Pith. sign in

REVIEW 2 major objections 5 minor 43 references

Nodal surfaces in $\mathbb{P}^3$ and coding theory

T0 review · 2 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Every sextic in $\mathbb{P}^3$ with 65 nodes has, up to isomorphism, the same associated binary linear code—the $[65,12]$ code with weights 24, 32, 40.

desk verdict A streamlined and honest re-proof of a known classification, with speculative but interesting septic candidates; the proof is only as solid as the unshipped LinCode computations behind it. read the letter →

arxiv 2505.17531 v1 pith:IZFUKCQM submitted 2025-05-23 math.CO cs.ITmath.AGmath.IT

Sascha Kurz This is my paper · ORCID
classification math.COcs.ITmath.AGmath.IT MSC 14J7094B05
keywords nodalsurfacesbinarylinearcodesBarthsexticdivisibleevensetsofnodesPlesspowermomentsresidualsexticsinP^3
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 proves that every nodal sextic surface in $\mathbb{P}^3$ with the maximum number of nodes—65, as in the Barth sextic—has one and the same associated binary linear code, up to isomorphism. That code is the $[65,12]_2$ code with nonzero weights $\{24,32,40\}$, weight enumerator $x^0y^{65}+390x^{24}y^{41}+3055x^{32}y^{33}+650x^{40}y^{25}$, and automorphism group of order $15600$. The proof derives strong restrictions from the first four Pless power moment identities and then completes the classification by exhaustive computer enumeration of small divisible codes. The result turns a geometric uniqueness question about maximally nodal sextics into a coding-theoretic statement and gives a concrete point of departure for the next open case, septics in $\mathbb{P}^3$.

What carries the argument

The load-bearing object is the binary linear code whose codewords are the even sets of nodes of a nodal surface: a set $N$ of nodes is even when some divisor $Q$ on the minimal resolution satisfies $2Q \sim \pi^{-1}(N)$. The argument runs on the first four Pless power moment identities, a coordinate-free rewrite of the MacWilliams identities, which are solved for the unknown weight coefficients $a_{24},a_{32},a_{40},a_{48}$ and force the presence of a weight-40 codeword. From there, the residual code construction—restricting the original code to the coordinates outside a codeword's support—reduces the classification to small 4-divisible codes, which are enumerated exhaustively along with the 8-divisible $[n,12,24]_2$ codes for $n\le65$. An extension argument placing the associated code inside a code one dimension larger then selects the unique $[65,12]$ code.

What would settle it

Run an independent exhaustive search (with different software or with publicly checked logs) for binary linear codes of length 65, dimension 12 and minimum distance 24 in which every nonzero weight is divisible by 8; if any code not isomorphic to the one in Lemma 4(3) appears, Theorem 5 is false. Alternatively, find a sextic in $\mathbb{P}^3$ with 65 nodes whose even-set code does not have weight enumerator $x^0y^{65}+390x^{24}y^{41}+3055x^{32}y^{33}+650x^{40}y^{25}$.

Watch

Extended reading notes

Core claim

The central claim, stated as Theorem 5, is that the binary linear code associated to any sextic in $\mathbb{P}^3$ with 65 nodes is unique up to isomorphism. The associated code has length 65, dimension 12, minimum distance 24, and its weight enumerator is $x^0y^{65}+390x^{24}y^{41}+3055x^{32}y^{33}+650x^{40}y^{25}$; its automorphism group has order $15600$. In the paper's own terms, the code is the unique 8-divisible $[65,12,24]_2$ code appearing as case (3) of Lemma 4. The argument first restricts the weight set to $\{24,32,40\}$ by moment identities and residual-code bounds, then uses an extension result to rule out the length-63 and length-64 candidates, leaving exactly the $[65,12]$ code.

Load-bearing premise

The proof depends on the completeness of the exhaustive computer search over small auxiliary codes; if the search missed any case, the uniqueness conclusion could be false.

