Pith. sign in

REVIEW 3 major objections 5 minor 35 references

Revisiting 3-Sasakian and $G_2$-structures

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The nearly-parallel G2 3-forms on the 7-sphere and Berger's space are shown to share one explicit SO(4)-invariant normal form, and the nearly-half-flat equations on their orbits fix the structure constants.

desk verdict Useful, honest survey with clean coordinate formulas for the squashed G2 structures on S7 and B7; the reverse-engineered uniqueness in Theorem 5.2(i) is conditional on an unproved ansatz, so take that claim with a grain of salt. read the letter →

arxiv 2412.19525 v1 pith:PL5ZYU5D submitted 2024-12-27 math.DG hep-th

Simon Salamon , Ragini Singhal This is my paper · ORCID
classification math.DGhep-th MSC 53C2553C2953C3053C10
keywords 3-Sasakianmanifoldsnearly-parallelG2structurescohomogeneity-oneactions7-sphereBergerspacenearly-half-flatSU(3)Spin(7)holonomyexceptional
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 tries to show that the two classic 7-dimensional model spaces for exceptional holonomy — the squashed 7-sphere and Berger's space SO(5)/SO(3) — carry nearly-parallel G2 (NP2) 3-forms that can be written in one explicit canonical shape under their cohomogeneity-one SO(4) actions. In each case the adapted basis of invariant 1-forms turns the 3-form into $\phi = dt \wedge (X_{12}+X_{34}+X_{56}) + X_{135}-X_{146}-X_{236}-X_{245}$, a normal form dictated by the 2x2 upper-triangular matrices in Corollaries 3.4 and 4.4. The paper then shows that the nearly-half-flat condition on the 6-dimensional SO(4) orbits, a 6-dimensional shadow of the full G2 equation, is strong enough to single out the constants that define the homogeneous structures; on Berger's space the full NP2 equations fix $(\lambda, a, b) = (2/\sqrt{5}, 1/2, 0)$ up to sign. Making these 3-forms explicit matters because they are the local models against which cohomogeneity-one constructions, associative submanifolds, and deformations of G2 geometry are tested.

What carries the argument

The machinery is the cohomogeneity-one SO(4) description of the two 7-manifolds, combined with the nearly-half-flat condition on the 6-dimensional principal orbits. The adapted 1-forms are generated by 2x2 upper-triangular matrices applied to the vertical-horizontal pairs $(f_k, e_k)$ for $S^{7}$ and $(p_k, n_k) = \tfrac12(f_k+e_k,\,-f_k+e_k)$ for $B^{7}$, with trigonometric coefficients $c_k = \cos(t+\vartheta_k)$, $s_k = \sin(t+\vartheta_k)$. The target normal form is $\phi = dt \wedge \xi + \mathrm{Re}\,\Xi$, where $\xi = X_{12}+X_{34}+X_{56}$ is a symplectic form on the orbits and $\Xi$ is a complex volume form; a 3-form of this shape is an NP2 structure precisely when the hypersurface equations (5.45) hold. The key identity used as a reverse-engineering tool is the nearly-half-flat equation $d(\mathrm{Re}\,\Xi) = (\mu/2)\,\xi^2$, which turns the geometric condition into algebraic equations for the constants $(\lambda, a, b)$ of the ansatz.

What would settle it

Compute the nearly-half-flat condition (5.46) and the second hypersurface equation (5.45) for the same 2x2 upper-triangular ansatz but with each entry allowed to be a general Fourier series in $c_k$ and $s_k$, requiring smooth extension over the singular orbits at $t=0$ and $t=\pi/3$. If any solution other than $(\lambda,a,b) = (2/\sqrt{5}, 1/2, 0)$ up to sign survives, Theorem 5.2(i)'s uniqueness claim is false; if none does, the finite ansatz was complete.

