From 12 to 6: Sharpening the Three-Charge Bound in Maxwell's Problem

A.Gabrielov; B.Shapiro; Dm.Novikov; T.Novikov

REVIEW 2 major objections 4 minor 9 references

For three positive point charges, the number of nondegenerate equilibrium points is at most six, for every charge strength, every exponent α>0, and every dimension n≥2—a sharpening of the previous upper bound of twelve.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-03 00:22 UTC pith:YDE5DRN7

load-bearing objection A real step forward on a classical problem: the new four-contact lemma gives 6 instead of 12, and the only genuine soft spot is that the genericity check at the witness is asserted rather than shown. the 2 major comments →

arxiv 2607.28785 v1 pith:YDE5DRN7 submitted 2026-07-30 math.CA

From 12 to 6: Sharpening the Three-Charge Bound in Maxwell's Problem

A.Gabrielov , Dm.Novikov , T.Novikov , B.Shapiro This is my paper

classification math.CA MSC 31B0534C0858K05

keywords point chargesequilibrium pointscritical pointspotential theorythree chargesfewnomialsupper boundsnondegenerate

verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper establishes that a potential generated by three positive point charges has at most six nondegenerate equilibrium points, for any charge magnitudes, any exponent α>0, and any dimension n≥2. The previous bound was twelve. The improvement comes from proving a missing auxiliary statement: an associated polynomial system Q=R=0 has at least four solutions, counted with multiplicity, in each open quadrant of the two-dimensional (f,g)-plane. The proof uses a separation lemma at the unique saddle point of a separated-variable first integral Φ=u(f)+s(g), followed by a four-contact lemma showing that the oval component of the contact curve Γ={Q=0} must meet the separatrix level at least four times. As a consequence, the number of nondegenerate equilibria for three charges is always 2, 4, or 6.

Core claim

The paper proves that a potential of the form Vα = Σ ζ_i ρ_i^{-α}, with three positive charges in R^n and α>0, has at most six nondegenerate critical points. The argument reduces critical points to intersections of two auxiliary curves in the positive quadrant of an (f,g)-plane, then sharpens an auxiliary count from twelve to six. The new ingredient is a proof that the polynomial system Q=R=0 has at least four solutions, counted with multiplicity, in every open quadrant. This is obtained by a separation lemma at the unique saddle point of the separated-variable first integral Φ=u(f)+s(g), followed by a four-contact lemma: a nonconstant function on the oval component of Γ={Q=0} must have tota

What carries the argument

The load-bearing mechanism is the four-contact lemma applied to the compact oval O⊂Γ, where Γ={Q=0} is the contact curve of two closed logarithmic one-forms defined from the three-charge potential. On O, the auxiliary polynomial R is proportional to the derivative of the first integral Φ restricted to O. The separated-variable form Φ=u(f)+s(g) gives Φ a unique nondegenerate saddle p in each open quadrant, and the separation lemma shows that the level set {Φ=c0} splits that quadrant into exactly two connected components separated by the line {f=f0}. A nonconstant function on a circle with at least two distinct critical points and an even number of odd-order critical points must have total cri

Load-bearing premise

The genericity proof uses exact rational arithmetic at one parameter value to certify that two auxiliary polynomials are coprime and that their resultants have no unexpected common factors; if that single computation, or its extension to a Zariski-open neighborhood, is mistaken, the four-contact lemma cannot be applied.

What would settle it

Run the exact-arithmetic computation at the witness parameter θ* = (3/10, 7/5, 7/10, 14/5, 1/2): if gcd(Q,R)≠1 or gcd(r2,r4) gains an extra real factor beyond f ξ3(f), the genericity step collapses. Independently, a numerical search for a three-charge configuration with eight or more nondegenerate equilibrium points would directly refute the bound of six.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

For every α>0 and every dimension n≥2, any configuration of three positive point charges has 2, 4, or 6 nondegenerate equilibrium points, never more.
The improvement is achieved by saturating the mixed-volume count: the three quadrants outside the positive one now contribute at least 16 of the 28 torus solutions counted by the Bernstein–Kushnirenko bound.
To reach the conjectured bound of 4 for three charges, one only needs to show that when the auxiliary curve γ2 crosses the oval in four points, the first Rolle step overcounts by at least 2; equivalently, that there is at most one local minimum.
The theorem covers all dimensions n≥2 because every equilibrium point lies in the affine span of the three charges, so the problem reduces to a plane.
All tested parameter sets show exactly four real solutions of Q=R=0 in each open quadrant, indicating that the new auxiliary count is generically sharp.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The separation mechanism—a separated one-well/one-hill first integral whose saddle level set splits a quadrant into exactly two components—is a general geometric fact that could yield four-contact lower bounds in other two-dimensional counting problems governed by closed one-forms.
If a configuration with six equilibria is eventually found, it would saturate the new bound and confirm that the method is sharp; the paper leaves open whether such a configuration exists for special values of α, so a targeted search near parameters with four positive-quadrant crossings is a natural next test.
The proof's reliance on exact rational arithmetic at one witness parameter is a testable point: an independent symbolic verification of the two resultant computations would make the genericity argument fully self-contained and could be automated for neighboring parameters.
The sharpness analysis suggests that the slack in the current method is concentrated in the first Rolle step, so future upper bounds for configurations of four or more charges should target that inequality rather than the mixed-volume count.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

