Pith. sign in

REVIEW 4 major objections 5 minor 40 references

Sharp estimates for the Fourier transform of surface-carried measures and maximal operators associated with hypersurfaces in $\mathbb{R}^4$ with vanishing Gaussian curvature

T0 review · 4 major / 5 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read For polynomial hypersurfaces in R^4 with zero Gaussian curvature, the Newton height h(φ) fixes the sharp Fourier decay rate 1/h(φ) and, for h(φ) ≥ 2, the exact L^p boundedness exponent of the maximal operator.

desk verdict Genuine but narrow advance on zero-curvature hypersurfaces in R^4; referee it, but verify the imported Hessian-zero classification. read the letter →

arxiv 2602.18163 v2 pith:CAXPYIB6 submitted 2026-02-20 math.CA

Isroil A. Ikromov , Gayrat Toshpulatov This is my paper · ORCID
classification math.CA MSC 42B2542B10
keywords Fouriertransformofsurface-carriedmeasuresmaximaloperatorshypersurfacesoscillatoryintegralsNewtonpolyhedronadaptedcoordinatesystemvanishingGaussiancurvatureheight
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

Polynomial hypersurfaces in R^4 whose Gaussian curvature vanishes identically are governed, this paper argues, by a single number: the height h(φ) of the polynomial phase φ that graphs the surface. For such surfaces the Fourier transform of the surface-carried measure decays like (1+|ξ|)^{−1/h(φ)} times at most one logarithm, with the decay sharp along the normal direction — the first sharp three-dimensional results of this kind. The same height controls the maximal operator: when h(φ) ≥ 2 it is bounded on L^p(R^4) exactly for p > h(φ), and the same criterion holds for h(φ) < 2 when the Hessian at the origin vanishes, while the two-curvature case gives p > 3/2. The route is a classification of polynomials with identically vanishing Hessian determinant into a two-variable form or the form Q1(x1) + Q2(x1)x2 + Q3(x1)x3, which makes an adapted coordinate system available and reduces the analysis to known two-dimensional estimates. As a consequence, all four oscillation and contact indices of the surface equal 1/h(φ).

What carries the argument

The central object is the Newton polyhedron of φ at the origin — the convex hull of the exponent vectors in the Taylor support, shifted by the positive orthant — and its height h(φ), the supremum over smooth coordinate systems of the distance from the origin to the polyhedron along the diagonal. The load-bearing structural input is a classification theorem from the algebraic literature: every polynomial in three variables with identically zero Hessian determinant is, after an invertible linear change of variables, either a function of at most two variables or of the form Q1(x1) + Q2(x1)x2 + Q3(x1)x3. From this normal form the paper constructs an adapted coordinate system in which the distanc

What would settle it

Find a three-variable polynomial φ with det(D²φ) ≡ 0 that cannot be reduced by any invertible linear change to two variables or to Q1(x1)+Q2(x1)x2+Q3(x1)x3 — counterexamples to the analogous classification are known in five variables, so three variables is the place to search. Alternatively, for a model phase such as φ = x1²x2 (height 2), compute lim_{ξ4→∞} ξ4^{1/2}∫e^{iξ4x1²x2}η(x)dx with η(0) ≠ 0: if the limit is not a non-zero constant, the sharpness claim (11) fails.

Watch

Extended reading notes

Core claim

The paper claims that for any polynomial φ: R^3 → R with φ(0)=0, ∇φ(0)=0 and det(D²φ) ≡ 0, the oscillatory integral giving the Fourier transform of the surface-carried measure satisfies |∫ e^{i(ξ4φ + ξ1x1 + ξ2x2 + ξ3x3)}η dx| ≤ C||η||_{C³}(log(2+|ξ|))^ν (1+|ξ|)^{−1/h(φ)}, with ν ∈ {0,1} tracking whether the principal face of the Newton polyhedron is a vertex when h ≥ 2; along the normal direction ξ1=ξ2=ξ3=0 the estimate is sharp, in that ξ4^{1/h}(log ξ4)^{−ν} times the integral tends to a non-zero constant. It further claims the maximal operator is bounded on L^p(R^4) for p > max{h(φ),2}, the necessary condition p > h(φ) is also sufficient when h(φ) ≥ 2, and also when h(φ) < 2 provided D²φ(0

Load-bearing premise

Everything hangs on the classification, supplied by another paper, that every three-variable polynomial whose Hessian determinant vanishes identically can be linearly transformed into either two variables or the form Q1(x1)+Q2(x1)x2+Q3(x1)x3; if that classification has an unlisted exceptional form, the sharp estimates do not cover the full class claimed.