Editorial extensions

If this is right

  • Every sextic in $\mathbb{P}^3$ with 65 nodes shares the same even-set code: $[65,12]$, with weight enumerator $x^0y^{65}+390x^{24}y^{41}+3055x^{32}y^{33}+650x^{40}y^{25}$.
  • The code's automorphism group has order $15600$, so any symmetry of the even-set structure of a 65-node sextic is constrained by this group.
  • The associated code is projective, which means its dual has minimum distance at least 3; this adds geometric restrictions on which subsets of the 65 nodes can be even.
  • The uniqueness applies to the code, not the surface: there is a 3-parameter family of sextics with 65 nodes, all giving the same code.
  • For septics, the paper produces explicit candidate codes that satisfy all known weight constraints up to the current bound, providing a target for further geometric exclusion.

Reading between the lines

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

  • Beyond the paper: if the enumeration is independently verified, the uniqueness result also gives a deciding test for whether a proposed 65-node sextic is new—compare its associated code to the unique one.
  • Beyond the paper: the same moment-identity and residual-code pipeline could be applied to the 64-node sextic case, where the paper notes only seven candidate codes remain.
  • Beyond the paper: the explicit septic candidate codes give a coding-theoretic sieve for the open problem $\mu(7)\in[99,104]$: showing that no such code is geometrically realizable would improve the upper bound.
  • Beyond the paper: the order-15600 automorphism group of the unique code may correspond to only a subgroup of the symmetries of any particular 65-node sextic, so one could test whether the full group is realized geometrically.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. The paper studies the binary linear code associated to a nodal sextic surface in P^3. Its main result, Theorem 5, asserts that for every sextic with the maximum possible number of 65 nodes, the associated code is unique up to isomorphism and is explicitly an [65,12,{24,32,40}] code with weight enumerator x^0 y^65 + 390 x^24 y^41 + 3055 x^32 y^33 + 650 x^40 y^25 and automorphism group of order 15600. The proof combines analytic coding-theoretic arguments (MacWilliams identities, residual codes, Lemma 1) with exhaustive computer enumerations using the LinCode package (Lemmas 2-4 and part of Theorem 5). The paper also states candidate codes relevant to the open problem of the maximum number of nodes of septics in P^3.

Significance. If the main theorem is correct, it resolves a natural uniqueness question and provides a self-contained statement of a result that has previously circulated only in unpublished preprints, notably [Kur20]. The explicit generator matrices and weight enumerators are useful, and Lemma 1 is a clean analytic argument that does not depend on computation. However, the central proof heavily depends on computer enumerations that are not shipped with the paper, so the result is conditional on the correctness and completeness of those computations as well as on the cited classifications. The paper would be considerably strengthened by making these computations reproducible or by providing certificates.

major comments (2)
  1. [Lemma 3 and proof of Theorem 5] Lemma 3 is load-bearing for Theorem 5, but its proof is a black-box LinCode enumeration: the paper provides neither the enumeration scripts, input files, parameters, version of LinCode, nor logs. The counts of 4-divisible [23,11], [24,11], [25,11] codes are only referenced to [DFG+11, Mil] and [Kur20]. Since the uniqueness conclusion of Theorem 5 depends on the completeness of these classifications, the paper should either include reproducible computational artifacts or give a verifiable description of the enumeration in sufficient detail for an independent check.
  2. [Proof of Theorem 5] The extension check in the proof of Theorem 5 is under-specified. The text says that LinCode verifies that none of the Lemma 4 codes can be extended to a [≤66,13,{24,32,40,48,56}]2 code, but the code C' from [End98] is stated to contain codewords with weights in {16,28,32,36,...}. It is not explained why failure of extension with weights only in {24,32,40,48,56} rules out all possibilities for dimension at least 13, nor is the relationship between the two LinCode steps (first proving non-extension, then extending C to C') made clear. The logical chain should be written out explicitly and the enumeration parameters reported.
minor comments (5)
  1. [Lemma 4, case (2)] The weight enumerator contains the typo '3087c32y32', which should be '3087x32y32'.
  2. [Lemma 6] The sentence 'Up to isomorphism there are unique [51,8,{24,32}] and [54,8,{24,32}] codes, two [55,8,{24,32}] codes, three two [56,8,{24,32}] codes' is garbled; 'three two' should presumably be 'three'.
  3. [Proof of Theorem 5] The phrase 'A generator matrix its given by' should be 'A generator matrix is given by', and the statement 'the code C' in the proof of Theorem 5 is unique' should specify 'unique up to isomorphism'.
  4. [Section 3 and 4] The notation '[≤66,13,{24,32,40,48,56}]2 code' is nonstandard and should be defined, for example by writing 'length n ≤ 66' explicitly whenever it is used.
  5. [Section 4] The phrase 'For each 0 ≤ µ ≤ 65 there exists a sextic' should be 'For every integer µ with 0 ≤ µ ≤ 65 there exists a sextic'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the uniqueness proof is a computer-assisted classification anchored by external coding-theory results; the unshipped enumeration details are reproducibility risks, not circular reductions.

