Einstein Manifolds, Self-Dual Weyl Curvature, and Conformally Kaehler Geometry

Claude LeBrun

REVIEW 1 major objections 4 minor 23 references

Einstein Manifolds, Self-Dual Weyl Curvature, and Conformally Kaehler Geometry

T0 review · 1 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Every simply connected compact oriented Einstein 4-manifold whose self-dual Weyl curvature has positive determinant everywhere is conformally Kähler, hence orientedly diffeomorphic to a del Pezzo surface.

desk verdict Theorem A's integral argument is clean and self-contained; Theorem C's final almost-Kähler-to-Kähler step is outsourced to a cited proposition whose hypotheses aren't checked—a soft spot to tighten, not a fatal flaw. read the letter →

arxiv 1908.01881 v2 pith:FFWCGKPN submitted 2019-08-05 math.DG

Claude LeBrun This is my paper · ORCID

classification math.DG MSC 53C2514J2632J1553C55

keywords Einsteinmetricself-dualWeylcurvatureconformallyKählerdelPezzosurfaceharmonicextremal4-manifold

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 establishes a local curvature test for being conformally Kähler in four dimensions: a simply connected compact oriented Einstein 4-manifold whose self-dual Weyl curvature $W^+$ satisfies $\det(W^+)>0$ at every point is conformal to an extremal Kähler metric, and therefore is orientedly diffeomorphic to a del Pezzo surface. Because the determinant condition is purely local and conformally invariant, it is used here as the first such local curvature criterion for this class of Einstein metrics. The proof also covers the wider class of metrics with harmonic self-dual Weyl curvature, $\delta W^+=0$, showing that with $b_+(M)\neq 0$ they are exactly the metrics $h=s^{-2}g$ obtained from positive-scalar Kähler surfaces by a standard conformal ansatz. A relaxed determinant inequality, $\det(W^+)\ge -\frac{5\sqrt2}{21\sqrt{21}}|W^+|^3$, is shown to force $\det(W^+)>0$ after passing to a double cover, giving the same conformally Kähler conclusion more broadly.

What carries the argument

The central object is the self-dual Weyl endomorphism $W^+:\Lambda^+\to\Lambda^+$. When $\det(W^+)>0$, its eigenvalues $(\alpha,\beta,\gamma)$ sum to zero, so exactly one is positive; the positive eigenline is a smooth real line bundle $L\subset\Lambda^+$. The proof rescales $h$ to $g=\alpha^{2/3}h$, forcing the top eigenvalue of $W^+_g$ to satisfy $\alpha_g f=1$ for $f=\alpha_h^{-1/3}$, and picks a global self-dual 2-form $\omega$ in $L$ with $|\omega|^2_g=2$, which defines an almost-complex structure. The key identity is the weighted Weitzenböck formula for $fW^+$ that follows from $\delta W^+=0$; integrating it against $\omega\otimes\omega$ and using the eigenvalue bounds forces $\int|\nabla\omega|^2\le 0$, hence $\nabla\omega=0$, so the almost-complex structure is integrable and $g$ is Kähler.

What would settle it

A compact simply connected oriented Einstein 4-manifold with $\det(W^+)>0$ at every point whose underlying smooth manifold is not orientedly diffeomorphic to one of the ten del Pezzo surfaces would refute Theorem A and its corollary.

Watch

Extended reading notes

Core claim

The central claim, stated as Theorem A, is that $\det(W^+)>0$ for a simply connected compact oriented Einstein 4-manifold forces the metric to be conformally Kähler, with the conformally rescaled metric extremal Kähler and of positive scalar curvature; the corollary identifies the underlying oriented manifold as a del Pezzo surface and asserts that the known Einstein metrics with this property sweep exactly one connected component of the Einstein moduli space. For the larger class of metrics with harmonic self-dual Weyl curvature, the paper proves the same conclusion without simple connectivity up to a double cover, and classifies the underlying manifolds as rational or ruled surfaces. The method is an explicit construction: positivity of the determinant makes the top eigenspace of $W^+$ a smooth line bundle, and the preferred conformal rescaling $g=\alpha^{2/3}h$ turns the harmonicity of $W^+$ into an integral identity whose only non-negative outcome is $\nabla\omega=0$, so the self-dual 2-form $\omega$ is covariantly constant and defines the Kähler form. Finally, the paper shows that a much weaker inequality involving $|W^+|$ already forces $\det(W^+)>0$, so the same conformally Kähler conclusion follows under weaker hypotheses.