Editorial extensions

If this is right

  • Every polynomial phase in the class has sharp Fourier decay exponent 1/h(φ) along the normal direction, with at most one logarithmic factor; no polynomial phase of this type decays faster.
  • When h(φ) ≥ 2, the maximal operator is bounded on L^p(R^4) exactly for p > h(φ), so the boundedness exponent equals the Newton height.
  • When h(φ) < 2 and the Hessian at the origin vanishes, boundedness holds exactly for p > h(φ); when two principal curvatures are non-zero, it holds for p > 3/2.
  • The uniform oscillation, oscillation, uniform contact, and contact indices all equal 1/h(φ), so the height determines the complete asymptotic picture for this class.
  • The estimates are stable under sufficiently small analytic perturbations of φ, and two long-standing conjectures linking Fourier decay to maximal boundedness are confirmed for these hypersurfaces.

Reading between the lines

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

  • The classification step is the true reach of the method: the paper's final section sketches the same results for smooth phases admitting the same normal form, but a smooth phase with det(D²φ) ≡ 0 outside the normal form would escape the exponent 1/h(φ).
  • The unresolved h(φ) < 2 case with exactly one non-zero principal curvature is tied to the same open problem in R^3; since the proof reduces the averaging operator to a two-variable phase, progress in three dimensions would likely settle it in R^4 as well.
  • Sharpness is proved only in the normal direction; testing (10) for oblique frequency directions — for example, numerically or via model phases with fold-type singularities — would show whether the logarithmic factor ν is genuinely needed off the normal.
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

4 major / 5 minor

Summary. The paper studies polynomial graph hypersurfaces in R^4 with identically zero Gaussian curvature, i.e. phases ϕ(x1,x2,x3) with ϕ(0)=0, ∇ϕ(0)=0, and det D²ϕ≡0. It proves: existence of adapted coordinates (Theorem 3.2) via the algebraic classification Theorem 3.1 quoted from [4]; sharp uniform Fourier decay of surface-carried measures with exponent 1/h(ϕ) and at most one logarithmic factor (Theorem 4.1); stability under small analytic perturbations (Corollary 4.2); L^p boundedness of the associated maximal operator for p>max{h(ϕ),2} (Theorem 5.1), with sharp threshold p>h(ϕ) when h(ϕ)≥2 (Corollary 5.3) and in the case h(ϕ)<2 with D²ϕ(0)=0 (Theorem 5.5); the integrability criterion for |ϕ|^{-1/p} (Proposition 5.2); and equality of the uniform oscillation, oscillation, uniform contact, and contact indices to 1/h(ϕ) (Theorem 6.1). The main arguments reduce the zero-Hessian polynomial to a two-variable phase or the normal form Q1(x1)+Q2(x1)x2+Q3(x1)x3, then apply prior 2D results of Varchenko, Ikromov–Müller, and others.

Significance. If the classification and reductions are correct, the paper gives a sharp, invariant description of Fourier decay and maximal-operator boundedness for a natural class of degenerate non-convex hypersurfaces in R^4. A notable strength is that the exponent is the coordinate-invariant Newton height h(ϕ), with no fitted constants, and the results confirm the Iosevich–Sawyer and Stein–Iosevich–Sawyer conjectures in this setting. The paper is, however, heavily dependent on the external algebraic classification Theorem 3.1, and several proof steps in the genuinely 3D cases are terse or contain incorrect statements. These issues are repairable, but they make the manuscript unsuitable for acceptance in its present form.

