Does \(''\Delta ^ 2r^{2-n}=0\) on a Riemannian manifold'' imply flatness ? (Q1083697)

At the beginning, the following simple result is proved: Let (M,g) be an n-dimensional connected analytic Riemannian manifold with \(n>2\) and let \(m\in M\). Denote by r the distance from p to m, where \(p\in M\) is sufficiently close to m. Finally, suppose \(\Delta r^{\ell}=0\) for \(\ell \neq 0\), all \(m\in M\), and all sufficiently small r. Then \(\ell =2-n\) and (M,g) is locally flat. The first author investigated, in a joint paper, the similar condition for the iterated Laplacian: \(\Delta^ 2r^{\ell}=0\), \(\ell \neq 0\). He proved that the only essential cases were \(\ell =2\), \(\ell =4-n\), or \(\ell =2-n\). The case \(\ell =4-n\) has been solved completely, the case \(\ell =2\) remains open, and the case \(\ell =2-n\) is conjectured to imply local flatness. The main purpose of the present paper is to prove the last conjecture under some additional assumptions, e.g., if (M,g) is locally a direct product, or 3-dimensional, or locally symmetric. The authors also show how the equation \(\Delta^ 2r^{2-n}=0\) is related to the theory of harmonic spaces.