A real step forward on a classical problem: the new four-contact lemma gives 6 instead of 12, and the only genuine soft spot is that the genericity check at the witness is asserted rather than shown.

read the letter

You should know this paper delivers what it promises. The bound on nondegenerate equilibria for three positive charges drops from 12 to 6, and the main new tool—the separation lemma at the saddle of a separated-variable first integral, leading to the four-contact lemma—is clever and carefully argued. The proof is not a black box; the geometry is clear and the key lemmas are rigorous.

What is new: Proposition 2.1, the four-solution proposition. The reduction to the auxiliary system Q=R=0 is inherited from the authors' earlier paper [4], and they say so explicitly. That is legitimate reliance, not a gap. The four-contact lemma (Lemma 4.2) is the real contribution: it rules out low-multiplicity configurations by parity and by the separation property of the saddle. I found the argument convincing.

The soft spot is Section 5. The genericity assumptions (G1) and (G2) are verified at a single rational witness θ* via exact arithmetic, and the paper says the computation is short and a script is available. But the actual output of the gcd computations in (5.5) is not reproduced. That is a minor weakness—exactly the kind of thing a referee can check in ten minutes—and it is not a reason to reject. The stress-test note worries that if the gcd claims were wrong the whole bound collapses; that is true, but there is no indication they are wrong, and the authors point to a reproducible script. Still, for a paper whose main theorem depends on it, embedding the output would have been better.

The perturbation argument at the end is standard and correct: the bound persists to all parameters by the implicit function theorem. The paper is honest about the remaining gap to Maxwell's conjectured bound of 4.

Who is this for? Anyone working on Maxwell's problem, fewnomials, or Rolle–Khovanskii theory. It deserves a serious referee. The main thing I'd ask the referee to do is re-run or verify the exact computation at θ* and check the subresultant degree-stability argument. If that checks out, this is a solid contribution.

Referee Report

2 major / 4 minor

Summary. The paper proves Theorem 1.1: for every n≥2 and every α>0, the potential of three positive point charges in R^n has at most six nondegenerate equilibrium points. The proof sharpens the authors' earlier bound of twelve by proving a four-contact lemma: on the oval component O of the contact curve Γ in a quadrant, the auxiliary polynomial R has at least four zeros counted with multiplicity. The key new ingredients are a separation lemma for the level sets of a separated primitive Φ at its unique saddle point, a parity/counting argument on a circle, and a genericity result (Proposition 5.1) verified by exact arithmetic at a single rational witness θ*. The paper states that a reproducible script verify_genericity.py accompanies the source.

Significance. If the result holds, this is a substantial improvement on the three-charge case of Maxwell's problem, reducing the best known upper bound from 12 to 6 and nearly reaching Maxwell's conjectural bound of 4. The four-contact lemma and the separation lemma are elegant and appear new; they are proven rigorously and in detail. The genericity verification is honest: the authors rely on exact rational arithmetic at a witness and provide a reproducibilty statement rather than numerical sampling. The paper also gives computational checks saturating the mixed-volume count, which clarifies the remaining slack in the method and strengthens confidence in the sharpness of the approach. The central derivation is sound, but the proof as written depends on an externally supplied computation that is not embedded in the manuscript.

major comments (2)

[Section 5, Eq. (5.5)] The proof of Proposition 5.1, and hence the entire Theorem 1.1, rests on the exact arithmetic assertions gcd(Q,R)=1 and gcd(r2,r4)=fξ3(f) at the witness θ*. These assertions are not demonstrated in the text; the script is only referenced. As written, the proof of (G1) and (G2) is conditional on an externally supplied computation. Please include the computation output or provide an appendix with the relevant resultant/subresultant values so a reader can verify (5.5) without running the script. This is load-bearing: if either gcd assertion fails, Proposition 2.1 cannot be applied to the three non-positive quadrants and the bound N≤6 collapses.
[Section 5, Eq. (5.6)] The step 'the subresultant criterion, together with (5.4), then shows...' needs more detail. To conclude that gcd(r2,r4)=fξ3(f) on a Zariski-open neighborhood of θ*, one must specify which subresultant is nonzero at θ* (the one forcing gcd degree 3) and which higher subresultants vanish. Without this information, the passage from the single-witness computation to the open-set statement is a gap, albeit a fillable one. The authors should present the subresultant chain or a direct certificate for (5.5).

