Pith. sign in

REVIEW 3 major objections 4 minor 43 references

Fredholm complexes of Hilbert C*-modules

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read The paper proves that a finite-length Hilbert C*-module complex is Fredholm if and only if its even Dirac operator is Fredholm, with the index valued in K0(A).

desk verdict A serious and mostly rigorous extension of Fredholm complex theory to Hilbert C*-modules with unbounded regular differentials, but the main Dirac equivalence currently rests on a deferred, nontrivial lifting theorem that a referee should insist on seeing. read the letter →

arxiv 2505.07568 v1 pith:YDENNVVM submitted 2025-05-12 math.OA math.KT

classification math.OAmath.KT MSC 19K56
keywords complexofHilbertC*-modulesFredholmevenDiracoperatorindexK-theoryHodgedecompositionunboundedregularoperatorsquasicomplex
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 Fredholm theory for complexes of Hilbert C*-modules, the noncommutative analogue of classical Hilbert complexes. Its central theorem states that a finite-length such complex is Fredholm exactly when its even Dirac operator is Fredholm, and that the index of the complex, defined as the index of that operator, lies in the K-theory group K0(A) of the coefficient C*-algebra. The paper also proves that the index is stable under small and relatively compact perturbations and gives Hodge-decomposition formulas for it. If correct, this extends the familiar index theory of elliptic complexes over spaces to arbitrary noncommutative C*-algebras.

What carries the argument

The even Dirac operator D+t is the central object: it is the block operator on the even and odd direct sums of the Hilbert C*-modules, with entries built from the differential maps t2k and the adjoints t*2k-1 in each diagonal block. Rolling up the cochain complex into this single regular operator is what carries the argument, because D+t is Fredholm precisely when the whole complex is Fredholm, and its K0(A)-index is the complex's index. Two supporting mechanisms make the reduction work: the bounded transform Ft = t(1 + t*t)-1/2, which turns unbounded regular operators into adjointable operators while preserving the Fredholm property, and the Calkin-functor that sends a Hilbert A-module E to the quotient LA(Sigma,E)/KA(Sigma,E), converting a quasicomplex into a family of exact complexes over C*-algebras C(Sigma). Characterizations of weak and strong Hodge decomposition are also needed, since not every closed submodule of a Hilbert C*-module is complemented.

What would settle it

A concrete counterexample would settle the central claim: exhibit a finite-length A-Hilbert quasicomplex that is A-Fredholm but whose image under the Calkin-functor is not an exact C(Sigma)-Hilbert complex for some Hilbert A-module Sigma, or, equivalently for complexes, a finite-length A-Hilbert complex whose even Dirac operator is A-Fredholm even though the complex admits no joint parametrix.

Watch

Extended reading notes

Core claim

Theorem 1.11, proved as Corollary 7.18, asserts that a finite-length A-Hilbert complex (E,t) is A-Fredholm if and only if its even Dirac operator D+t is A-Fredholm, and the Fredholm index of the complex is defined as ind((E,t)) := ind(D+t) in K0(A). The proof passes through quasicomplexes: a quasicomplex of adjointable operators is Fredholm exactly when its even Dirac operator is Fredholm, and a complex of unbounded regular operators is Fredholm exactly when its bounded transform is, so the result transfers to the unbounded setting. Along the way the paper shows that the index is invariant under small gap-topology perturbations (Theorem 1.15) and under relatively compact perturbations of the differentials (Theorem 1.16). It also computes the index as [ker(D+t)]0 - [ker(D-t)]0 when a weak Hodge decomposition exists, and as the alternating sum of K-theory classes of the cohomology groups when a strong Hodge decomposition exists (Theorem 1.14).

Load-bearing premise

The whole equivalence rests on Theorem 7.14, whose proof is not given: an A-Hilbert quasicomplex is A-Fredholm if and only if, for every Hilbert A-module Sigma, the induced C(Sigma)-Hilbert complex is exact; if that characterization fails for some C*-algebra, the reduction to the Dirac operator and the K0(A)-index construction would not go through.