major comments (4)
  1. [Theorem 3.1; used throughout §§4–5] The entire reduction rests on Theorem 3.1, quoted from [4, Thm 3.3] without proof. The statement asserts an invertible matrix A∈R^{3×3}, while the classification in the literature is often stated over C. The authors should either prove the real version or explain why the complex normal form can be chosen real. If the quoted theorem has any unlisted exceptional case, Theorems 4.1, 5.1, 5.5, and 6.1 do not cover the full polynomial class announced in the abstract. This is a verification risk rather than a detected counterexample, but it is load-bearing.
  2. [Theorem 4.1, proof, Part 2, Case 1] In the proof of sharpness (11), the text says that Erdélyi's lemma gives a nonzero constant C independent of x2 and x3. For the phase λ x1^{ν1} Q̃(x1,x2,x3), the leading asymptotic constant is proportional to Q̃(0,x2,x3)^{-1/ν1}, which depends on x2,x3. The conclusion (11) can still be recovered by dominated convergence with a continuous nonzero limit function of (x2,x3), but the proof as written is incorrect. Please correct this step.
  3. [Theorem 5.1, proof, Case 1] The statement that the hyperplane {t1=ν1} touches the Newton polyhedron of ϕ 'only at (ν1,0,0)' is false. For ϕ=x1^{ν1}Q̃ with Q̃(0)≠0, the Newton polyhedron is the half-space t1≥ν1, so the supporting hyperplane contains the entire unbounded face {ν1}×R^2_+. The dyadic decomposition that follows can still be justified by the subsequent verification that ∂_{x1}²φ_k≠0 on the annulus, but the geometric claim should be corrected or removed.
  4. [Corollary 4.2] The proof of the stability result is only a paragraph. Corollary 3.4 gives a local analytic change reducing the unperturbed ϕ to a function of two variables, but the perturbed phase ϕ+Φ is not a small 2D perturbation of the reduced phase after that change: it becomes a 3D perturbation, and the linear frequency terms are transformed. Karpushkin's 2D stability theorem therefore does not apply directly. Please provide the reduction, for example by splitting according to the size of the transformed ξ3-frequency and then applying the 2D stability result with parameters.
minor comments (5)
  1. [Theorem 5.1, proof, Case 2] The displayed bound after 'this is equivalent to' contains a typo: 2^{k(1/p - 3k/(2ν2+1))} should read 2^{k(1/p - 3/(2ν2+1))}; the next sentence uses the corrected exponent.
  2. [Theorem 3.2, proof, Part 1(a)] In the displayed formula for Φ(y), 'Q(φ_1^ν(y))' should be 'Q(φ_1(y))'.
  3. [Section 4, proof of Theorem 4.1, Part 1] The estimate ∫_R ||η(·,·,x3)||_{C^3(R^2)} dx3 ≤ diam(U)||η||_{C^3(R^2)} is notated inconsistently; the norm on the right should be over R^3 or over U. Similarly, in Part 2 Case 1, 'C^1' norms are used without precise domains.
  4. [Corollary 5.3] The phrase 'the maximal operator M with ρ(x0)>0' is imprecise because the operator in (15) is defined through η, the projected density. It would be clearer to say 'with η(0)>0' throughout.
  5. [Section 4, definition of ν(ϕ)] The definition of ν(ϕ) refers to Varchenko's exponent and the dimension of the principal face, but the precise relation is only sketched. A short explanation or a pointer to the exact statement in [19] and [38] would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the arguments reduce to an external algebraic classification and prior lower-dimensional theorems; no parameter is fitted and no target estimate is assumed by construction.