full rationale

Walking the derivation chain: the associated code C of a 65-nodal sextic inherits 8-divisibility, n ≤ 65, k ≥ 12, and d ≥ 24 from Beauville, Catanese, and Endraß (external). Lemma 1 bounds n and forces a weight-40 codeword purely from Pless power moments. Lemma 2 excludes weights 64 and 56 via residual-code length arguments, citing [KK20] for the non-existence of a 4-divisible code of length 9; that is a parameter-free, externally checkable classification, not a restatement of Theorem 5. Lemma 3 is the load-bearing step: the paper reports an exhaustive LinCode enumeration of the 4-divisible residual codes (counts attributed to the external sources [DFG+11, Mil]) and then of the 8-divisible [n,12,24] codes, with the generator matrices given in the author's own [Kur20]. This is a computer-assisted classification whose stated assumptions (divisibility, dimension, minimum distance, effective length) do not include the target statement, so under the stated review rules it counts as independent support even though scripts and logs are not shipped; the remark that the computation now takes under two hours on a single core is a runtime report, not a definitional input. Lemma 4 then intersects that classification with [JR97]'s exclusion of weights 48 and 64, yielding the three candidates. Theorem 5's selection of the [65,12,24] code uses Endraß's external existence theorem for the extension C' and a further LinCode non-extension check. No equation is set equal to its own input, and no fitted parameter is renamed as a prediction. The paper itself flags an omitted proof: the weight-48 exclusion is relegated to "exhaustive computer enumeration within seconds" without details, and the non-extension check in Theorem 5 is stated tersely. Those are completeness and reproducibility risks, not circularity, so the honest finding is no significant circularity.

Assumptions & free parameters 0 free parameters · 9 assumptions · 0 invented entities

No constants are fitted to data. The central claim rests on standard coding-theory identities and on external computational classifications, several from the author's own prior work; the completeness and correctness of those computations are the main unverified inputs.

assumptions (9)
  • standard math MacWilliams identities and the first four Pless power moments are valid for binary linear codes.
    Used in Lemma 1 to derive bounds on weights; standard coding theory.
  • domain assumption The code associated to a nodal surface is a binary linear code whose length n <= m, dimension k >= m - ceil(s^3/2) + 2s^2 - 3s + 1, divisibility 4 or 8 depending on degree, and minimum distance d >= 2*ceil(s(s-2)/2).
    Cited from [Bea79], [JR97], [Cat81], [End98]; used to fix parameters for the sextic case.
  • standard math The residual code of a q^r-divisible code is q^(r-1)-divisible.
    Used throughout Lemmas 1-3, cited [HKK18, Lemma 7].
  • domain assumption No 4-divisible binary code of length 9 exists.
    Used in Lemma 2 to exclude weight-56 codewords; cited [KK20].
  • domain assumption The lists of 4-divisible [23,11], [24,11], [25,11] codes from [DFG+11] and [Mil] are complete.
    Used in Lemma 3 as the starting set for the residual-code enumeration.
  • domain assumption The [64,13,24] code exhibited is the unique 8-divisible [64,13,24] code.
    Used in the remark after Lemma 1 and indirectly in the classification of Lemma 3; cited from [Kur20].
  • ad hoc to paper LinCode exhaustively and correctly enumerates all codes up to isomorphism in the claimed parameter ranges.
    The proof of Lemma 3 and Theorem 5 relies on these computations; no scripts or certificates are supplied.
  • domain assumption For a sextic with 65 nodes there exists a code C' properly containing C with dim(C') = dim(C)+1 and C'\C weights in {16,28,32,36,...}.
    Used in the proof of Theorem 5; cited from [End98].
  • domain assumption Minimum length data for [n,8,24] codes from [BJV00] are correct.
    Used in Lemma 6 for the [56,k,{24,32}] classification.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Nodal surfaces in $\mathbb{P}^3$ and coding theory." pith.science (2026). https://pith.science/paper/IZFUKCQM

