Pith. sign in

REVIEW 3 major objections 7 minor 35 references

On rank two Kleinian groups with three parabolics

T0 review · 3 major / 7 minor · reviewed 2026-07-09 · glm-5.2

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Three parabolics break the two-bridge link mold

desk verdict Novel example of a lattice in the parabolic locus, but the SnapPy verification in Theorem 3.6 has a gap read the letter →

arxiv 2607.07424 v1 pith:NPRUTRGS submitted 2026-07-08 math.GR math.GT

Alex Elzenaar This is my paper · ORCID
classification math.GRmath.GT MSC 20H1022E4030F4057K32
keywords KleiniangroupsparabolicrepresentationsgenustwohandlebodytunnelnumberonelinksmaximalcuspSchottkyspacecharactervarietyBorromeanrings
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper studies what happens when you take a genus-two handlebody — the natural home of a free group on two generators — and ask for discrete representations into the isometry group of hyperbolic 3-space that send three disjoint simple closed curves on the boundary surface to parabolic isometries. The motivating question is whether the well-understood theory for two parabolic generators (where the only finite-covolume groups come from two-bridge links and their quotients) extends cleanly to the general rank-two case with three parabolics. The author proves two main results. First, if such a representation has a non-empty domain of discontinuity (meaning the group acts nicely on part of the Riemann sphere), then it must be a maximal cusp group on the boundary of genus-two Schottky space — confirming the expected generalization of the infinite-covolume picture. Second, and unexpectedly, there exist finite-covolume discrete representations sending all three curves to parabolics that do not arise from tunnel number one links at all, even up to finite-index subgroups. The concrete example is a 3-fold quotient of the Borromean rings complement, constructed from an embedding of the 8_5 knot on the genus-two surface. This demonstrates that the structure of the parabolic locus is genuinely more complicated than the two-parabolic-generator setting would suggest.

What carries the argument

The argument combines several tools: (1) topological inequalities due to Bers and others bounding the Euler characteristic of the boundary and the number of cusps/singularities of the quotient orbifold in terms of the group rank, which force the group to be a maximal cusp when the domain of discontinuity is non-empty; (2) explicit trace coordinates on the character variety of the free group F_2, using the parametrisation by traces of generators and their product; (3) a computational identification strategy where the author guesses the orbifold using classification results for groups generated by two elliptics or parabolics, then verifies the guess independently by matching trace parameters;

What would settle it

If the cited classification of discrete cofinite groups generated by two elliptics or parabolics turns out to be incomplete or incorrect in a way that excludes the orbifold type the author identifies, the identification of the example as a 3-fold quotient of the Borromean rings would need re-verification through an alternative route.

Watch

Extended reading notes

Core claim

The central discovery is that the classification of discrete representations of the rank-two free group into PSL(2,C) sending three disjoint curves to parabolics does not reduce to the two-bridge link story. Specifically, the author proves that infinite-covolume groups with non-empty domain of discontinuity are forced to be maximal cusp groups (Theorem 2.5), but then exhibits a finite-covolume lattice — concretely, a 3-fold quotient of the Borromean rings complement arising from the 8_5 pretzel knot — where all three curves are parabolic yet the group is not virtually a tunnel number one link group (Theorem 3.6). The mechanism carrying the proof of Theorem 3.6 is the identification of an orb

Load-bearing premise

The proof of Theorem 3.6 initially identifies the quotient orbifold by appealing to a classification of discrete cofinite groups generated by two elliptics or parabolics that is cited as 'To appear'; while the author notes the final verification is independent (comparing trace parameters), the initial identification step depends on that external classification being complete and correct.