Load-bearing premise

At the end of the proof of Theorem 3.1, the argument assumes that a previously published proposition, cited but not stated, converts an almost-Kähler metric into a Kähler metric under hypotheses that may hold only for the original metric $h$ rather than the rescaled metric $g$; if that proposition requires $\delta W^+=0$ to hold for $g$ itself, the proof of Theorem C has a gap.

Editorial extensions

If this is right

Every simply connected compact oriented Einstein 4-manifold with $\det(W^+)>0$ is orientedly diffeomorphic to one of the ten del Pezzo surfaces: $S^2\times S^2$ or the nine manifolds $\mathbb{CP}^2\#m\overline{\mathbb{CP}}^2$, $0\le m\le 8$.
On each such manifold, the known Einstein metrics with $\det(W^+)>0$ fill exactly one connected component of the Einstein moduli space $\mathcal{E}(M)$.
More generally, any compact oriented 4-manifold with harmonic self-dual Weyl curvature, $b_+(M)\neq 0$, and $\det(W^+)>0$ is conformal to a positive-scalar Kähler metric with $h=s^{-2}g$, and hence is orientedly diffeomorphic to a rational or ruled surface: $\mathbb{CP}^2$, $(\Sigma\times S^2)\#k\,\mathbb{CP}^2$, or a nontrivial $S^2$-bundle over $\Sigma$.
Without simple connectivity, the only possible fundamental groups are trivial or $\mathbb{Z}_2$, so any compact oriented Einstein example reduces to the simply connected case by passing to a double cover.
The relaxed condition $\det(W^+)\ge -\frac{5\sqrt2}{21\sqrt{21}}|W^+|^3$ with $\delta W^+=0$ already forces $\det(W^+)>0$, so it leads to the same conformally Kähler conclusion.

Reading between the lines

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

The conformal normalization $\alpha_g f=1$ effectively promotes a pointwise eigenvalue condition to a global geometric structure; a natural test is whether this construction yields a canonical conformal representative on non-compact or orbifold 4-manifolds with the same determinant sign.
The constant $-\frac{5\sqrt2}{21\sqrt{21}}$ in Theorem C is the exact value at the eigenvalue ratio $\beta/\alpha=1/4$, so it may be the sharp threshold separating conformally Kähler from non-Kähler behavior; looking for equality cases could reveal borderline almost-Kähler metrics that are not Kähler.
Because Theorem C's last step invokes a cited proposition without restating it, a direct proof that the almost-Kähler metric $g$ with $W^+_g(\omega,\omega)>0$ and the conformally weighted harmonicity condition is Kähler would make the theorem independent of that external result.
The classification in Theorem B suggests that $\det(W^+)>0$ exactly characterizes the conformal classes of positive-scalar Kähler metrics on rational and ruled surfaces; checking whether the moduli of such metrics is naturally parameterized by conformal classes of extremal Kähler metrics would connect the Einstein and extremal-Kähler moduli problems.

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.

Desk Editor's Note

read the letter

Good paper, and I'd send it to a referee. The centerpiece is a new proof of Wu's theorem (det W^+ > 0 forces conformal Kählerity for simply-connected Einstein 4-manifolds), and the proof here is better than the announced one in one respect: it's a direct integral argument. Theorem 2.1 is self-contained: choose f so α f ≡ 1, use Lemma 1's inequality and the Weitzenböck formula, and the integral identity forces ∇ω ≡ 0. That's clean, and I verified the inequalities. The algebraic Lemma 3 for Theorem C also checks out; the constant is right.

The genuinely new results are Theorem B (harmonic self-dual Weyl curvature, b+ ≥ 1, det W^+ > 0 implies conformally Kähler) and Theorem C (the sharp lower bound on det W^+ that implies positivity). Those are real extensions, not just rehashes, and the moduli-space applications in the corollaries are reasonable.