full rationale

The paper's central claims are not circular. The algebraic reduction in Theorem 3.1 is quoted from an external source, de Bondt--van den Essen [4], and does not itself contain the Fourier or maximal estimates. Theorem 3.2 and Corollary 3.4 construct adapted coordinate systems by explicit calculations; in the two-variable case they invoke the prior two-dimensional adapted-coordinate theorem [18], which is a published result whose assumptions do not include the present four-dimensional conclusions. Theorem 4.1 reduces the three-variable oscillatory integral to the two-dimensional theorem [19, Theorem 1.1], Varchenko's theorem [38], and Erdelyi's lemma, none of which contains the claimed uniform three-variable estimate. The exponent -1/h(phi) is determined by the intrinsic Newton-polyhedron height h(phi), which is defined independently of the desired estimate and is not fitted to any data or to the conclusion. Theorem 5.1 similarly uses prior maximal-operator results [17, Theorem 1.3] and [21, Theorem 7.1] after either reducing to at most two variables or performing a dyadic decomposition; these are lower-dimensional or structurally simpler results, not the target theorem. The necessary condition in Proposition 5.2 is proved directly from the normal forms with explicit integrability computations. Theorem 6.1 is an immediate consequence of the established estimates and the known index relationships from [17, Theorem 1.14]; the indices are not defined to equal 1/h by fiat. Several heavily used citations are by the same authors, but they are genuine published theorems with independent content and do not assume the current results. The main verification risk, the unproved external classification Theorem 3.1, is a concern about correctness of an external algebraic input, not a circularity.

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

The paper's narrative rests on the cited algebraic classification of Hessian-zero polynomials and on the authors' prior sharp 2D estimates; these are external published results, not derived here. No numerical constants are fitted; h(φ) is a coordinate-invariant geometric quantity, and the indices β_u, β, γ_u, γ are existing definitions.

assumptions (6)
  • standard math Theorem 3.1 (de Bondt–van den Essen classification): every polynomial φ: R^3 → R with det(D²φ)≡0 is, after an invertible linear change, either independent of one variable or of the form Q1(x1)+Q2(x1)x2+Q3(x1)x3.
    Used in §§3–5 to reduce all phase functions to tractable normal forms; not proved in the paper.
  • standard math Ikromov–Müller [19, Theorem 1.1]: sharp uniform oscillatory estimates for two-variable phases.
    Applied in §4 Part 1 and Part 2 Case 2 to the inner two-dimensional integrals.
  • standard math Ikromov–Müller et al. [17, Theorem 1.3] and Buschenhenke–Ikromov–Müller [6, Theorem 1.2]: maximal operator bounds for hypersurfaces in R^3.
    Used in §5 Part 1 and Theorem 5.5(ii) after reduction to surfaces depending on two variables.
  • standard math Varchenko [38, Theorem 0.6]: sharp two-dimensional asymptotic expansions with exponent 1/h.
    Used in §4 to establish the sharpness limit (11).
  • standard math Karpushkin [25]: stability of Varchenko's estimate under small analytic perturbations in two dimensions.
    Invoked in the proof of Corollary 4.2.
  • domain assumption Transversality assumption for the maximal operator: affine tangent planes to S do not pass through the origin.
    Standard in the maximal-operator setting (§5, after (14)); without it, the boundedness behaviour can change significantly.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Sharp estimates for the Fourier transform of surface-carried measures and maximal operators associated with hypersurfaces in $\mathbb{R}^4$ with vanishing Gaussian curvature." pith.science (2026). https://pith.science/paper/CAXPYIB6

@misc{pith2026260218163,
  author       = {Pith},
  title        = {Pith review of: Sharp estimates for the Fourier transform of surface-carried measures and maximal operators associated with hypersurfaces in $\mathbbR^4$ with vanishing Gaussian curvature},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CAXPYIB6}},
  note         = {Machine review of arXiv:2602.18163}
}
abstract

In this paper, we study problems related to harmonic analysis on hypersurfaces in $\mathbb{R}^4 $ with zero Gaussian curvature and given as graphs of polynomial functions. We derive sharp uniform estimates with respect to the direction of frequencies for the Fourier transform of measures supported on such hypersurfaces. Additionally, we study the $L^p$-boundedness problem of maximal operators associated with hypersurfaces. We determine the exact value of the boundedness exponent in terms of the heights of these hypersurfaces.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 1 linked inside Pith

  1. [4]

    M. C. de Bondt and A. van den Essen. Singular Hessians.J. Algebra, 282(1), 195–204, 2004

    2004

  2. [1]

    V. I. Arnold. Remarks on the method of stationary phase and on the Coxeter numbers. Uspekhi Mat. Nauk., 28(5), 17–44, 1973. English transl.Russ. Math. Surv.,28(5), 19-58, 1973

    1973

  3. [2]