@misc{pith2026250517531,
  author       = {Pith},
  title        = {Pith review of: Nodal surfaces in $\mathbbP^3$ and coding theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IZFUKCQM}},
  note         = {Machine review of arXiv:2505.17531}
}
abstract

To each nodal hypersurface one can associate a binary linear code. Here we show that the binary linear code associated to sextics in $\mathbb{P}^3$ with the maximum number of $65$ nodes, as e.g. the Barth sextic, is unique. We also state possible candidates for codes that might be associated with a hypothetical septic attaining the currently best known upper bound for the maximum number of nodes.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 43 canonical work pages

  1. [1]

    Two projective surfaces with many nodes, admitting the symmetries of the icosahedron

    Wolf Paul Barth. Two projective surfaces with many nodes, admitting the symmetries of the icosahedron. Journal of Algebraic Geometry , 5(1):173--186, 1996

    work page 1996

  2. [2]

    The maximum number of double points on a surface

    Alfred Barnard Basset. The maximum number of double points on a surface. Nature , 73(1889):246, 1906

    work page 1906

  3. [3]

    Dual transform and projective self-dual codes

    Iliya Bouyukliev and Stefka Bouyuklieva. Dual transform and projective self-dual codes. Advances in Mathematics of Communications , 18:328--341, 2024

    work page 2024

  4. [4]

    Computer classification of linear codes

    Iliya Bouyukliev, Stefka Bouyuklieva, and Sascha Kurz. Computer classification of linear codes. IEEE Transactions on Information Theory , 67:7807--7814, 2021

    work page 2021

  5. [5]

    Sur le nombre maximum de points doubles d’une surface dans P ^3 ( (5)= 31 )

    Arnaud Beauville. Sur le nombre maximum de points doubles d’une surface dans P ^3 ( (5)= 31 ). Journ\'ees de G\'eom\'etrie alg\'ebraique d'Angers , pages 207--215, 1979

    work page 1979

  6. [6]

    The smallest length of eight-dimensional binary linear codes with prescribed minimum distance

    Iliya Bouyukliev, David Benjamin Jaffe, and Vesselin Vavrek. The smallest length of eight-dimensional binary linear codes with prescribed minimum distance. IEEE Transactions on Information Theory , 46(4):1539--1544, 2000

    work page 2000

  7. [7]

    Cusps and codes

    Wolf Paul Barth and S awomir Rams. Cusps and codes. Mathematische Nachrichten , 280(1-2):50--59, 2007

    work page 2007

  8. [8]

    Babbage's conjecture, contact of surfaces, symmetric determinantal varieties and applications

    Fabrizio Catanese. Babbage's conjecture, contact of surfaces, symmetric determinantal varieties and applications. Inventiones Mathematicae , 63(3):433--465, 1981

    work page 1981

Show all 43 references
  1. [9]

    Generalized K ummer surfaces and differentiable invariants of N oether- H orikawa surfaces I

    Fabrizio Catanese. Generalized K ummer surfaces and differentiable invariants of N oether- H orikawa surfaces I . In Eugenio Calabi, editor, Manifolds and Geometry , volume 36 of Symposia Mathematica , pages 132--177. Cambridge University Press, Cambridge, 1996

    1996

  2. [10]

    A memoir on cubic surfaces

    Arthur Cayley. A memoir on cubic surfaces. Philosophical Transactions of the Royal Society of London , 159:231--326, 1869

  3. [11]

    Constructing sextic surfaces with a given number d of nodes

    Fabrizio Catanese and Giuseppe Ceresa. Constructing sextic surfaces with a given number d of nodes. Journal of Pure and Applied Algebra , 23(1):1--12, 1982