Editorial extensions

If this is right

  • The existence of finite-covolume groups in the parabolic locus that are not tunnel number one link groups means any complete classification of discrete representations in Par(α,β,γ) must accommodate a broader class of orbifolds than previously expected.
  • The intersection of parabolic loci for distinct curve systems (Theorem 3.8) suggests the deformation space structure for rank-two groups with three parabolics has a richer intersection pattern than the two-parabolic case, where pleating rays do not collide.
  • The computational identification strategy — guess the orbifold using classification theory, then verify by matching traces — provides a practical template for solving similar identification problems for other curve systems on the genus-two surface.
  • The rarity of such finite-covolume examples (the author notes they do not appear often in computer experiments) raises the question of what special geometric or algebraic conditions on the curve system are necessary for them to exist.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The 3-fold cover being the Borromean rings complement (which has tunnel number 2 and rank 3 fundamental group) while the quotient orbifold group is rank 2 suggests a nontrivial relationship between the tunnel number of the cover and the rank of the quotient group that may constrain which curve systems can produce such examples.
  • The fact that the curve system has only Z/2 x Z/2 symmetry yet the orbifold has an order-3 cone axis raises the question of whether the order of the cone axis is determined by the algebraic structure of the words representing the curves, rather than by the symmetry of the embedding.
  • A systematic search over pretzel knots P(a,2,c) for small odd a and c could test whether finite-covolume non-link groups appear only for specific arithmetic conditions on a and c, or whether they are more broadly distributed.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 7 minor

Summary. The paper studies discrete representations of the free group of rank 2 into PSL(2,C) such that three disjoint simple closed curves on the genus-2 boundary surface are sent to parabolic elements. Two main results are proved: (1) Theorem 2.5, stating that if the domain of discontinuity is non-empty and all three curves are parabolic, then the group is a maximal cusp group on the boundary of genus-2 Schottky space; (2) Theorem 3.6, exhibiting a finite-covolume representation in a parabolic locus that sends all three curves to parabolics and is not virtually a tunnel number one link group, realized concretely via a 3-fold quotient of the Borromean rings complement. A supporting result, Theorem 3.8, gives an example of a discrete group lying in the intersection of parabolic loci for two distinct curve systems.

Significance. The paper addresses a concrete instance of the identification problem for Kleinian groups, a classical and actively studied topic. Theorem 2.5 gives a clean generalization of the two-parabolic-generator theory to the full rank-2 setting, and its proof via the inequalities of Lemmas 2.2–2.4 is elegant and self-contained. Theorem 3.6, if correct, would be genuinely surprising: it would show that the parabolic locus contains finite-covolume groups not arising from tunnel number one link complements, contradicting a natural conjecture. The explicit construction via the 8_5 knot (pretzel P(3,2,3)) and the connection to the Borromean rings is concrete and falsifiable. The author is transparent about the computational nature of the Theorem 3.6 verification and about the dependence on the classification of Chesebro–Martin–Schillewaert [9] (to appear). Theorem 3.8 is a nice bonus result, carefully verified by matching three traces against the SnapPy census manifold m003(0,0).