Watch

Extended reading notes

Core claim

The central discovery is that the squashed NP2 structure on $S^{7}$ and the homogeneous NP2 structure on Berger's space $B^{7}$ can both be presented, in adapted SO(4)-invariant bases, by the same canonical 3-form $\phi = dt \wedge \xi + \mathrm{Re}\,\Xi$, with $\xi = X_{12}+X_{34}+X_{56}$ and $\Xi = (X_1+iX_2)\wedge(X_3+iX_4)\wedge(X_5+iX_6)$. For $S^{7}$ this is Theorem 3.5, with the basis of Corollary 3.4 built from a 2x2 upper-triangular matrix whose entries involve $c_k = \cos(t+\vartheta_k)$ and $s_k = \sin(t+\vartheta_k)$. For $B^{7}$ the analogous statement is Corollary 4.4, with the basis $(Y_i)$ built from $(p_k, n_k)$ and a matrix of the same type. Section 5 turns the tables: rather than deriving the structures from known metrics, it runs the upper-triangular ansatz in reverse and asks which constants solve the nearly-half-flat equation on the 6-dimensional orbits. For $S^{7}$ the NHF condition alone gives a short list of possibilities (Proposition 5.1); for $B^{7}$, imposing the full NP2 equations leaves exactly $(\lambda, a, b) = (2/\sqrt{5}, 1/2, 0)$ up to sign, the value $b = 0$ being the condition for smooth extension over the singular orbits (Theorem 5.2).

Load-bearing premise

The uniqueness results in Theorem 5.2 hold only for the finite ansatz of upper-triangular 2x2 matrices with the specific trigonometric functions, and the paper does not prove that the true NP2 structures must lie inside that ansatz rather than in a more general Fourier series.

Editorial extensions

If this is right

  • The two homogeneous NP2 structures on S^7 and B^7 are now available in an explicit interval-times-SO(4) chart, so local computations of their G2 invariants can be done directly from the 3-form and the matrices in Corollaries 3.4 and 4.4.
  • Because the same canonical 3-form $\phi = dt \wedge \xi + \mathrm{Re}\,\Xi$ appears on both spaces, the algebraic part of the G2 structure — calibrations, associative submanifold equations, the form of $\ast\phi$ — is identical for the two, with only the underlying basis and structural equations differing.
  • The nearly-half-flat condition on 6-dimensional hypersurfaces is strong enough to pin down the constants of the homogeneous NP2 structures up to a short list, so the NHF system can be used as a practical selection rule in cohomogeneity-one constructions.
  • For B^7, the value $b = 0$ forced by the full NP2 equations is exactly what permits smooth extension over the singular orbits SO(4)/O(2), showing that the homogeneous NP2 structure is the boundary of the ansatz that survives the singularity.
  • The explicit upper-triangular forms connect these NP2 structures to the 3-Sasakian description of S^7 and to the principal-subalgebra description of B^7, making transparent the different roles of the SO(4) factors at the singular orbits.

Reading between the lines

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

  • A natural extension is to allow each slot of the 2x2 matrices in Theorem 5.2 to carry a genuine Fourier series; the authors leave this open, and solving the resulting ODE system would either produce new cohomogeneity-one NP2 structures or prove rigidity of the homogeneous ones.
  • Because the canonical form is shared verbatim by S^7 and B^7, one can carry out a single interval-times-SO(4) computation of, say, the infinitesimal deformation space or the associator equation and then instantiate it on both spaces, which would give a direct PDE check of results obtained by twistor methods.
  • The same upper-triangular ansatz, with $b = 0$ ensuring smooth extension, could be tried on the principal S^3 bundles over Hitchin's orbifolds $O_k$ for $k > 5$; the paper poses the existence of SO(4)-invariant NP2 structures there as an open question.
  • The NHF solutions of Proposition 5.1 not containing the homogeneous S^7 structure are natural candidates for SO(4)-invariant nearly-half-flat SU(3) structures on the orbits; whether any of them can be extended off the hypersurfaces to an NP2 manifold is left open by the paper.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper revisits the geometry of regular 3-Sasakian 7-manifolds and their nearly-parallel G2 (NP2) structures, with a focus on cohomogeneity-one actions of SO(4). It derives explicit orthonormal coframings for the squashed NP2 structure on S7 (Corollary 3.4 and Theorem 3.5) and for the Berger space B7 = SO(5)/SO(3) (Corollary 4.4 and Proposition 4.3), showing in each case that the 3-form takes the canonical form φ = dt∧(X12+X34+X56)+X135−X146−X236−X245. Sections 2–4 develop the quaternionic-form machinery of Bryant–Salamon and Galicki–Salamon, and Section 5 studies the induced nearly-half-flat SU(3) structures on the six-dimensional principal orbits, claiming that the NHF condition, together with the full NP2 equations, determines the constants in the ansatz. The paper's central explicit formulas are well supported, but the uniqueness claims in Section 5 are not fully proven.