Editorial extensions

If this is right

  • For any finite-length A-Hilbert complex, the Fredholm index in K0(A) is well-defined and equals ind(D+t), so index computations reduce to a single regular operator.
  • The index is invariant under small gap-topology perturbations and under relatively compact perturbations of the differential maps, giving homotopy invariance for continuous families of complexes.
  • Under weak Hodge decomposition the index is [ker(D+t)]0 - [ker(D-t)]0; under strong Hodge decomposition it equals the alternating sum Σk (-1)k [Hk((E,t))]0, matching the classical Euler-characteristic formula.
  • The Fredholm property of a complex is equivalent to that of its adjoint, bounded-transform, graph-norm, and Laplace-operator versions, and the index of the adjoint complex is (-1)N+1 times the original index.
  • For complexes over the compact operators on a Hilbert space, and for A-elliptic complexes on compact manifolds, the Fredholm property is tied to finite-dimensional cohomology and the index is computed by the Euler characteristic of the cohomology K-classes.

Reading between the lines

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

  • If the unproved characterization behind Theorem 7.14 is given a full proof, the same strategy would likely extend to infinite-length quasicomplexes and to sequences whose successive compositions are merely compact, transferring index-stability results from bounded operators.
  • The reduction to a single Dirac operator suggests that index computations for noncommutative elliptic complexes could be performed using known K0(A)-valued index formulas for first-order regular operators, such as Callias-type indices.
  • For C*-algebras in which every closed submodule is complemented, such as algebras of compact operators, weak Hodge decomposition always holds, so the formula [ker(D+t)]0 - [ker(D-t)]0 may hold unconditionally there; this is a testable special case of the paper's results.
  • The tensor-product construction for adjointable complexes points toward product formulas for the K0-valued index and toward a noncommutative topological-index theorem, but the paper does not develop these consequences.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper develops a theory of Fredholm complexes of Hilbert C*-modules over a fixed (possibly noncommutative) C*-algebra A. An A-Hilbert complex is a finite sequence of regular operators between Hilbert A-modules satisfying the complex property, and it is called A-Fredholm when it admits a joint parametrix up to compact operators. The main results are Theorem 1.11, asserting that a finite-length A-Hilbert complex is A-Fredholm if and only if its even Dirac operator is A-Fredholm; Definition 1.12, which defines the Fredholm index in K0(A) via the even Dirac operator; Theorem 1.14, giving Hodge-decomposition formulas for the index; and Theorems 1.15 and 1.16, establishing stability of the index under small gap-topology perturbations and relatively compact perturbations. The paper also constructs the bounded transform, adjoint, and graph-norm complexes, characterizes weak and strong Hodge decompositions, discusses Putinar's functor and quasicomplexes, and treats direct sums, exact sequences, tensor products, complexes over the compact operators, and C*-elliptic complexes.

Significance. If the central equivalence holds, this paper provides a substantial noncommutative generalization of the classical Fredholm complex index theory of Segal, Atiyah-Bott, and Brüning-Lesch, with an index taking values in K0(A). The paper is careful about the failure of the Hilbert projection theorem for Hilbert C*-modules and supplies detailed proofs for many technical parts, including the regularity of the Dirac operator (Theorem 3.12), the Hodge decomposition characterizations (Section 4), and the perturbation results (Section 9). It also connects the abstract theory to concrete examples in Section 11. The main weakness is that several load-bearing statements are deferred to external sources or to the first author's thesis, in particular Theorem 7.14, which underpins the main equivalence between Fredholmness of a complex and Fredholmness of its Dirac operator.