major comments (3)
  1. Proof of Theorem 3.6: the algebraic identities 'X = AB and Y = ABABA' (with A = x²Y = X⁻²Y, B = yX = Y⁻¹X) are incorrect. Direct computation in the free group gives AB = X⁻²Y · Y⁻¹X = X⁻¹ (not X) and ABABA = X⁻⁴Y (not Y). The group equality G = ⟨A,B⟩ = ⟨X,Y⟩ does still hold (since AB = X⁻¹ yields X, and then Y = X²A), so the invocation of the classification [9] for groups generated by a parabolic and an elliptic is not invalidated. However, the stated formulas must be corrected, and any downstream computation that relies on them should be rechecked.
  2. Proof of Theorem 3.6, SnapPy verification: the author computes tr²(bAB) from SnapPy and describes it as 'the trace-squared of the product of the meridian generators.' The meridian generators are b and bA (as stated in the SnapPy output immediately above), so their product is b · bA = b²A, not bAB. The element actually computed (bAB) is different. The author should compute tr²(b²A) instead, or clarify what is being matched.
  3. Proof of Theorem 3.6, trace matching: the verification only checks one trace match — tr²(AB) = −1 from the abstract parameters (2) against tr²(bAB) = −1 from SnapPy. But the author then invokes the fact that 'rank two subgroups of PSL(2,C) are determined by the set of traces of two generators and their product,' which requires matching all three traces (tr²(g₁), tr²(g₂), tr²(g₁g₂)) for the same generating pair. From the parameters (2): tr²(A) = tr²(X⁻²Y) = 4 (parabolic), tr²(B) = tr²(Y⁻¹X) = 1 (elliptic of order 3), tr²(AB) = tr²(X⁻¹) = −1. The SnapPy elements b, bA, bAB have trace-squared values (1, 1, −1) — all elliptic. No SnapPy element with tr² = 4 is computed to correspond to the parabolic generator A. The identification between {A, B} and the SnapPy generators is therefore not established. The author must either (i) find the correct SnapPy elements corresponding to A and B and ver
minor comments (7)
  1. The paper cites [9] (Chesebro–Martin–Schillewaert) as 'To appear.' Since the classification result is used to narrow the candidate orbifolds in Theorem 3.6, the author should confirm its current status and, if possible, provide a more precise reference (e.g., preprint number or journal acceptance).
  2. In the SnapPy transcript on p. 10, the output matrices are displayed with truncated decimals. While the author notes this, it would help to state the precision used and confirm that the trace-squared values match to the expected numerical tolerance.
  3. Lemma 2.2 proof: the claim that 'Ab(π₁(N)) does not contain any finite order elements' could use a brief justification (e.g., N is a manifold, so π₁(N) is torsion-free).
  4. The footnote on p. 6 crediting a StackExchange post for the 'half lives, half dies' theorem is unusual for a journal paper; the standard reference is Hatcher's notes or the original source by Culler–Gordon–Luecke–Shalen.
  5. Figure 9: the caption 'The threefold cover of the 3/10 2-bridge link orbifold is the Borromean rings' could be more informative — it would help to indicate which covering map is intended.
  6. In Example 3.5, the two sets of trace parameters are labeled (2) and (3), but these labels are not introduced before they appear. A brief sentence orienting the reader would help.
  7. The abstract states 'hitherto unexpected finite covolume groups'; this is a strong claim that should be softened or contextualized (e.g., 'not predicted by the existing two-parabolic theory').

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for a careful reading and for identifying genuine errors in the proof of Theorem 3.6. The referee's first two comments are correct: the algebraic identities for X and Y in terms of A and B contain sign errors, and the SnapPy trace computation does not match the claimed generating pair. We will correct these. The third comment raises a more substantive issue about whether the trace identification is fully established. We address each point below.

read point-by-point responses

  1. Referee: Proof of Theorem 3.6: the algebraic identities 'X = AB and Y = ABABA' (with A = x²Y = X⁻²Y, B = yX = Y⁻¹X) are incorrect. Direct computation in the free group gives AB = X⁻²Y · Y⁻¹X = X⁻¹ (not X) and ABABA = X⁻⁴Y (not Y). The group equality G = ⟨A,B⟩ = ⟨X,Y⟩ does still hold (since AB = X⁻¹ yields X, and then Y = X²A), so the invocation of the classification [9] for groups generated by a parabolic and an elliptic is not invalidated. However, the stated formulas must be corrected, and any downstream computation that relies on them should be rechecked.