Significance. If the explicit formulas in Sections 3 and 4 are correct, the paper provides a useful unified presentation of two known homogeneous NP2 structures in cohomogeneity-one coordinates, clarifying the contrast between S7 and B7 through upper-triangular 2×2 matrices. The derivation of the squashed S7 structure from the Bryant–Salamon and Galicki–Salamon formalisms is a valuable contribution, and the comparison with Ziller's round-metric expression is helpful. The paper does not ship machine-checked proofs or reproducible code; the verifications are hand calculations with some Maple checks that are not included. The main gap is that Theorem 5.2(i), which is the basis for the claimed uniqueness of the B7 constants, is explicitly unproved and rests on an uncontrolled finite ansatz. This tempers the significance of the Section 5 conclusions, although it does not invalidate the canonical descriptions in Sections 3 and 4.

major comments (3)
  1. [Section 5, Theorem 5.2(i)] Theorem 5.2(i) asserts that the full NP2 equations (5.45) and (5.46) hold for the B7 ansatz if and only if (λ, a, b) = (±2/√5, 1/2, 0). The proof of part (i) is explicitly omitted: the text states "We omit the proof of (i)", and the Maple verification is not shipped. Because this statement is the basis for the claimed uniqueness of the constants and for the concluding remark that the NHF condition determines the homogeneous structures, the claimed 'iff' is currently unsubstantiated. The authors should either supply a complete proof or explicitly reformulate the statement as a conditional result within the ansatz.
  2. [Section 5, ansatz before Theorem 5.2] The proof of Theorem 5.2 restricts invariant 3-forms to the finite upper-triangular ansatz (Y_{2k−1}, Y_{2k}) = ((2λ, 2λ a c_k), (b, 2s_k)) (p_k, n_k). The authors themselves acknowledge at the end of the section that "one could insert less trivial Fourier series in each slot". Without a completeness argument for this finite trigonometric family, the 'iff' in Theorem 5.2(i) and the phrase "essentially unique within a 3-parameter family" prove uniqueness only inside the ansatz, not among all SO(4)-invariant structures. This should be stated explicitly, since the paper's stronger claim that the NHF condition determines the homogeneous structures depends on it.
  3. [Section 5, Proposition 5.1] Proposition 5.1's conclusion that (5.46) holds if and only if (a,b) ∈ {(0,−1), (0,0), (±2,1)} is likewise derived within the same upper-triangular ansatz with constant coefficients λ, a, b. The proof is explicit, so this is a scope issue rather than an internal inconsistency. Nevertheless, the phrase "reverse engineer the constants occurring in Corollary 3.4" should be qualified as "within this ansatz" to avoid overgeneralizing the conclusion.
minor comments (5)
  1. [Throughout] There are several typos and OCR artifacts: "conditin" in the proof of Proposition 2.1, "stuctures" in Remark 3.1, "annhilator" in Section 4, and "greaterorequalslant" in the Introduction. These should be corrected.
  2. [Section 5, after Proposition 5.1] The sentence "any positive definite linear combination of g7 and h7 on S7 is associated to an NHF structure" is stated without proof or derivation; please add a brief justification or label it as an observation.
  3. [References] Reference [25] (Kawai) appears in the bibliography but I did not find it cited in the text; please check the citation map.
  4. [Equation (4.40)] The displayed 5×5 matrix has a stray comma in the lower-right entry; please fix the formatting.
  5. [Section 5, Theorem 5.2] The sentence "Both parts were checked by hand and using Maple, but in different orders" is vague; specifying which equations were checked and how would help reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the explicit canonical forms are derived from stated invariant data, and the Section 5 constants are obtained by solving PDE conditions rather than by fitting; the main caveat is an omitted proof for Theorem 5.2(i).