major comments (3)
  1. [§7.5.2, Theorem 7.14] The proof of Theorem 7.14 is not given; the text states only that it is similar to [ST98, Theorem 5.1.3]. This theorem is load-bearing: its converse direction is used in Proposition 7.17 to prove that A-Fredholmness of the quasicomplex implies A-Fredholmness of the even Dirac operator, and hence Theorem 1.11 follows. The converse requires lifting exactness of every induced C(Σ)-complex to a chain contraction by adjointable operators on the original modules, and because Hilbert C*-modules do not in general satisfy the Hilbert projection theorem, the Banach-space argument from [ST98] cannot be transplanted without checking that the relevant submodules are complementable. The authors should provide a complete proof or a precise statement of the additional hypotheses needed.
  2. [§5, Lemma 5.5 and related statements] Lemma 5.5, Corollary 5.6, Theorem 5.7, and Lemma 5.9 are stated without proofs and deferred to the master's thesis [VV24]. These chain-homotopy and cohomology-map results are used in the proof of Theorem 7.14 and therefore in the central equivalence Theorem 1.11. Since [VV24] is not a published, readily available source, the paper should include the proofs or at least the precise statements needed for Theorem 7.14.
  3. [§8.3, Remark 8.12] Remark 8.12 explicitly leaves to the reader the well-definedness of the Putinar index, namely the fact that the index of Tev does not depend on the choice of quasicomplex parametrix. This invariance is used in Theorem 8.15 to identify the Putinar index with the Dirac-operator index. The omitted proof should be included, since without it the comparison between the two index constructions is conditional.
minor comments (4)
  1. [§4.2, Theorem 4.9] The proof of the implication (iii)⇒(i) says 'the reader should simply remove the overline and replace polar decomposition by closed range'; this is not a proof and should be rewritten as an explicit argument.
  2. [§11.2] There is a typo in the phrase 'intregro-differential operators'; it should read 'integro-differential operators'.
  3. [§7.5.1, Theorem 7.13] The notation C(Σ) for the Calkin algebra LA(Σ)/KA(Σ) may be confused with the algebra of continuous functions on a space; a different symbol or an explicit warning would improve readability.
  4. [References] The reference [ST98] is a preprint and not widely available; the proof of Theorem 7.14 should either be included in the paper or the relevant theorem statement from [ST98] should be reproduced.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular reduction; Theorem 1.11 rests on an unproved Putinar characterization, a completeness gap rather than circularity.

full rationale

I walked the derivation chain of Theorems 1.11-1.16. The Fredholm index of a complex is defined as ind(D^+_t) (Definition 1.12), and the alternative index formulas for weak/strong Hodge decomposition, the adjoint complex, stability under perturbations, and the Putinar Tev-index are all proved as theorems using Exel's index calculus and the bounded transform; they are not restatements of the definition. Theorem 1.11 is proved via Corollary 7.18 and Proposition 7.17, and the converse direction of Proposition 7.17 uses Theorem 7.14, the characterization of A-Fredholm quasicomplexes by exactness of all induced C(Sigma)-complexes. The proof of Theorem 7.14 is not supplied: the text says 'The proof is similar to [ST98, Theorem 5.1.3], using the results for chain homotopies in Section 5,' and Section 5 states several lemmas without proofs, deferred to [VV24], the first author's master's thesis written under the second author's supervision. This is a genuine robustness/deferral issue: the main Fredholm-Dirac equivalence is conditional on an omitted argument. It is not, however, circularity in the sense targeted here. [ST98] is an external source, and the paper nowhere defines A-Fredholmness or the index in terms of the exactness characterization, nor does it fit a parameter and call the fit a prediction. The self-citations [VV24] and [vdD25] do not assume the target index theorem. There are no fitted inputs, no author-uniqueness theorem forced by citation, and no renaming of a known result as a new derivation. I therefore find no circular step; the low score records only the load-bearing deferral of Theorem 7.14 and the related Section 5 material, which is a completeness concern rather than a circularity.

Assumptions & free parameters 0 free parameters · 8 assumptions · 0 invented entities

The paper uses standard background theorems and imported results; it introduces no new entities or fitted parameters. Its contribution is structural: definitions and connections between existing objects.