    Authors: The referee is correct. We have verified that AB = X⁻¹ (not X) and ABABA = X⁻⁴Y (not Y) by direct computation in the free group. The group equality G = ⟨A,B⟩ = ⟨X,Y⟩ does still hold, as the referee notes: from AB = X⁻¹ we recover X = (AB)⁻¹, and then Y = X²A = (AB)⁻²A. The correct expressions are X = (AB)⁻¹ and Y = (AB)⁻²A. We will replace the incorrect formulas in the proof of Theorem 3.6 with these corrected ones. We have rechecked all downstream computations: the trace values tr²(A) = 4, tr²(B) = 1, and tr²(AB) = −1 are unchanged because they depend only on A and B themselves (which are correctly defined), not on the erroneous expressions for X and Y in terms of A and B. The Farey word computation W_{3/10} = bAbaBabAbaBaBAbABaBA is expressed in terms of A and B directly and is unaffected. The corrected formulas will appear in the revised manuscript. revision: yes

  2. Referee: Proof of Theorem 3.6, SnapPy verification: the author computes tr²(bAB) from SnapPy and describes it as 'the trace-squared of the product of the meridian generators.' The meridian generators are b and bA (as stated in the SnapPy output immediately above), so their product is b · bA = b²A, not bAB. The element actually computed (bAB) is different. The author should compute tr²(b²A) instead, or clarify what is being matched.

    Authors: The referee is correct that the description 'the trace-squared of the product of the meridian generators' is inaccurate. The meridian generators reported by SnapPy are b and bA, so their product is b · bA = b²A, not bAB. The element bAB that we computed is instead the product of the first meridian b with the second meridian composed with the second meridian again — that is, bAB is a different element of the group. We will correct the description: bAB is not the product of the two meridian generators, but rather a specific element whose trace-squared we are matching against tr²(AB) from the abstract parameters. The numerical value tr²(bAB) = −1 from SnapPy does match tr²(AB) = −1 from the parameters (2), and this is the comparison we intend. However, as the referee's third comment correctly identifies, this single trace match is insufficient by itself to establish the full identification. We address that issue in our response to the third comment. revision: yes

  3. Referee: Proof of Theorem 3.6, trace matching: the verification only checks one trace match — tr²(AB) = −1 from the abstract parameters (2) against tr²(bAB) = −1 from SnapPy. But the author then invokes the fact that 'rank two subgroups of PSL(2,C) are determined by the set of traces of two generators and their product,' which requires matching all three traces (tr²(g₁), tr²(g₂), tr²(g₁g₂)) for the same generating pair. From the parameters (2): tr²(A) = tr²(X⁻²Y) = 4 (parabolic), tr²(B) = tr²(Y⁻¹X) = 1 (elliptic of order 3), tr²(AB) = tr²(X⁻¹) = −1. The SnapPy elements b, bA, bAB have trace-squared values (1, 1, −1) — all elliptic. No SnapPy element with tr² = 4 is computed to correspond to the parabolic generator A. The identification between {A, B} and the SnapPy generators is therefore not established. The author must either (i) find the correct SnapPy elements corresponding to A and B and ver

    Authors: The referee raises a valid and important objection. The trace matching as currently presented is incomplete. The generating pair {A, B} from the abstract parameters has trace-squared values (tr²(A), tr²(B), tr²(AB)) = (4, 1, −1), where A is parabolic. The SnapPy generators b and bA both have tr² = 1 (both elliptic of order 3), so they cannot correspond to A and B directly. The element bAB with tr² = −1 matches tr²(AB), but without also matching tr²(A) = 4 and tr²(B) = 1 against the corresponding SnapPy elements, the identification is not fully established by the trace triple alone. We have two options for the revision, and we are committed to carrying out at least one of them. Option (i): find the correct SnapPy elements corresponding to A and B. Since A is parabolic (tr² = 4) and B is elliptic of order 3 (tr² = 1), we need to locate a parabolic element in the SnapPy fundamental group of L6a2(3,0) and an order-3 elliptic whose product has tr² = −1. The SnapPy output already shows that bA has tr² = 1 (order 3 elliptic), so B could correspond to bA or a conjugate. We need to find a parabolic element p in the SnapPy group such that tr²(p · bA) = −1. This is a finite computation that we will perform and report in the revision. Option (ii): alternatively, we can verify the identification by checking that the relator W_{3/10} = bAbaBabAbaBaBAbABaBA is the identity in the abstract group G (defined by parameters (2)) and that no order-2 element exists, which together with the classification [9] forces G to be the orbifold L6a2(3,0). This approach does not require matching all three traces against SnapPy directly, but relies on the classification theorem. We believe option (i) is cleaner and will pursue it as the primary fix, supplementing with option (ii) if needed. If we are, revision: partial

