum number of non-overlapping congruent copies of \(D\) that can touch \(D\) without overlapping \(D\). (The bodies might touch at only 1 point, e.g., two squares touching only at a vertex.) Newton, in 1694, proposed that \(N(B^3)= 12\) for the ball \(B^3\), whereas Gregory conjectured that \(N(B^3)= 13\). Newton was right! Even today \(N(B^d)\) is known only for \(d= 2,3,8,24\), and not much is known about \(N(D)\) in the plane. If \(P_n\) denotes the regular \(n\)-gon, it is known that \(N(P_3)= 12\), \(N(P_4)= 8\) and \(N(P_n)= 6\) for \(n\geq 5\). In the paper under review the authors prove that, for an arbitrary compact \(D\) in the plane, \(N(D)\geq 6\). (This inequality was proved by Hadwiger for convex \(D\).) Also, they determine upper and lower bounds for the Newton number of isosceles triangles. The proofs are constructive, elegant and elementary; the bibliography is good and the exposition is pleasant.