    1982

  4. [12]

    Even sets of nodes are bundle symmetric

    Gianfranco Casnati and Fabrizio Catanese. Even sets of nodes are bundle symmetric. Journal of Differential Geometry , 47(2):237--256, 1997

    1997

  5. [13]

    Errors in the paper:`` E ven sets of nodes are bundle symmetric''

    Gianfranco Casnati and Fabrizio Catanese. Errors in the paper:`` E ven sets of nodes are bundle symmetric''. Journal of Differential Geometry , 50(2):415--415, 1998

    1998

  6. [14]

    Varieties of nodal surfaces, coding theory and discriminants of cubic hypersurfaces

    Fabrizio Catanese, Stephen Coughlan, Davide Frapporti, Alessandro Verra, and Michael Kiermaier. Varieties of nodal surfaces, coding theory and discriminants of cubic hypersurfaces. arXiv preprint 2206.05492 , 2022

    2022 arXiv

  7. [15]

    Even sets of nodes on sextic surfaces

    Fabrizio Catanese and Fabio Tonoli. Even sets of nodes on sextic surfaces. Journal of the European Mathematical Society , 9(4):705--737, 2007

    2007

  8. [16]

    Doran, Michael G

    Charles F. Doran, Michael G. Faux, Sylvester James Gates, Tristan H \"u bsch, Kevin M. Iga, Gregory D. Landweber, and Robert L. Miller. Codes and supersymmetry in one dimension. Advances in Theoretical and Mathematical Physics , 15(6):1909--1970, 2011

    1909

  9. [17]

    Alternating bilinear forms over GF (q)

    Philippe Delsarte and Jean-Marie Goethals. Alternating bilinear forms over GF (q) . Journal of Combinatorial Theory, Series A , 19(1):26--50, 1975

    1975

  10. [18]

    Some new results on the minimum length of binary linear codes of dimension nine

    Stefan Dodunekov, Sugi Guritman, and Juriaan Simonis. Some new results on the minimum length of binary linear codes of dimension nine. IEEE Transactions on Information Theory , 45(7):2543--2546, 1999

    1999

  11. [19]

    Codes and projective multisets

    Stefan Dodunekov and Juriaan Simonis. Codes and projective multisets. The Electronic Journal of Combinatorics , 5:1--23, 1998

    1998

  12. [20]

    Minimal even sets of nodes

    Stephan Endra . Minimal even sets of nodes. Journal f\"ur die reine und angewandte Mathematik (Crelles Journal) , 503:87--108, 1998. arXiv preprint alg-geom/9710025

    1998 arXiv

  13. [21]

    Mass formulas for self-dual codes over Z _4 and F _q+u F _q

    Philippe Gaborit. Mass formulas for self-dual codes over Z _4 and F _q+u F _q . IEEE Transactions on Information Theory , 42(4):1222--1228, 1996

    1996

  14. [22]

    Partial spreads and vector space partitions

    Thomas Honold, Michael Kiermaier, and Sascha Kurz. Partial spreads and vector space partitions. In Marcus Greferath, Mario Osvin Pav c evi \'c , Natalia Silberstein, and Mar \' a \'A ngeles V \'a zquez-Castro, editors, Network Coding and Subspace Designs , pages 131--170. Spri...

    2018

  15. [23]

    Optimal binary linear codes of length 30

    David Benjamin Jaffe. Optimal binary linear codes of length 30 . Discrete Mathematics , 223(1-3):135--155, 2000

    2000

  16. [24]

    A sextic surface cannot have 66 nodes

    David Benjamin Jaffe and Daniel Ruberman. A sextic surface cannot have 66 nodes. Journal of Algebraic Geometry , 6(1):151--168, 1997

    1997

  17. [25]

    Dieter Jungnickel and Vladimir D. Tonchev. The classification of antipodal two-weight linear codes. Finite Fields and Their Applications , 50:372--381, 2018

    2018

  18. [26]

    On the lengths of divisible codes

    Michael Kiermaier and Sascha Kurz. On the lengths of divisible codes. IEEE Transactions on Information Theory , 66(7):4051--4060, 2020

    2020

  19. [27]