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; the paper's central results are verified against external benchmarks (SnapPy census, trace computations) rather than being defined by their inputs.

full rationale

The paper's two main results (Theorem 3.6 and Theorem 3.8) are verified independently of the author's prior work. Theorem 3.6 identifies a finite covolume group by matching trace parameters against a SnapPy census manifold (L6a2(3,0)), and Theorem 3.8 identifies another group against the census manifold m003(0,0) by comparing three traces and invoking Mostow–Prasad rigidity. The author's prior work [13, 14, 15] provides coordinate systems and Farey word tables used as computational tools, but these are not used to define the target results. The classification of Chesebro–Martin–Schillewaert [9] is used only to generate an initial guess of the orbifold type; the author explicitly states that 'once we have our guess then the verification is independent of the general theory (we just need to compare the trace parameters).' The self-citations are therefore not load-bearing for the final identification. The skeptic's concerns about algebraic errors in the proof of Theorem 3.6 (wrong formulas for AB and ABABA, checking the wrong SnapPy element, incomplete trace matching) are correctness issues, not circularity issues—they concern whether the verification was done correctly, not whether the verification is circular by construction. No step in the derivation chain reduces to its own inputs by definition or by a self-citation chain that is itself unverified.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical entities or particles. The mathematical objects (parabolic locus, maximal cusp groups) are standard. The free parameters are the trace coordinates of the representations, which are determined by solving the defining equations of the parabolic locus.

free parameters (1)
  • Trace parameters (t_X, t_Y, t_XY, v) = See equations (2)-(5) for specific numerical values
    The representations are parameterized by traces of generators and their product. The specific values are found by solving the parabolicity equations tr^2=4 for the three curves.
assumptions (4)
  • standard math Ahlfors' finiteness theorem for torsion groups (used in Lemma 2.2)
    Invoked in Lemma 2.2 to ensure finitely many ideal singular arcs, standard background result.
  • standard math Half Lives, Half Dies theorem (used in Lemma 2.4)
    Invoked in Lemma 2.4 to bound the number of ends, standard 3-manifold topology result.
  • domain assumption Classification of discrete cofinite groups generated by a pair of elliptics or parabolics [9]
    Invoked in Theorem 3.6 proof to identify the group as a Heckoid or 2-bridge link group. This is a domain result cited as 'To appear'.
  • standard math Mostow-Prasad rigidity
    Invoked in Theorem 3.8 to conclude isometry from group isomorphism, standard result.

how reviews work

0 comments
Cite this review

Pith. "Pith review of On rank two Kleinian groups with three parabolics." pith.science (2026). https://pith.science/paper/NPRUTRGS

@misc{pith2026260707424,
  author       = {Pith},
  title        = {Pith review of: On rank two Kleinian groups with three parabolics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NPRUTRGS}},
  note         = {Machine review of arXiv:2607.07424}
}
abstract

The free group of rank $2$ is the fundamental group of the genus $2$ handlebody $\mathcal{H}$. We study discrete representations of this group into $ \mathsf{PSL}(2,\mathbb{C}) $ so that three disjoint simple closed curves on the conformal boundary $\partial_\infty \mathcal{H} $ are sent to parabolic elements. We show that the only infinite covolume groups of this form are maximal cusp groups on the boundary of genus $2$ Schottky space. We also exhibit hitherto unexpected finite covolume groups which do not arise from Heegaard splitting presentations of tunnel number $1$ links.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 35 canonical work pages

  1. [9]

    arXiv:2512.17044 [math.GT](cit. on pp. 2, 4). REFERENCES 13

    work page arXiv

  2. [1]