Now the soft spot, which is exactly where the stress-test note points: Theorem C's proof stops at dω = 0 and then says 'by [16, Proposition 2]' the almost-Kähler metric is actually Kähler. The proposition is not stated, and the paper doesn't explicitly verify its hypotheses for the conformally rescaled metric g. The text says 'h = f² g satisfies δ(W^+) = 0' — that indicates the proposition is being invoked with h as the metric with harmonic self-dual Weyl curvature, not g. I think that's probably the correct reading, and if so the inference is sound. But it's not checkable from this paper alone. A referee should ask the author to state the proposition and show that the hypotheses are met. This is a gap in exposition, not a revealed flaw in the mathematics; it doesn't affect Theorem A, whose proof is independent of this step.

The citation pattern is fine. The paper leans on the author's own earlier work [13] and [16], but those are published with independent proofs, and the present article uses them for clearly identified classification and upgrade steps. No circularity. No data, no fitting; the curvature conditions are pointwise and the conclusions are derived, not fitted.

Final word: this is a solid contribution to 4-dimensional conformal and Einstein geometry. The intended audience is specialists, but the main proof is accessible to anyone comfortable with Weitzenböck formulas. It deserves a serious referee. My recommendation: send it to peer review, and have the referee require the author to clarify the cited proposition in Section 3 before it's accepted.

Referee Report

1 major / 4 minor

Summary. This paper proves a characterization of conformally Kähler Einstein metrics on compact oriented 4-manifolds via the sign of det(W^+), the determinant of the self-dual Weyl curvature. Theorem A states that a simply-connected compact oriented Einstein 4-manifold with det(W^+) > 0 is conformal to an orientation-compatible extremal Kähler metric. The paper gives an independent proof of Wu's announced result by constructing a preferred conformal rescaling g = α^{2/3} h from the top eigenvalue α of W^+, selecting a global self-dual eigenform ω, and deriving via a Weitzenböck formula and an integral inequality that ∇ω = 0. Theorem B extends the conclusion to harmonic self-dual Weyl curvature with b_+(M) ≠ 0, and Theorem C replaces the condition det(W^+) > 0 by the weaker inequality det(W^+) ≥ -5√2/(21√21) |W^+|^3, yielding det(W^+) > 0 and conformal Kählerity after passing to a double cover. The paper also includes classification corollaries for del Pezzo surfaces and the connectedness of the relevant Einstein moduli space component.

Significance. If the results are correct, this is a significant advance: it converts a non-local, harmonic-form characterization of conformally Kähler Einstein metrics into a purely local curvature inequality, and it provides a new proof of Theorem A that is largely self-contained. The main integral argument in Theorem 2.1 is elegant and the algebraic estimates have correct constants. The paper is also valuable for extending the method to harmonic self-dual Weyl metrics and for giving a clean treatment of the double-cover cases. The reliance on the author's earlier classification results [13] and [16] is acceptable since those are published with independent proofs, though the final almost-Kähler-to-Kähler step in Theorem C needs explicit hypothesis checking. Overall the central claim is defensible and the method is novel, with no free parameters and a clear geometric conclusion.

major comments (1)

[§3, proof of Theorem 3.1] The final step of the proof of Theorem 3.1 invokes [16, Proposition 2] to conclude that an almost-Kähler metric g (with dω = 0, |ω| = √2, and W^+(ω,ω) > 0) is actually Kähler, but the proposition is neither stated nor proved, and its hypotheses are not checked for the conformally rescaled metric g. The paper establishes δ_h W_h^+ = 0 for the original metric h = f^2 g, and the weighted conformal invariance gives δ_g(f W_g^+) = 0, which is the premise used in the Weitzenböck formula (9). It is not verified whether [16, Proposition 2] requires the harmonicity condition δ_g W_g^+ = 0 on the almost-Kähler metric g itself or merely the weighted condition on the conformal class. Since Theorem C and Propositions 3.2 and 3.3 depend on this step, the authors should either state [16, Proposition 2] and confirm that the conformal class (or the specific representative g) satisfies its hypotheses, or supply a direct proof of the almost-Kähler-to-Kähler upgrade.