minor comments (4)

[Section 5, paragraph after (5.5)] The referenced script verify_genericity.py should be included as an ancillary file or appendix with version/checksum information, so that the exact computation is reproducible from the published version.
[Equation (4.1)] Add parentheses for clarity: Q = -((1+2α)/(α+1)^2) ξ1ξ2ξ3 - fgQ1, to avoid misreading the sign.
[Figure 1 caption] The caption text 'equilibria = 1 2; 2 crosses O+ , budget 2 + 2 = 4' is cryptic; please expand it so the reader can interpret the figure's relation to the proof.
[Section 6] The statement 'Q=R=0 has exactly four real solutions in each open quadrant' is a computational check for ten parameter sets; the paper correctly does not use it as evidence for Theorem 1.1. It may help to explicitly say this check is not part of the proof.

Circularity Check

0 steps flagged

No significant circularity: the four-solution proposition is proved, not assumed; reliance on prior work is legitimate.

full rationale

The paper's central new claim, Proposition 2.1, is established in Sections 3–4 by an independent argument: Lemma 3.1 constructs a unique saddle point of the separated primitive and Lemma 4.2 proves at least four contacts via a topological separation argument. This is not assumed or fitted. The author's earlier paper [4] is used for the reduction to the polynomial system Q=R=0, for the identity (4.1), for the multiplicity-six analysis, and for the auxiliary counts N1=2 and N3=0; this is legitimate reliance on a published external result, not a self-referential definition or a fitted parameter renamed as a prediction. Section 5's genericity proof depends on exact rational arithmetic at a single witness and on the cited script verify_genericity.py; even if one worries about the computation not being embedded, that is a reproducibility or verification concern, not circularity, because the witness computation is a concrete check and the Zariski-open conclusion is a standard transfer argument. The final perturbation argument is a standard persistence argument for nondegenerate equilibria, not a tautology. No equation in the paper reduces by construction to its own input, and no load-bearing step is justified solely by a self-citation that is itself unverified. Therefore the derivation chain is not circular.

Axiom & Free-Parameter Ledger

0 free parameters · 4 axioms · 0 invented entities

The proof introduces no free parameters or new postulated entities. It relies on standard results in fewnomial theory and on the authors' prior paper [4] for the central reduction; the new contribution is the four-solution proposition, proven here.

axioms (4)

standard math Rolle–Khovanskii theorem (Khovanskii's fewnomial theory)
Invoked to obtain inequalities (2.9) and (2.13); foundational result in the paper's method.
standard math Bernstein–Kushnirenko theorem
Used for the mixed-volume count (2.16) bounding common zeros of Q and R in (C*)^2.
domain assumption Morse-theoretic parity relation #{equilibria}=2m0+2 (from [4])
Equation (1.2) is stated as recalled from [4]; used only to infer m0≤2 from the bound, not in the main counting.
domain assumption Reduction formulas (2.1)–(2.15) and identity (4.1) from prior work [4]
The entire reduction to the algebraic count N, the definitions of Q and R, and the boundedness of Γ come from the authors' earlier paper. This is legitimate prior published work but is self-cited.

pith-pipeline@v1.3.0-alltime-deepseek · 7692 in / 14976 out tokens · 144552 ms · 2026-08-03T00:22:17.861850+00:00 · methodology

0 comments

read the original abstract

In \cite{GNS} we proved that, for every $\alpha>0$, the potential of three positive point charges has at most $12$ nondegenerate equilibrium points. We also observed that the same method would give the sharper bound $6$ if a certain auxiliary polynomial system $Q=R=0$ had at least four solutions, counted with multiplicity, in each open quadrant of the $(f,g)$-plane. Here we prove this four-solution statement. The main new ingredient is a separation argument at the unique saddle point of a separated-variable first integral. Consequently, the upper bound for three charges improves from $12$ to $6$.

Figures

Figures reproduced from arXiv: 2607.28785 by A.Gabrielov, B.Shapiro, Dm.Novikov, T.Novikov.