assumptions (8)
  • standard math Standard Hilbert C*-module toolkit: inner products, adjointable operators, compact and finite-rank operators, bounded transform, graph norms.
    Used throughout Sections 2 and 3; the paper cites Lance [Lan95] and Wegge-Olsen [WO93].
  • standard math Regular operators on Hilbert C*-modules satisfy the standard functional calculus and resolvent identities listed in Proposition 2.2.
    Foundational to the bounded transform and Dirac operator definitions; taken from [Lan95, Chapters 9-10].
  • standard math Theorem 3.17 from [LM19] on sums of regular self-adjoint operators, used to prove regularity of the Laplace operator.
    Invoked in Theorem 3.18 to show the k-th Laplace operator is self-adjoint and regular.
  • domain assumption Putinar's characterization of Fredholm quasicomplexes via exactness after applying the functor phi_Sigma, as in [Put82] and [ST98, Theorem 5.1.3].
    Theorem 7.14 relies on a proof 'similar to [ST98, Theorem 5.1.3]', not reproduced here; this is the weakest assumption.
  • standard math Exel's index map properties for adjointable A-Fredholm operators [Exe93, Section 3], including additivity, stability, and parametrix behavior.
    Used in Sections 8 and 9 to build the index and prove its properties.
  • standard math K-theory of C*-algebras and finite-rank Hilbert modules [Exe93, RLL00], including K0(A) as the target group for the index.
    Assumed background throughout the index sections.
  • domain assumption The C*-algebra A is fixed and complex; all Hilbert A-modules are over the same A, and finite-length complexes are assumed when needed.
    Sets the scope of the paper (Sections 1.1, 1.4).
  • domain assumption For the C*-elliptic example, the integro-differential operator calculus and symbol map of [ST01, Chapter 2] are taken as given.
    Used in Section 11.2 to derive A-elliptic complexes as examples.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fredholm complexes of Hilbert C*-modules." pith.science (2026). https://pith.science/paper/YDENNVVM

@misc{pith2026250507568,
  author       = {Pith},
  title        = {Pith review of: Fredholm complexes of Hilbert C*-modules},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YDENNVVM}},
  note         = {Machine review of arXiv:2505.07568}
}
read the original abstract

We investigate complexes of Hilbert C*-modules, which are cochain complexes with (unbounded) regular operators between Hilbert C*-modules as differential maps. In particular, we provide various equivalent characterizations of the Fredholm property for such complexes of Hilbert C*-modules, and we define the Fredholm index taking values in the K-theory group of the C*-algebra. Among other properties of this index, we prove the stability under small or relatively compact perturbations, and we obtain alternative expressions for the index under the existence of a (weak or strong) Hodge decomposition.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 28 canonical work pages

  1. [1]

    M. F. Atiyah and R. Bott. A L efschetz fixed point formula for elliptic complexes I 10.2307/1970694 . Annals of Mathematics. Second Series , 86:374--407, 1967

  2. [2]

    M. F. Atiyah and I. M. Singer. The index of elliptic operators I 10.2307/1970715 . Annals of Mathematics. Second Series , 87:484--530, 1968

  3. [3]

    Ambrozie and F.-H

    C.-G. Ambrozie and F.-H. Vasilescu. Banach space complexes 10.1007/978-94-011-0375-6 , volume 334 of Mathematics and its Applications . Kluwer Academic Publishers Group, Dordrecht, 1995

  4. [4]

    Bayramov

    S. Bayramov. On stability of index of F redholm complexes on the C^ -algebra 10.1080/00036810701822306 . Applicable Analysis. An International Journal , 87(4):409--419, 2008

  5. [5]

    Baki \'c and B

    D. Baki \'c and B. Gulja s . http://www.acta.hu/periodica.html?periodica=2002681&paramtipus_ertek=publication&param_ertek=2662 Hilbert C *-modules over C *-algebras of compact operators . Acta Universitatis Szegediensis. Acta Scientiarum Mathematicarum , 68(1-2):249--269, 2002

  6. [6]

    Br \"u ning and M

    J. Br \"u ning and M. Lesch. Hilbert complexes 10.1016/0022-1236(92)90147-B . Journal of Functional Analysis , 108(1):88--132, 1992

  7. [7]

    H. O. Cordes and J. P. Labrousse. The invariance of the index in the metric space of closed operators https://www.jstor.org/stable/24900877. J. Math. Mech. , 12:693--719, 1963

  8. [8]

    Dixmier and A

    J. Dixmier and A. Douady. http://www.numdam.org/item?id=BSMF_1963__91__227_0 Champs continus d'espaces hilbertiens et de C *-alg \`e bres . Bulletin de la S oci \'e t \'e M ath \'e matique de F rance , 91:227--284, 1963

Show all 43 references
  1. [9]