full rationale

The derivation chain is self-contained at its core. Proposition 2.1 solves for λ from the condition dϕ = µ∗ϕ using the displayed expressions for dΘ and dΥ, so λ = 2/√5 is derived, not assumed. Theorems 3.5 and 4.4 are explicit substitutions of the invariant bases (Xi) and (Yi) into that 3-form, with the basis relations proved in Theorem 3.3 and Corollary 4.4 from the Maurer–Cartan and Lie-algebra data. In Section 5, Proposition 5.1 and Theorem 5.2 start from a finite 3-parameter ansatz and impose (5.46) or (5.45)–(5.46); the constants λ, a, b are solved from the resulting algebraic system rather than fitted to the target conclusion, so the logic is not circular. The paper's self-citations, e.g. [12] and [20], are supplemented by self-contained proofs in the text; [33] is contextual. Two caveats should be weighed, but they are completeness or correctness issues rather than circularity. Theorem 5.2(i) is asserted with "We omit the proof of (i), since analysis of (5.45) is somewhat fruitless," so the claimed iff cannot be independently checked from the paper, and the Maple verification is not shipped. Also, the finite Fourier ansatz in Theorem 5.2 is acknowledged to be incomplete: "one could insert less trivial Fourier series in each slot." Thus the uniqueness statement is conditional on that ansatz being exhaustive. None of these gaps makes any equation equivalent to its own input by construction.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The paper is a computational differential-geometry study that builds on established 3-Sasakian and G2 constructions. It introduces no new entities and has no residual free parameters: all constants are determined by the geometry or by the consistency equations. The main load-bearing inputs are prior results on self-dual Einstein manifolds, the squashed NP2 metric, the known 3-form on B7, the cohomogeneity-one group diagram, a uniqueness result for connection forms, and a restrictive ansatz for the invariant 3-forms.

assumptions (6)
  • domain assumption Existence of a self-dual Einstein 4-manifold with positive scalar curvature and the associated 3-Sasakian principal bundle P^- from [28] and [12].
    Invoked from the start; the Spin(7) and G2 forms are built on the total space of the spin bundle Sigma^- over such an M.
  • domain assumption The squashed metric h7 = (4/5) sum (beta_i)^2 + pi* g4 on P^- is Einstein with an NP2 structure whose cone has holonomy Spin(7).
    Used as the known background in Proposition 2.1 to fix lambda = 2/sqrt(5); the proof is modelled on [20].
  • domain assumption The SO(5)-invariant 3-form on B7 is the one computed in [21] and [6], and it satisfies the NP2 condition.
    Proposition 4.3 normalizes and rewrites this known form in the basis (gamma_i); the existence of the homogeneous NP2 structure is taken as established.
  • domain assumption The cohomogeneity-one SO(4) action on B7 has group diagram with principal isotropy Z2^2 and singular orbits SO(4)/O(2), with weights (3,1) and (1,3) from [31] and [23].
    Used in Section 4 to build the bases (p_k, n_k) and the pullback of the Maurer-Cartan form in (4.40).
  • ad hoc to paper The invariant 3-forms on the principal orbits are captured by the finite upper-triangular 2x2 ansatz with trigonometric coefficients.
    This ansatz underlies Proposition 5.1 and Theorem 5.2; the authors admit that 'one could insert less trivial Fourier series in each slot', so completeness of the ansatz is not proven.
  • standard math The connection 1-form phi is uniquely determined by equations (1.4) and (1.5).
    Lemma 3.2 verifies only these equations and concludes phi is given by (3.29), relying on uniqueness from Cartan/Levi-Civita theory.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Revisiting 3-Sasakian and $G_2$-structures." pith.science (2026). https://pith.science/paper/PL5ZYU5D

@misc{pith2026241219525,
  author       = {Pith},
  title        = {Pith review of: Revisiting 3-Sasakian and $G_2$-structures},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PL5ZYU5D}},
  note         = {Machine review of arXiv:2412.19525}
}
abstract

The algebra of exterior differential forms on a regular 3-Sasakian 7-manifold is investigated, with special reference to nearly-parallel $G_2$ 3-forms. This is applied to the study of 3-forms invariant under cohomogeneity-one actions by $SO(4)$ on the 7-sphere and on Berger's space $SO(5)/SO(3)$.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 35 canonical work pages

  1. [35]