minor comments (4)

[§2, proof of Theorem A] Theorem A asserts that h is conformal to an extremal Kähler metric, but the proof of Theorem 2.1 and the surrounding discussion only explicitly show that h = s^{-2} g for a Kähler metric g of positive scalar curvature. Please add a sentence explaining why g is extremal, for example by citing the relevant result from [13] or [7].
[§3, Theorem 3.1 computation] In the chain of integral identities after Eq. (9), the equality ∫⟨ω,(d+d*)^2ω⟩ = 2∫|dω|^2 uses the fact that |d*ω| = |dω| for a self-dual 2-form ω, which follows from *ω = ω. This step is not stated; a short parenthetical justification would improve readability.
[General] The proof of Proposition 2.3 depends on the author's earlier work [13], and the proof of Theorem C depends on [16, Proposition 2]. These are published results, but since they are used in load-bearing positions, it would be helpful to state exactly which statements are being imported, particularly for [16, Proposition 2].
[General] There are a few typographical and formatting issues, such as 'Propostion' in Section 2 and inconsistent capitalization of 'del Pezzo'; a careful proofread is recommended.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the conformal-rescaling argument is self-contained; the author's self-citations are load-bearing but are published independent theorems, and the flagged [16] issue is a rigor gap, not a circular step.

full rationale

The central derivation is self-contained and does not reduce to its own inputs. In Section 2, the metric h with δW^+ = 0 and det(W^+) > 0 is rescaled by f = α_h^{-1/3}; the integral identity derived from the weighted Weitzenböck formula (9), together with Lemmas 1 and 2, yields ∇ω = 0 and hence a Kähler metric g without assuming the conclusion. The del Pezzo corollary and the final almost-Kähler-to-Kähler upgrade in Theorem 3.1 are imported from the author's earlier papers [13] and [16, Prop. 2]; these are published theorems with independent proofs and are not restatements of the present target, so they are self-citations rather than circularity. The reader-identified concern that [16, Prop. 2] may require harmonicity of W^+ with respect to g rather than the weighted condition δ_g(fW_g^+) = 0 is a possible rigor gap in the last step of Theorem C, not a definitional or fitted-input circle. No fitted parameter is renamed as a prediction, and no equation is reused as its own conclusion.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

The proofs introduce no free parameters or invented entities. The central calculation relies on standard Weitzenbock identities and on several prior theorems, including two by the author, that are cited but not reproved.

assumptions (7)

standard math Weitzenbock formula (7): (d+d*)^2 ω = ∇*∇ω - 2W+(ω) + (s/3)ω for self-dual 2-forms.
Used in the integration-by-parts argument in Theorem 2.1 and Theorem 3.1.
domain assumption Weighted conformal invariance: if δW+ = 0 for h, then δ(fW+) = 0 for g = f^{-2}h, along with Weitzenbock formula (9).
Standard conformal geometry fact from [18], used to set up the integral identity.
domain assumption Derdzinski's theorem: a Kähler surface of scalar curvature s > 0 yields h = s^{-2}g with δW+ = 0 and det(W+) > 0.
Provides the converse direction of Theorem B and the known examples for the corollary.
domain assumption Yau's theorem: a compact Kähler surface with positive scalar curvature has h^{2,0} = 0 and hence b+(M) = 1.
Used to identify b+ and harmonic self-dual 2-forms in Propositions 2.2 and 2.3.
domain assumption [16, Proposition 2]: an almost-Kähler 4-manifold with W+(ω,ω) > 0 and harmonic self-dual Weyl curvature is Kähler with positive scalar curvature.
Load-bearing for Theorem C; the paper cites it without proof.
domain assumption LeBrun's classification [13]: Einstein metrics on compact complex surfaces; used to conclude a conformally Kähler Einstein metric with b+ = 1 is a del Pezzo surface.
Load-bearing for Proposition 2.3 and the corollary.
standard math b+(M) = 1 + 2h^{2,0} for compact Kähler surfaces.
Used to relate geometric genus and b+.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Einstein Manifolds, Self-Dual Weyl Curvature, and Conformally Kaehler Geometry." pith.science (2026). https://pith.science/paper/FFWCGKPN