    Eschmeier

    J. Eschmeier. Tensor products and elementary operators 10.1515/crll.1988.390.47 . Journal f \"u r die reine und angewandte Mathematik , 1988(390):47--66, 1988

  2. [10]

    R. Exel. A F redholm operator approach to M orita equivalence 10.1007/BF00961067 . K -Theory , 7(3):285--308, 1993

  3. [11]

    Frank and K

    M. Frank and K. Sharifi. https://www.jstor.org/stable/24715484 Generalized inverses and polar decomposition of unbounded regular operators on H ilbert C *-modules . Journal of Operator Theory , 64(2):377--386, 2010

  4. [12]

    a user Advanced Texts Basler Lehrb \

    J. M. Gracia-Bond \' a, J. C. V \'a rilly, and H. Figueroa. Elements of Noncommutative Geometry 10.1007/978-1-4612-0005-5 . Birkh \"a user Advanced Texts Basler Lehrb \"u cher. Birkh \"a user Boston, MA, 2001

  5. [13]

    M. Joachim. Unbounded F redholm operators and K -theory 10.1142/9789812704443\_0009 , pages 177--199. World Sci. Publ., River Edge, NJ, 2003

  6. [14]

    T. Kato. Perturbation theory for linear operators 10.1007/978-3-642-66282-9 . Classics in Mathematics. Springer-Verlag, Berlin, 1995. Reprint of the 1980 edition

  7. [15]

    Kaad and M

    J. Kaad and M. Lesch. A local global principle for regular operators in H ilbert C *-modules 10.1016/j.jfa.2012.03.002 . Journal of Functional Analysis , 262(10):4540--4569, 2012

  8. [16]

    Kr \'y sl

    S. Kr \'y sl. Hodge theory for elliptic complexes over unital C *-algebras 10.1007/s10455-013-9394-9 . Annals of Global Analysis and Geometry , 45:197--210, 2014

  9. [17]

    Kr \'y sl

    S. Kr \'y sl. Hodge theory for complexes over C *-algebras with an application to A -ellipticity 10.1007/s10455-015-9449-1 . Annals of Global Analysis and Geometry , 47(4):359--372, 2015

  10. [18]

    Kr \'y sl

    S. Kr \'y sl. Elliptic complexes over C *-algebras of compact operators 10.1016/j.geomphys.2015.12.001 . Journal of Geometry and Physics , 101:27--37, 2016

  11. [19]

    Kustermans

    J. Kustermans. The functional calculus for regular operators in H ilbert C *-modules revisited 10.48550/arXiv.funct-an/9706007 . arXiv.org , 1997

  12. [20]

    E. C. Lance. Hilbert C *-modules: A toolkit for operator algebraists 10.1017/CBO9780511526206 , volume 210 of London Mathematical Society Lecture Note Series . Cambridge University Press, Cambridge, 1995

  13. [21]

    Lesch and B

    M. Lesch and B. Mesland. Sums of regular self-adjoint operators in H ilbert C *-modules 10.1016/j.jmaa.2018.11.059 . Journal of Mathematical Analysis and Applications , 472(1):947--980, 2019

  14. [22]

    B. Magajna. Hilbert C *-modules in which all closed submodules are complemented 10.1090/S0002-9939-97-03551-X . Proceedings of the American Mathematical Society , 125(3):849--852, 1997

  15. [23]

    A. S. Mi s c enko and A. T. Fomenko. The index of elliptic operators over C -algebras 10.1070/IM1980v015n01ABEH001207 . Math. USSR Izvestija , 15(1):87--112, 1980

  16. [24]

    G. J. Murphy. C *-algebras and operator theory 10.1016/C2009-0-22289-6 . Academic Press, Inc., Boston, MA, 1990

  17. [25]

    Niknam and K

    A. Niknam and K. Sharifi. The A tkinson theorem in H ilbert C^ -modules over C^ -algebras of compact operators 10.1155/2007/53060 . Abstract and Applied Analysis , pages Art. ID 53060, 7, 2007

  18. [26]

    R. S. Palais. https://www.jstor.org/stable/j.ctt1bd6k0x Seminar on the A tiyah- S inger index theorem , volume No. 57 of Annals of Mathematics Studies . Princeton University Press, Princeton, NJ, 1965. With contributions by M. F. Atiyah, A. Borel, E. E. Floyd, R. T. Seeley, W....