**Figure 1.** Figure 1: A positive-quadrant configuration with four equilibria. The orange curve is the oval O+ ⊂ Γ; the green and purple curves are γ2 and γ1, respectively; and the dashed curve is the separatrix level Φ = c0. Red points mark critical points of Φ [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗

discussion (0)

Reference graph

Works this paper leans on

9 extracted references · 1 linked inside Pith

[1]

Arathoon, G

Ph. Arathoon, G. Ball, and M. D. Kvalheim,The Maxwell conjecture is false, arXiv:2607.27197 (2026)

Pith/arXiv arXiv 2026
[2]

D. N. Bernstein,The number of roots of a system of equations, Funct. Anal. Appl.9(1975), 183–185

1975
[3]

Edelsbrunner, C

H. Edelsbrunner, C. Fillmore, and G. Oliveira,Counting equilibria of the electrostatic potential, Proc. Lond. Math. Soc. (3)132(2026), no. 6, Paper No. e70163, 32 pp

2026
[4]

Gabrielov, D

A. Gabrielov, D. Novikov, and B. Shapiro,Mystery of point charges, Proc. Lond. Math. Soc. (3)95(2007), no. 2, 443–472

2007
[5]

A. G. Khovanskii,Fewnomials, Translations of Mathematical Monographs, vol. 88, American Mathematical Society, Providence, RI, 1991

1991
[6]

J. C. Maxwell,A Treatise on Electricity and Magnetism, vol. 1, Clarendon Press, Oxford, 1873
[7]

Tsai,Maxwell’s conjecture on three point charges with equal magnitudes, Phys

Y.-L. Tsai,Maxwell’s conjecture on three point charges with equal magnitudes, Phys. D309 (2015), 86–98

2015
[8]

Warwick,Masters of Theory: Cambridge and the Rise of Mathematical Physics, University of Chicago Press, Chicago, 2003; Chapter 6, pp

A. Warwick,Masters of Theory: Cambridge and the Rise of Mathematical Physics, University of Chicago Press, Chicago, 2003; Chapter 6, pp. 286–356

2003
[9]

Zolotov,Upper bounds for the number of isolated critical points via the Thom–Milnor theorem, Anal

V. Zolotov,Upper bounds for the number of isolated critical points via the Thom–Milnor theorem, Anal. Math. Phys.13(2023), Paper No. 81, 18 pp. Department of Mathematics, Purdue University, West Lafayette, Indiana 47907, USA Department of Mathematics, Weizmann Institute of Science, Rehovot, Israel Department of Economics, Northwestern University, Evanston...

2023

[1] [1]

Arathoon, G

Ph. Arathoon, G. Ball, and M. D. Kvalheim,The Maxwell conjecture is false, arXiv:2607.27197 (2026)

Pith/arXiv arXiv 2026

[2] [2]

D. N. Bernstein,The number of roots of a system of equations, Funct. Anal. Appl.9(1975), 183–185

1975

[3] [3]

Edelsbrunner, C

H. Edelsbrunner, C. Fillmore, and G. Oliveira,Counting equilibria of the electrostatic potential, Proc. Lond. Math. Soc. (3)132(2026), no. 6, Paper No. e70163, 32 pp

2026

[4] [4]

Gabrielov, D

A. Gabrielov, D. Novikov, and B. Shapiro,Mystery of point charges, Proc. Lond. Math. Soc. (3)95(2007), no. 2, 443–472

2007

[5] [5]

A. G. Khovanskii,Fewnomials, Translations of Mathematical Monographs, vol. 88, American Mathematical Society, Providence, RI, 1991

1991

[6] [6]

J. C. Maxwell,A Treatise on Electricity and Magnetism, vol. 1, Clarendon Press, Oxford, 1873

[7] [7]

Tsai,Maxwell’s conjecture on three point charges with equal magnitudes, Phys

Y.-L. Tsai,Maxwell’s conjecture on three point charges with equal magnitudes, Phys. D309 (2015), 86–98

2015

[8] [8]

Warwick,Masters of Theory: Cambridge and the Rise of Mathematical Physics, University of Chicago Press, Chicago, 2003; Chapter 6, pp

A. Warwick,Masters of Theory: Cambridge and the Rise of Mathematical Physics, University of Chicago Press, Chicago, 2003; Chapter 6, pp. 286–356

2003

[9] [9]

Zolotov,Upper bounds for the number of isolated critical points via the Thom–Milnor theorem, Anal

V. Zolotov,Upper bounds for the number of isolated critical points via the Thom–Milnor theorem, Anal. Math. Phys.13(2023), Paper No. 81, 18 pp. Department of Mathematics, Purdue University, West Lafayette, Indiana 47907, USA Department of Mathematics, Weizmann Institute of Science, Rehovot, Israel Department of Economics, Northwestern University, Evanston...

2023