@misc{pith2026190801881,
  author       = {Pith},
  title        = {Pith review of: Einstein Manifolds, Self-Dual Weyl Curvature, and Conformally Kaehler Geometry},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FFWCGKPN}},
  note         = {Machine review of arXiv:1908.01881}
}

read the original abstract

Peng Wu recently announced a beautiful characterization of conformally Kaehler, Einstein metrics of positive scalar curvature on compact oriented 4-manifolds via the condition det (W^+) > 0. In this note, we buttress his claim by providing an entirely different proof of his result. We then present further consequences of our method, which builds on techniques previously developed in (LeBrun 2015).

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 23 canonical work pages

[13]

184 of Lecture Notes in Pure and Appl

, Einstein metrics on complex surfaces , in Geometry and Physics (Aarhus, 1995), vol. 184 of Lecture Notes in Pure and Appl. Ma th., Dekker, New York, 1997, pp. 167–176

work page 1995
[16]

Global Anal

, Einstein metrics, harmonic forms, and symplectic four- manifolds, Ann. Global Anal. Geom., 48 (2015), pp. 75–85

work page 2015
[1]

Barth, C

W. Barth, C. Peters, and A. V an de Ven , Compact Complex Surfaces, vol. 4 of Ergebnisse der Mathematik und ihrer Grenzgebiete (3), Springer-Verlag, Berlin, 1984

work page 1984
[2]

M. Berger, Sur les vari´ et´ es d’Einstein compactes, in Comptes Rendus de la IIIe R´ eunion du Groupement des Math´ ematiciens d’Exp ression Latine (Namur, 1965), Librairie Universitaire, Louvain, 1 966, pp. 35– 55

work page 1965
[3]

A. L. Besse , Einstein Manifolds, vol. 10 of Ergebnisse der Mathematik und ihrer Grenzgebiete (3), Springer-Verlag, Berlin, 1987

work page 1987
[4]

Besson, G

G. Besson, G. Courtois, and S. Gallot , Entropies et rigidit´ es des espaces localement sym´ etriques de courbure stricteme nt n´ egative, Geom. and Func. An., 5 (1995), pp. 731–799

work page 1995
[5]

B ¨ohm, Inhomogeneous Einstein metrics on low-dimensional spheres and other low-dimensional spaces , Invent

C. B ¨ohm, Inhomogeneous Einstein metrics on low-dimensional spheres and other low-dimensional spaces , Invent. Math., 134 (1998), pp. 145– 176

work page 1998
[6]

C. P. Boyer, K. Galicki, and J. Koll ´ar, Einstein metrics on spheres, Ann. of Math. (2), 162 (2005), pp. 557–580

work page 2005

Show all 23 references

[7]

X. X. Chen, C. LeBrun, and B. Weber , On conformally K¨ ahler, Einstein manifolds , J. Amer. Math. Soc., 21 (2008), pp. 1137–1168

2008
[8]

Derdzi ´nski, Self-dual K¨ ahler manifolds and Einstein manifolds of dimension four , Compositio Math., 49 (1983), pp

A. Derdzi ´nski, Self-dual K¨ ahler manifolds and Einstein manifolds of dimension four , Compositio Math., 49 (1983), pp. 405–433

1983
[9]

M. J. Gursky , Four-manifolds with δW + = 0 and Einstein constants of the sphere , Math. Ann., 318 (2000), pp. 417–431

2000
[10]

N. J. Hitchin , Compact four-dimensional Einstein manifolds , J. Dif- ferential Geometry, 9 (1974), pp. 435–441

1974
[11]

, On the curvature of rational surfaces , in Diﬀerential geometry (Proc. Sympos. Pure Math., Vol. XXVII, Part 2, Stanford Univ ., Stan- ford, Calif., 1973), Amer. Math. Soc., Providence, R. I., 19 75, pp. 65–80

1973
[12]

LeBrun , Einstein metrics and Mostow rigidity , Math

C. LeBrun , Einstein metrics and Mostow rigidity , Math. Res. Lett., 2 (1995), pp. 1–8. 16

1995
[14]