  19. [27]

    G. K. Pedersen. Analysis now 10.1007/978-1-4612-1007-8 , volume 118 of Graduate Texts in Mathematics . Springer-Verlag, New York, 1989

  20. [28]

    M. Putinar. https://www.jstor.org/stable/24714066 Some invariants for semi- F redholm systems of essentially commuting operators . Journal of Operator Theory , 8(1):65--90, 1982

  21. [29]

    R rdam, F

    M. R rdam, F. Larsen, and N. Laustsen. An introduction to K -theory for C *-algebras 10.1017/CBO9780511623806 , volume 49 of London Mathematical Society Student Texts . Cambridge University Press, Cambridge, 2000

  22. [30]

    Rempel and B.-W

    S. Rempel and B.-W. Schulze. Index theory of elliptic boundary problems 10.1515/9783112707159 . Akademie-Verlag, Berlin, 1982

  23. [31]

    R. T. Seeley. Integro-differential operators on vector bundles 10.2307/1994203 . Transactions of the American Mathematical Society , 117:167--204, 1965

  24. [32]

    G. Segal. Fredholm complexes 10.1093/qmath/21.4.385 . The Quarterly Journal of Mathematics. Oxford. Second Series , 21:385--402, 1970

  25. [33]

    K. Sharifi. The gap between unbounded regular operators https://www.jstor.org/stable/24715970. Journal of Operator Theory , 65(2):241--253, 2011

  26. [34]

    Schulze and N

    B.-W. Schulze and N. N. Tarkhanov. Elliptic complexes of pseudodifferential operators on manifolds with edges https://publishup.uni-potsdam.de/opus4-ubp/frontdoor/index/index/year/2008/docId/2321. Preprint, 1998

  27. [35]

    Yu. P. Solovyov and E. V. Troitsky. C^* -algebras and elliptic operators in differential topology 10.1090/mmono/192 , volume 192 of Translations of Mathematical Monographs . American Mathematical Society, Providence, RI, 2001. Translated from the 1996 Russian original by Troitsky

  28. [36]

    E. V. Troitsky. The index of equivariant elliptic operators over C *-algebras 10.1007/BF00140752 . Annals of Global Analysis and Geometry , 5(1):3--22, 1987

  29. [37]

    Vasilescu

    F.-H. Vasilescu. https://www.jstor.org/stable/24713845 Stability of the index of a complex of B anach spaces . Journal of Operator Theory , 2(2):247--275, 1979

  30. [38]

    van den Dungen

    K. van den Dungen. Generalised D irac- S chr \"o dinger operators and the C allias theorem 10.1017/fms.2024.157 . Forum Math. Sigma , 13:e11, 2025

  31. [39]

    Villegas-Villalpando

    B. Villegas-Villalpando. https://bonnus.ulb.uni-bonn.de/permalink/49HBZ_ULB/idtnkp/alma991045524486406467 Fredholm C omplexes of H ilbert C *- M odules . Master's thesis, R heinische F riedrich- W ilhelms- U niversit \"a t B onn, 2024

  32. [40]

    C. Wahl. http://nyjm.albany.edu:8000/j/2009/15_319.html Homological index formulas for elliptic operators over C *-algebras . New York Journal of Mathematics , 15:319--351, 2009

  33. [41]

    C. A. Weibel. An introduction to homological algebra 10.1017/CBO9781139644136 , volume 38 of Cambridge Studies in Advanced Mathematics . Cambridge University Press, Cambridge, 1994

  34. [42]

    R. O. Wells, Jr. Differential analysis on complex manifolds 10.1007/978-0-387-73892-5 , volume 65 of Graduate Texts in Mathematics . Springer, New York, third edition, 2008. With a new appendix by Oscar Garcia-Prada

  35. [43]

    N. E. Wegge-Olsen. K -theory and C *-algebras: A friendly approach 10.1093/oso/9780198596943.001.0001 . Oxford Science Publications. The Clarendon Press, Oxford University Press, New York, 1993

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.