    Classification of non-free Kleinian groups generated by two parabolic transformations

    Hirotaka Akiyoshi, Ken’ichi Ohshika, John Parker, Makoto Sakuma and Han Yoshida. “Classification of non-free Kleinian groups generated by two parabolic transformations”. In:Trans. Am. Math. Soc.374.3 (2021), pp. 1765–1814.DOI: 10.1090/tran/8246(cit. on p. 1)

    work page doi:10.1090/tran/8246(cit 2021

  3. [2]

    Inequalities for finitely generated Kleinian groups

    Lipman Bers. “Inequalities for finitely generated Kleinian groups”. In:J. Anal. Math.18 (1967), pp. 23–41.DOI:10.1007/BF02798032(cit. on p. 6)

    work page doi:10.1007/bf02798032(cit 1967

  4. [3]

    Bridson and André Haefliger.Metric spaces of non-positive curvature

    Martin R. Bridson and André Haefliger.Metric spaces of non-positive curvature. Grundlehren Math. Wiss. 319. Berlin: Springer, 1999 (cit. on p. 6)

    work page 1999

  5. [4]

    The classification of Kleinian surface groups. II: The ending lamination conjecture

    Jeffrey F. Brock, Richard D. Canary and Yair N. Minsky. “The classification of Kleinian surface groups. II: The ending lamination conjecture”. In:Ann. Math. (2)176.1 (2012), pp. 1–149.DOI: 10.4007/annals.2012.176.1.1 (cit. on p. 1)

    work page doi:10.4007/annals.2012.176.1.1 2012

  6. [5]

    The cusped hyperbolic census is complete

    Benjamin A. Burton. “The cusped hyperbolic census is complete”. In:Trans. Am. Math. Soc.(2017). Online first.DOI:10.1090/tran/6767(cit. on p. 12)

    work page doi:10.1090/tran/6767(cit 2017

  7. [6]

    A census of cusped hyperbolic 3-manifolds

    Patrick J. Callahan, Martin V. Hildebrand and Jeffrey R. Weeks. “A census of cusped hyperbolic 3-manifolds”. In:Math. Comput.68.225 (1999), 321–332, microfiche suppl.URL: http : / / www . jstor . org / stable / 2585113(cit. on p. 12)

    work page 1999

  8. [7]

    Ap- proximation by maximal cusps in boundaries of deformation spaces of Kleinian groups

    Richard D. Canary, Marc Culler, Sa’ar Hersonsky and Peter B. Shalen. “Ap- proximation by maximal cusps in boundaries of deformation spaces of Kleinian groups.” In:J. Differ. Geom.64.1 (2003), pp. 57–109.DOI: 10 . 4310 / jdg / 1090426888(cit. on p. 3)

    work page 2003

Show all 35 references
  1. [8]

    Eric Chesebro.Geometry for Kleinian groups generated by a parabolic pair

  2. [10]

    Martin and Jeroen Schillewaert.Classification and presentations of Heckoid groups

    Eric Chesebro, Gaven J. Martin and Jeroen Schillewaert.Classification and presentations of Heckoid groups. To appear (cit. on pp. 10, 11)

  3. [11]

    The tree of knot tunnels

    Sangbum Cho and Darryl McCullough. “The tree of knot tunnels”. In:Geom. Topol.13.2 (2009), pp. 769–815.DOI:10.2140/gt.2009.13.769(cit. on p. 3)

    2009 doi

  4. [12]

    Dehn filling and the geo- metry of unknotting tunnels

    Daryl Cooper, David Futer and Jessica S. Purcell. “Dehn filling and the geo- metry of unknotting tunnels”. In:Geom. Topol.17.3 (2013), pp. 1815–1876.DOI: 10.2140/gt.2013.17.1815(cit. on p. 3)

    2013 doi

  5. [13]