    Classification of -divisible linear codes spanned by codewords of weight

    Michael Kiermaier and Sascha Kurz. Classification of -divisible linear codes spanned by codewords of weight . IEEE Transactions on Information Theory , 69(6):3544--3551, 2023

    2023

  20. [28]

    U ber die F l \

    Ernst Eduard Kummer. \"U ber die F l \"a chen vierten G rades mit sechzehn singul \"a ren P unkten. Collected papers II , pages 418--432, 1864

  21. [29]

    Classification of 8 -divisible binary linear codes with minimum distance 24

    Sascha Kurz. Classification of 8 -divisible binary linear codes with minimum distance 24 . arXiv preprint 2012.06163 , 2020

    2012 arXiv

  22. [30]

    Computer classification of linear codes based on lattice point enumeration

    Sascha Kurz. Computer classification of linear codes based on lattice point enumeration. In International Congress on Mathematical Software , pages 97--105. Springer, 2024

    2024

  23. [31]

    Hypersurfaces with many singularities -- History, Constructions, Algorithms, Visualization

    Oliver Labs. Hypersurfaces with many singularities -- History, Constructions, Algorithms, Visualization . PhD thesis, Mainz, Germany, 2005

    2005

  24. [32]

    A septic with 99 real nodes

    Oliver Labs. A septic with 99 real nodes. Rendiconti del Seminario Matematico della Università di Padova , 116:299--313, 2006. arXiv preprint math/0409348

    2006 arXiv

  25. [33]

    Combinatorial problems of elementary abelian groups

    Florence Jessie MacWilliams. Combinatorial problems of elementary abelian groups . PhD thesis, Radcliffe College, 1962

    1962

  26. [34]

    A theorem on the distribution of weights in a systematic code

    Florence Jessie MacWilliams. A theorem on the distribution of weights in a systematic code. Bell System Technical Journal , 42(1):79--94, 1963

    1963

  27. [35]

    Robert L. Miller. Doubly-even codes. https://rlmill.github.io/de_codes/. Accessed: 2025-04-11

    2025

  28. [36]

    On nodal determinantal quartic hyperfurfaces in P ^4

    Kjell Fredrik Pettersen. On nodal determinantal quartic hyperfurfaces in P ^4 . PhD thesis, Oslo, Norway, 1998

    1998

  29. [37]

    Power moment identities on weight distributions in error correcting codes

    Vera Pless. Power moment identities on weight distributions in error correcting codes. Information and Control , 6(2):147--152, 1963

    1963

  30. [38]

    On W ahl's proof of (6)= 65

    Roberto Pignatelli and Fabio Tonoli. On W ahl's proof of (6)= 65 . Asian Journal of Mathematics , 13(3):307--310, 2009

    2009

  31. [39]

    On the distribution of surfaces of the third order into species, in reference to the absence or presence of singular points, and the reality of their lines

    Ludwig Schl \"a fli. On the distribution of surfaces of the third order into species, in reference to the absence or presence of singular points, and the reality of their lines. Philosophical Transactions of the Royal Society of London , (153):193--241, 1863

  32. [40]

    Mac W illiams identities and coordinate partitions

    Juriaan Simonis. Mac W illiams identities and coordinate partitions. Linear Algebra and its Applications , 216:81--91, 1995

    1995

  33. [41]

    Semicontinuity of the spectrum and an upper bound for the number of singular points of the projective hypersurface

    Alexander Nicholaevich Varchenko. Semicontinuity of the spectrum and an upper bound for the number of singular points of the projective hypersurface. In Doklady Akademii Nauk SSSR , volume 270, pages 1294--1297, 1983

    1983

  34. [42]

    Genus three curves and 56 nodal sextic surfaces

    Bert Van Geemen and Yan Zhao. Genus three curves and 56 nodal sextic surfaces. Journal of Algebraic Geometry , 27(4):583--592, 2018

    2018

  35. [43]

    Enumeration of combinations of rational double points on quartic surfaces

    Jin-Gen Yang. Enumeration of combinations of rational double points on quartic surfaces. AMS/IP Studies in Advanced Mathematis , 5:275--312, 1997

    1997

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.