Diﬀerential Geom., 90 (2012), pp

, On Einstein, Hermitian 4-manifolds , J. Diﬀerential Geom., 90 (2012), pp. 277–302

2012
[15]

Reine Angew

, Einstein manifolds and extremal K¨ ahler metrics, J. Reine Angew. Math., 678 (2013), pp. 69–94

2013
[17]

Odaka, C

Y. Odaka, C. Spotti, and S. Sun , Compact moduli spaces of del Pezzo surfaces and K¨ ahler-Einstein metrics, J. Diﬀerential Geom., 102 (2016), pp. 127–172

2016
[18]

Penrose and W

R. Penrose and W. Rindler , Spinors and space-time. Vol. 2 , Cam- bridge University Press, Cambridge, 1986. Spinor and twist or methods in space-time geometry

1986
[19]

Sung , K¨ ahler surfaces of positive scalar curvature, Ann

M.-H. Sung , K¨ ahler surfaces of positive scalar curvature, Ann. Global Anal. Geom., 15 (1997), pp. 509–518

1997
[20]

Tian , On Calabi’s conjecture for complex surfaces with positive ﬁr st Chern class , Invent

G. Tian , On Calabi’s conjecture for complex surfaces with positive ﬁr st Chern class , Invent. Math., 101 (1990), pp. 101–172

1990
[21]

M. Y. W ang and W. Ziller , Einstein metrics on principal torus bundles, J. Diﬀerential Geom., 31 (1990), pp. 215–248

1990
[22]

Wu , Einstein four-manifolds of nonnegative determinant half W eyl curvature

P. Wu , Einstein four-manifolds of nonnegative determinant half W eyl curvature. e-print arXiv:1903.11818 [math.DG]

1903 arXiv
[23]

S. T. Yau , On the curvature of compact Hermitian manifolds , Invent. Math., 25 (1974), pp. 213–239. Department of Mathematics, Stony Brook University (SUNY), Stony Brook, NY 11794-3651 USA e-mail: claude@math.sunyb.edu Keywords: Einstein metric, Del Pezzo surface, Weyl curvat...

1974

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [13]

184 of Lecture Notes in Pure and Appl

, Einstein metrics on complex surfaces , in Geometry and Physics (Aarhus, 1995), vol. 184 of Lecture Notes in Pure and Appl. Ma th., Dekker, New York, 1997, pp. 167–176

work page 1995

[2] [16]

Global Anal

, Einstein metrics, harmonic forms, and symplectic four- manifolds, Ann. Global Anal. Geom., 48 (2015), pp. 75–85

work page 2015

[3] [1]

Barth, C

W. Barth, C. Peters, and A. V an de Ven , Compact Complex Surfaces, vol. 4 of Ergebnisse der Mathematik und ihrer Grenzgebiete (3), Springer-Verlag, Berlin, 1984

work page 1984

[4] [2]

M. Berger, Sur les vari´ et´ es d’Einstein compactes, in Comptes Rendus de la IIIe R´ eunion du Groupement des Math´ ematiciens d’Exp ression Latine (Namur, 1965), Librairie Universitaire, Louvain, 1 966, pp. 35– 55

work page 1965

[5] [3]

A. L. Besse , Einstein Manifolds, vol. 10 of Ergebnisse der Mathematik und ihrer Grenzgebiete (3), Springer-Verlag, Berlin, 1987

work page 1987

[6] [4]

Besson, G

G. Besson, G. Courtois, and S. Gallot , Entropies et rigidit´ es des espaces localement sym´ etriques de courbure stricteme nt n´ egative, Geom. and Func. An., 5 (1995), pp. 731–799

work page 1995

[7] [5]

B ¨ohm, Inhomogeneous Einstein metrics on low-dimensional spheres and other low-dimensional spaces , Invent

C. B ¨ohm, Inhomogeneous Einstein metrics on low-dimensional spheres and other low-dimensional spaces , Invent. Math., 134 (1998), pp. 145– 176

work page 1998

[8] [6]

C. P. Boyer, K. Galicki, and J. Koll ´ar, Einstein metrics on spheres, Ann. of Math. (2), 162 (2005), pp. 557–580

work page 2005