    Dunfield, Matthias Goerner and Jeffrey R

    Marc Culler, Nathan M. Dunfield, Matthias Goerner and Jeffrey R. Weeks. SNAPPY3.3, a computer program for studying the geometry and topology of 3-manifolds. Available athttp://snappy.computop.org. (Cit. on pp. 10, 12)

  6. [14]

    Alex Elzenaar.From disc patterns in the plane to character varieties of knot groups. 2025. arXiv:2503.13829 [math.GT](cit. on pp. 7, 9)

    2025 arXiv

  7. [15]

    Approximations of the Riley slice

    Alex Elzenaar, Gaven J. Martin and Jeroen Schillewaert. “Approximations of the Riley slice”. In:Expo. Math.41.1 (2023), pp. 20–54.DOI: 10.1016/j. exmath.2022.12.002(cit. on p. 2)

    2023 doi

  8. [16]

    The combinatorics of Farey words and their traces

    Alex Elzenaar, Gaven J. Martin and Jeroen Schillewaert. “The combinatorics of Farey words and their traces”. In:Groups Geom. Dyn.20.1 (2026), pp. 215–256. DOI:10.4171/ggd/832(cit. on p. 10)

    2026 doi

  9. [17]

    cornell.edu/~hatcher/3M/3M.pdf(visited on 26/02/2026) (cit

    Allen Hatcher.Notes on basic3-manifold topology.URL: https://pi.math. cornell.edu/~hatcher/3M/3M.pdf(visited on 26/02/2026) (cit. on p. 6)

    2026

  10. [18]

    A computer generated census of cusped hyperbolic 3-manifolds

    Martin Hildebrand and Jeffrey Weeks. “A computer generated census of cusped hyperbolic 3-manifolds”. In:Computers and mathematics. (Proceedings of a conference, June 13-17, 1989, Massachusetts Institute of Technology, Cambridge, MA). Ed. by Erich Kaltofen and Stephen M. Watt. ...

    1989

  11. [19]

    Geometric finiteness and uniqueness for Kleinian groups with circle packing limit sets

    Linda Keen, Bernard Maskit and Caroline Series. “Geometric finiteness and uniqueness for Kleinian groups with circle packing limit sets”. In:J. Reine Angew. Math.436 (1993), pp. 209–219.URL: https://eudml.org/doc/153503 (cit. on pp. 3, 7)

    1993

  12. [20]

    The Riley slice of Schottky space

    Linda Keen and Caroline Series. “The Riley slice of Schottky space”. In:Proc. Lond. Math. Soc. (3)69.1 (1994), pp. 72–90.DOI: 10.1112/plms/s3-69.1.72 (cit. on p. 1)

    1994 doi

  13. [21]

    On a volume invariant of 3-manifolds

    Marc Kegel, Arunima Ray, Jonathan Spreer, Em Thompson and Stephan Tillmann. “On a volume invariant of 3-manifolds”. In:Expo. Math., 125782 (2026). In press, journal pre-proof.DOI: 10.1016/j.exmath.2026.125782 (cit. on p. 12)

    2026 doi

  14. [22]

    The Riley slice revisited

    Yohei Komori and Caroline Series. “The Riley slice revisited”. In:The Epstein Birthday Schrift. Ed. by Igor Rivin, Colin Rourke and Caroline Series. Geom. Topol. Monogr. 1. Warwick: University of Warwick, Institute of Mathematics, 1998, pp. 303–316 (cit. on p. 1)

    1998

  15. [23]

    Kleinian groups and 3-dimensional topology. A survey

    Albert Marden. “Kleinian groups and 3-dimensional topology. A survey”. In: A crash course on Kleinian groups. Lectures given at a special session at the January 1974 meeting of the American Mathematical Society at San Francisco. Ed. by Lipman Bers and Irwin Kra. Lect. Notes Ma...

