Conway irreducible hyperbolic knots with two common covers (Q2388661)

[\textit{B. Zimmerman}, Topology Appl.79, No, 2, 143--157 (1997; Zbl 0888.57006)] has shown that if \(n>m>2\) are not coprime, hyperbolic knots are determined by their \(m\)-fold and \(n\)-fold cyclic branched covers. If the knots are \(\pi\)-hyperbolic, the result is true for \(m=2\) as well. A link \(L=L_1\cup\cdots\cup L_k\) is \((2\pi/n_1,\dots,2\pi/n_k)\)-hyperbolic if the orbifold with underlying space \(S^3\) and singular set \(L\) is hyperbolic and the order of the singularity of \(L_i\) is \(n_i\). If \(n_i=n\) for all \(i,L\) is called \(2\pi/n\)-hyperbolic. The author constructs a specific \(\pi\)-hyperbolic, three component link, \(L\). If \(n,m\geq 2\) are coprime, a cover of \(S^3\) ramified of order \(n\) along one component of \(L\) (again \(S^3)\), followed by a cover of the resulting \(S^3\) ramified of order \(m\) along the lift of another component of \(L\), results in the final lift of the third component being a knot in \(S^3\). Two knots are constructed in this manner giving the first main result of this paper: If \(n,m\geq 2\) are coprime there exists a pair of non-equivalent Conway irreducible hyperbolic knots with the same \(n\)-fold and \(m\)-fold cyclic branched covers. The author had previously found a gap in a claim that Conway irreducible hyperbolic knots are determined by their 2-fold and \(n\)-fold cyclic branched covers for any \(n>2\). The above result shows this is not true. However, the result can be obtained if the hypothesis is strengthened to include that the knots are \(\pi\)-hyperbolic and one of them admits a unique Seifert surface up to isotopy.