    V. I. Arnold, S. M. Gusein-Zade, V. Varchenko.Singularities of differentiable maps. Vol. I, The classification of critical points, Caustics and Wave Fronts.Birkh¨ auser, Boston-Basel-Stuttgart, 1985

    1985

  4. [3]

    M. F. Atiyah. Resolution of Singularities and division of distributions.Comm. Pure and Appl. Math., 23(2), 145–150, 1970

    1970

  5. [5]

    Bourgain

    J. Bourgain. Averages in the plane over convex curves and maximal operators.J. Anal. Math., 47, 69–85, 1986

    1986

  6. [6]

    Buschenhenke, S

    S. Buschenhenke, S. Dendrinos, I. A. Ikromov, D. M¨ uller. Estimates for maximal functions associated to hypersurfaces inR 3 with heighth <2: Part I.Trans. Amer. Math. Soc., 372(2), 1363–1406, 2019

    2019

  7. [7]

    Buschenhenke, I

    S. Buschenhenke, I. A. Ikromov, D. M¨ uller. Estimates for maximal functions associated to hypersurfaces inR 3 with heighth <2: Part II. A geometric conjecture and its proof for generic 2-surfaces,Ann. Sc. Norm. Super. Pisa Cl. di Sc., (5) XXVI, 1765–1877, 2025

    2025

  8. [8]

    T. C. Collins, A. Greenleaf, M. Pramanik. A multi-dimensional resolution of singularities with applications to analysis.American Journal of Mathematics,135(5), 1179–1252, 2013. 28

    2013

Show all 40 references
  1. [9]

    M. V. Fedoryuk.The Saddle-Point Method.Nauka, Moscow, 1977

    1977

  2. [10]

    Hironaka

    H. Hironaka. Resolution of singularities of an algebraic variety over a field of characteristic zero I.Ann. of Math., 79(1), 109–203, 1964

    1964

  3. [11]

    Gordan, M

    P. Gordan, M. Noether. ¨Uber die algebraischen Formen deren Hesse’sche Determinante identisch verschwindet.Math. Ann.,10, 547–568, 1876

  4. [12]

    Greenleaf

    A. Greenleaf. Principal curvature in harmonic analysis.Indiana Math. J., 30, 519–537, 1981

    1981

  5. [13]

    Greenblatt

    M. Greenblatt. Newton polygons and local integrability of negative powers of smooth functions in the plane.Trans. Amer. Math. Soc.,358, 657–670, 2006

    2006

  6. [14]

    P. T. Gressman. Uniform estimates for cubic oscillatory integrals.Indiana Univ. Math. J., 57(7), 3419–3442, 2008

    2008

  7. [15]

    O. Hesse. ¨Uber die Bedingung, unter welcher eine homogene ganze Function von n unabh¨ angigen Variabeln durch line¨ are Substitutionen von n andern unabh¨ angigen Variabeln auf eine homogene Function sich zur¨ uckf¨ uhren l¨ aßt, die eine Variable weniger enth¨ alt.Journal f¨...

  8. [16]

    O. Hesse. Zur Theorie der ganzen homogenen Functionen.Journal f¨ ur die reine und angewandte Mathematik,56, 263–269, 1859

  9. [17]

    I. A. Ikromov, M. Kempe, D. M¨ uller. Estimates for maximal functions associated to hypersurfaces inR 3 and related problems of harmonic analysis.Acta Math.204, 151–271, 2010

    2010

  10. [18]

    I. A. Ikromov, M. M¨ uller. On adapted coordinate systems.Trans. Amer. Math. Soc.,363(6), 2821–2848, 2011

    2011

  11. [19]

    I. A. Ikromov, D. M¨ uller. Uniform estimates for the Fourier transform of surface carried measures inR 3 and an application to Fourier restriction.J. Fourier Anal. Appl.17(6), 1292–1332, 2011

    2011

  12. [20]

    I. A. Ikromov, D. M¨ uller.Fourier restriction for hypersurfaces in three dimensions and Newton polyhedra. Annals of Mathematics Studies 194, Princeton University Press, Princeton and Oxford 2016; 260 pp

    2016

  13. [21]