    1974 doi

  16. [24]

    Grundlehren Math

    Bernard Maskit.Kleinian groups. Grundlehren Math. Wiss. 287. Berlin etc.: Springer-Verlag, 1988 (cit. on p. 8)

    1988

  17. [25]

    Oxford Math

    Katsuhiko Matsuzaki and Masahiko Taniguchi.Hyperbolic manifolds and Kleinian groups. Oxford Math. Monogr. Oxford: Clarendon Press, 1998 (cit. on pp. 1, 3, 6, 11, 12). 14 REFERENCES

    1998

  18. [26]

    The classification of Kleinian surface groups. I: Models and bounds

    Yair Minsky. “The classification of Kleinian surface groups. I: Models and bounds”. In:Ann. Math. (2)171.1 (2010), pp. 1–107.DOI: 10.4007/annals. 2010.171.1(cit. on p. 1)

    2010 doi

  19. [27]

    Identifying tunnel number one knots

    Kanji Morimoto, Makoto Sakuma and Yoshiyuki Yokota. “Identifying tunnel number one knots”. In:J. Math. Soc. Japan48.4 (1996), pp. 667–688.DOI: 10.2969/jmsj/04840667(cit. on pp. 3, 7)

    1996 doi

  20. [28]

    A quadratic parabolic group

    Robert Riley. “A quadratic parabolic group”. In:Math. Proc. Camb. Philos. Soc. 77 (1975), pp. 281–288.DOI:10.1017/S0305004100051094(cit. on p. 1)

    1975 doi

  21. [29]

    An elliptical path from parabolic representations to hyperbolic structures

    Robert Riley. “An elliptical path from parabolic representations to hyperbolic structures”. In:Topology of low-dimensional manifolds. Proceedings of the second Sussex conference, 1977. Ed. by Roger Fenn. Lect. Notes Math. 722. Cham: Springer, 1979, pp. 99–113.DOI:10.1007/bfb00...

    1977 doi

  22. [30]

    Applications of a computer implementation of Poincaré’s theorem on fundamental polyhedra

    Robert Riley. “Applications of a computer implementation of Poincaré’s theorem on fundamental polyhedra”. In:Math. Comput.40 (1983), pp. 607–632.DOI: 10.2307/2007537(cit. on p. 2)

    1983 doi

  23. [31]

    Parabolic representations of knot groups. I

    Robert Riley. “Parabolic representations of knot groups. I”. In:Proc. Lond. Math. Soc. (3)24 (1972), pp. 217–242.DOI: 10.1112/plms/s3-24.2.217 (cit. on p. 1)

    1972 doi

  24. [32]

    Seven excellent knots

    Robert Riley. “Seven excellent knots”. In:Low-dimensional topology. Proceed- ings of the Conference on Topology in Low Dimension, Bangor, 1979. Vol. 1. Ed. by R. Brown and T. L. Thickstun. Lond. Math. Soc. Lect. Note Ser. 48. Cambridge: Cambridge University Press, 1982, pp. 81...

    1979

  25. [33]

    The topology, geometry and algebra of unknotting tunnels

    Makoto Sakuma. “The topology, geometry and algebra of unknotting tunnels”. In:Chaos Solitons Fractals9.4-5 (1998), pp. 739–748.DOI: 10.1016/S0960- 0779(97)00101-X(cit. on p. 3)

    1998 doi

  26. [34]

    The Maskit embedding of the twice punctured torus

    Caroline Series. “The Maskit embedding of the twice punctured torus”. In: Geom. Topol.14.4 (2010), pp. 1941–1991.DOI: 10.2140/gt.2010.14.1941 (cit. on p. 2)

    2010 doi

  27. [35]

    Three dimensional manifolds, Kleinian groups and hy- perbolic geometry

    William P. Thurston. “Three dimensional manifolds, Kleinian groups and hy- perbolic geometry”. In:Bull. Am. Math. Soc., New Ser.6 (1982), pp. 357–379. DOI:10.1090/S0273-0979-1982-15003-0(cit. on p. 1)

    1982 doi

Pith tools

Reviewed July 9, 2026 · model on record in the stance chip above.