    Ziller: On the geometry of cohomogeneity one manifol ds with positive curvature

    W. Ziller: On the geometry of cohomogeneity one manifol ds with positive curvature. In: Riemannian Topology and Geometric Structures on Manifolds, Progress M ath. 271 (2009) 233–262. 2, 8, 11, 17 University of Münster, Einsteinstrasse 62, 48149 Münster, Germany. rsinghal@uni-muenster.de King’s College London, Strand, London, WC2R 2LS, UK. simon.salamon@kcl.ac.uk

    work page 2009

  2. [12]

    Bryant, S.M

    R.L. Bryant, S.M. Salamon: On the construction of compl ete metrics with exceptional holonomy. Duke Math. J. 58 (1989) 829–850. 2, 3, 4, 5

    work page 1989

  3. [21]

    Goette, N

    S. Goette, N. Kitchloo, K. Shankar: Diffeomorphism type of the Berger space SO(5)/SO (3). Amer. J. Math. 126 (2004) 395–416. 13, 17

    work page 2004

  4. [6]

    G. Ball, J. Madnick: Associative Submanifolds of the Ber ger Space. Commun. Analysis Geometry, to appear. 2, 12, 13, 16

    work page

  5. [20]

    Galicki, S

    K. Galicki, S. Salamon: Betti numbers of 3-Sasakian man ifolds. Geom. Dedicata 63 (1996) 45–68. 5, 6, 7

    work page 1996

  6. [32]

    Salamon: Riemannian Geometry and Holonomy Groups, L ongman Scientific and Technical, Har- low, Essex, 1989

    S. Salamon: Riemannian Geometry and Holonomy Groups, L ongman Scientific and Technical, Har- low, Essex, 1989. 2

    work page 1989

  7. [33]

    Singhal: Nearly half-flat SU (3)-structures on S3 × S3

    R. Singhal: Nearly half-flat SU (3)-structures on S3 × S3. Differ. Geom. Appl. 97 (2024) 102187. 18

    work page 2024

  8. [1]

    Abbena, S

    E. Abbena, S. Salamon, A. Gray: Modern Differential Geome try of Curves and Surfaces with Math- ematica, 3rd edition, Chapman & Hall/CRC, 2006. 7

    work page 2006

Show all 35 references
  1. [2]

    Agricola, G

    I. Agricola, G. Dileo, L. Stecker: Curvature properties of 3-(α, δ )-Sasaki manifolds. arXiv:2206.05150. 13

    arXiv

  2. [3]

    Agricola, T

    I. Agricola, T. Friedrich: 3-Sasakian manifolds in dime nsion seven, their spinors and G2 -structures. J. Geom. Physics 60 (2010) 326–332

    2010

  3. [4]

    Atiyah, N.J

    M.F. Atiyah, N.J. Hitchin: The Geometry and Dynamics of M agnetic Monopoles. Porter lectures, Princteon University Press, 1988. 17

    1988

  4. [5]

    G. Ball, J. Madnick: Associative submanifolds of squash ed 3-Sasakian manifolds. arXiv:2208.10622 2

    arXiv

  5. [7]

    Bär: Real Killing spinors and holonomy

    C. Bär: Real Killing spinors and holonomy. Commun. Math. Phys. 154 (1993) 509–521. 1, 7

    1993

  6. [8]

    Bazaikin: Noncompact Riemannian spaces with the h olonomy group Spin(7) and 3-Sasakian manifolds

    Ya.V. Bazaikin: Noncompact Riemannian spaces with the h olonomy group Spin(7) and 3-Sasakian manifolds. Proc. Steklov Inst. Math. 263(2008) 2–12. 8, 13

    2008

  7. [9]

    Berger: Les variétés riemanniennes homogènes normal es simplement connexes à courbure stricte- ment positive

    M. Berger: Les variétés riemanniennes homogènes normal es simplement connexes à courbure stricte- ment positive. Ann. Scuola Norm. Sup. Pisa 15 (1961) 179–246 . 2, 4, 12

    1961

  8. [10]

    Boyer, K

    C. Boyer, K. Galicki: Sasakian Geometry. Oxford Mathem atical Monographs, Oxford University Press, 2008. 7

    2008

  9. [11]

    Bryant: Some remarks on G2 structures

    R.L. Bryant: Some remarks on G2 structures. Proc. Gokova Geometry-Topology Conference 20 05, eds. S. Akbulut, T. Onder, R.J. Stern, International Press ( 2006) 75–109

