Sums of numbers with small partial quotients. II (Q5960981)

For a positive integer \(m\) let \(F(m)=\{[t,a_1,a_2,\dots];t\in\mathbb Z\), \(1\leq a_i\leq m\), \(i\geq 1\}\), where \([a_0,a_1,a_2,\dots]\) is the continued fraction. Define the sum of two sets of real numbers \(A\) and \(B\) by \(A+B=\{a+b;a\in A, b\in B\}\). Denote \(A-B\) by \(A+(-B)\), where \(-B=\{-b;b\in B\}\). \textit{M. Hall jun.} [Ann. Math. (2) 48, 966-993 (1947; Zbl 0030.02201)] proved that \(F(4)+F(4)=\mathbb R\). \textit{B. Diviš} [Acta Arith. 22, 157-173 (1973; Zbl 0254.10026)] and \textit{T. W. Cusick} [Proc. Am. Math. Soc. 38, 253-254 (1973; Zbl 0259.10028)] proved independently that \(F(3)+F(3)\neq\mathbb R\). \textit{J. L. Hlavka} [Trans. Am. Math. Soc. 211, 123-134 (1975; Zbl 0313.10032)] proved that \(F(m)+F(n)=\mathbb R\) holds for \((m,n)\) (\(m\neq n\)) equal to \((2,7)\) or \((3,4)\), but does not hold for \((m,n)=(2,4)\). He conjectured that it would not hold for both \((m,n)=(2,5)\) and \((m,n)=(2,6)\). The author disproves Hlavka's conjecture, showing that \(F(5)\pm F(2)=\mathbb R\). For Part I, cf. Proc. Am. Math. Soc. 130, 637-642 (2002; Zbl 0992.11044).