    I. A. Ikromov, S. E. Usmanov. On boundedness of maximal operators associated with hypersurfaces.Journal of Mathematical Sciences,264(6), 715–745, 2022

    2022

  14. [22]

    Iosevich

    A. Iosevich. Maximal operators associated to families of flat curves in the plane.Duke Math. J., 76(2), 633–644, 1994

    1994

  15. [23]

    Iosevich, E

    A. Iosevich, E. Sawyer. Oscillatory integrals and maximal averages over homogeneous surfaces.Duke Math. J.,82(1), 103–141, 1996

    1996

  16. [24]

    Iosevich, E

    A. Iosevich, E. Sawyer. Maximal averages over surfaces.Adv. Math.,132, 46–119, 1997

    1997

  17. [25]

    V. N. Karpushkin. A theorem on uniform estimates for oscillatory integrals with a phase depending on two variables.Trudy Sem. Petrovsk., 10, 150–169, 1984 (in Russian); English translation inJ. Soviet Math.,35, 2809–2826, 1986. 29

    1984

  18. [26]

    J. B. Lee, J. Lee, S. Oh. Maximal averages and non-transversality.arXiv preprint, arXiv:2601.01880

  19. [27]

    C. Lossen. When does the Hessian determinant vanish identically?.Bulletin of the Brazilian Mathematical Society,35(1), 71–82, 2004

    2004

  20. [28]

    S. Oh. Maximal estimates for averages over degenerate hypersurfaces.To appear in Trans. Amer. Math. Soc., arxiv:2401.16881,2025

    2025 arXiv

  21. [29]

    M. Pasch. Zur Theorie der Hesseschen Determinante.Journal f¨ ur die reine und angewandte Mathematik,80, 169–176, 1875

  22. [30]

    D. H. Phong, E. M. Stein. The Newton polyhedron and oscillatory integral operators.Acta Math., 179(1), 105–152, 1997

    1997

  23. [31]

    D. H. Phong, E. M. Stein, J. A. Sturm. On the growth and stability of real-analytic functions. Amer. J. Math., 121, 519–554, 1999

    1999

  24. [32]

    H. Schulz. Convex hypersurfaces of finite type and the asymptotics of their Fourier transforms.Indiana Univ. Math. J.40, 1267–1275, 1991

    1991

  25. [33]

    C. D. Sogge, Maximal operators associated to hypersurfaces with one nonvanishing principal curvature, in Fourier Analysis and Partial Differential Equations (Miraflores de la Sierra, 1992), Stud. Adv. Math., pp. 317–323. CRC, Boca Raton, FL, 1995

    1992

  26. [34]

    C. D. Sogge, E. M. Stein. Averages of functions over hypersurfaces inR n.Invent. Math., 82(3), 543–556, 1985

    1985

  27. [35]

    E. M. Stein.Harmonic Analysis: Real-Variable Methods, Orthogonality, and Oscillatory Integrals.Princeton Mathematical Series, vol. 43. Princeton University Press, Princeton, 1993

    1993

  28. [36]

    E. M. Stein, R. Shakarchi.Functional analysis: Introduction to further topics in analysis. Vol. 4, Princeton University Press, 2011

    2011

  29. [37]

    E. M. Stein. Maximal functions. I. Spherical means.Proc. Nat. Acad. Sci. U.S.A.,73(7), 2174–2175, 1976

    1976

  30. [38]

    A. N. Varchenko. Newton polyhedra and estimates of oscillating integrals.Funct. Anal, and Appl., 10(3), 175–196, 1976

    1976

  31. [39]

    Watanabe, M

    J. Watanabe, M. Bondt. On the theory of Gordan-Noether on homogeneous forms with zero Hessian (improved version). InInternational Conference on Polynomial Rings and Affine Algebraic Geometry,pp. 73–107. Cham: Springer International Publishing, 2018

    2018

  32. [40]

    Zimmermann.OnL p-estimates for maximal averages over hypersurfaces not satisfying the transversality condition.PhD thesis, University of Kiel, 2014

    E. Zimmermann.OnL p-estimates for maximal averages over hypersurfaces not satisfying the transversality condition.PhD thesis, University of Kiel, 2014. 30

    2014

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.