    2006

  10. [13]

    Dearricott: A 7-manifold with positive curvature

    O. Dearricott: A 7-manifold with positive curvature. D uke Math. J. 158 (2011) 307–346. 17

    2011

  11. [14]

    Dearricott: Positively curved self-dual Einstein m etrics on weighted projective planes

    O. Dearricott: Positively curved self-dual Einstein m etrics on weighted projective planes. Ann. Global Anal. Geom. 27 (2005) 79–86. 1

    2005

  12. [15]

    Fernández, S

    M. Fernández, S. Ivanov, V. Muñoz, L. Ugarte, Nearly hyp o structures and compact nearly Kähler 6-manifolds with conical singularities. J. Lond. Math. Soc . 78 (2008) 580–604. 18

    2008

  13. [16]

    A. Fino, P. Nurowski: Analog of selfduality in dimensio n nine. J. Reine Angew. Math. 699 (2015) 67–110. 9

    2015

  14. [17]

    Foscolo, M

    L. Foscolo, M. Haskins: New G2 holonomy cones and exotic nearly Kähler structures on S6 and S3 × S3. Annals of Math. 185 (2017) 59–130. 19

    2017

  15. [18]

    Friedrich, I

    Th. Friedrich, I. Kath, A. Moroianu, U. Semmelmann: On n early parallel G2 -structures. J. Geom. Physics 23 (1997) 259–286. 8

    1997

  16. [19]

    Galicki, H.B

    K. Galicki, H.B. Lawson Jr.: Quaternionic reduction an d quaternionic orbifolds. Math. Annalen 282 (1988) 1–22. 1

    1988

  17. [22]

    Gray: The structure of nearly Kähler manifolds

    A. Gray: The structure of nearly Kähler manifolds. Math . Annalen 223 (1976) 233–248 4 3-SASAKIAN AND G2 -STRUCTURES 21

    1976

  18. [23]

    Grove, B

    K. Grove, B. Wilking, W. Ziller: Positively curved coho mogeneity one manifolds and 3-Sasakian geometry. J. Differ. Geom. 78 (2008) 33–111. 2, 10, 15, 17

    2008

  19. [24]

    Hitchin: A new family of Einstein metrics

    N.J. Hitchin: A new family of Einstein metrics. In: Mani folds and Geometry (Pisa, 1993). Sympos. Math. XXXVI (1996) 190–222. 1, 17

    1996

  20. [25]

    Kawai: Some associative submanifolds of the squashe d 7-sphere

    K. Kawai: Some associative submanifolds of the squashe d 7-sphere. Quart. J. Math. 66 (2015) 861–

    2015

  21. [26]

    Kennon, J

    A. Kennon, J. Lotay: Geometric flows of G2 -structures on 3-Sasakian 7-manifolds. J. Geom. Physics 187 (2023) 104793

    2023

  22. [27]

    Kobak, A

    P.Z. Kobak, A. Swann: Classical nilpotent orbits as hyp erkähler quotients. Intern. J. Math. 7 (1996) 193–210. 16

    1996

  23. [28]

    Konishi: On manifolds with Sasakian 3-structure ove r quaternion Kaehler manifolds

    M. Konishi: On manifolds with Sasakian 3-structure ove r quaternion Kaehler manifolds. Kodai Math. Sem. Rep. 26(1975) 194–200. 1, 2

    1975

  24. [29]

    Kronheimer: Instantons and the geometry of the nil potent variety

    P.B. Kronheimer: Instantons and the geometry of the nil potent variety. J. Differ. Geom. 32 (1990) 473–490

    1990

  25. [30]

    Lotay: Associative submanifolds of the 7-sphere

    J.D. Lotay: Associative submanifolds of the 7-sphere. Proc. London Math. Soc. 105 (2012) 1183–1214. 2

    2012

  26. [31]

    Podestà, L

    F. Podestà, L. Verdiani: Positively curved 7-dimensio nal manifolds. Quart. J. Math. 50 (1999) 497–

    1999

  27. [34]

    Swann: HyperKähler and quaternionic Kähler geometr y

    A. Swann: HyperKähler and quaternionic Kähler geometr y. Math. Ann. 289 (1991) 421–450. 7

    1991

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.