Dehn surgery and (1,1)-knots in lens spaces (Q876547): Difference between revisions
From MaRDI portal
Latest revision as of 16:40, 25 June 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Dehn surgery and (1,1)-knots in lens spaces |
scientific article |
Statements
Dehn surgery and (1,1)-knots in lens spaces (English)
0 references
18 April 2007
0 references
A knot in the \(3\)-sphere is called doubly primitive if the knot lies on a genus two Heegaard surface \(S\) of the \(3\)-sphere and represents a free generator of both the fundamental groups of the genus two handlebodies bounded by \(S\). This notion was introduced by John Berge in an unpublished manuscript [Some knots with surgeries yielding lens spaces]. From its definition, any doubly primitive knot admits an integral surgery, called Berge's surgery, yielding a lens space. It is conjectured that doubly primitive knots are the only knots that admit Dehn surgery yielding lens spaces. Consider the lens space \(L\) obtained from Berge's surgery on a doubly primitive knot \(K\) in the \(3\)-sphere. Let \(K^*\) be the dual knot, that is, the core of the attached solid torus, in \(L\). Thus \(K^*\) admits Dehn surgery yielding the \(3\)-sphere. Berge showed that \(K^*\) is a \((1,1)\)-knot in \(L\), and moreover, \(K^*\) can be put in some canonical form with respect to a Heegaard splitting of \(L\). Such a canonical form is parameterized by a single integer \(u\), so \(K^*\) is represented as \(K(L;u)\). This claim is also proved in the appendix of the paper under review. Conversely, the author gives a handy necessary condition for Dehn surgery on a \((1,1)\)-knot \(K(L;u)\) in \(L\) to yield the \(3\)-sphere. (But the condition is not sufficient.) The proof uses the wave theorem by \textit{T. Homma, M. Ochiai} and \textit{M. Takahashi} [Osaka J. Math. 17, 625--648 (1980; Zbl 0454.57002)] on genus two Heegaard diagrams of the \(3\)-sphere. As a corollary, it is shown that a lens space with fundamental group of order less than \(18\) cannot be obtained by Berge's surgery on a non-torus doubly primitive knot. This is a strong evidence toward a conjecture by \textit{S. A. Bleiler} and \textit{R. A. Litherland} [Proc. Am. Math. Soc. 107, No. 4, 1127--1131 (1989; Zbl 0686.57007)]. Unfortunately, as the author writes, it remains possible that a doubly primive knot admits another surgery other than Berge's surgery, yielding a lens space.
0 references
Dehn surgery
0 references
doubly primitive knot
0 references
lens space
0 references
dual knot
0 references
\((1
0 references
1)\)-knot
0 references
0 references