The boundary correspondence under quasiconformal mappings and VMO-Teichmuller space

Liu Tailiang; Shen Yuliang

REVIEW 2 major objections 7 minor 25 references

The boundary correspondence under quasiconformal mappings and VMO-Teichmuller space

T0 review · 2 major / 7 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper proves that a homeomorphism of the real line is strongly vanishing symmetric if and only if it extends to a quasiconformal self-map of the upper half-plane whose complex dilatation induces a strongly vanishing Carleson measure…

desk verdict A solid, novel contribution to the VMO-Teichmüller program that fixes a real conformal-invariance defect; accept after minor revision. read the letter →

arxiv 2411.16042 v1 pith:OEEMEWWV submitted 2024-11-25 math.CV

Liu Tailiang , Shen Yuliang This is my paper · ORCID

classification math.CV MSC 30C6230F6030H35

keywords quasiconformalmappingstronglyvanishingsymmetrichomeomorphismCarlesonmeasureVMO-TeichmüllerspaceCMOBMOconformalweldingCayleytransform

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 introduces a subclass of quasisymmetric homeomorphisms of the real line, the strongly vanishing symmetric homeomorphisms, defined by requiring the logarithmic derivative to lie in the space CMO(R), the closure of compactly supported continuous functions in the BMO norm. The central theorem states that a homeomorphism belongs to this class exactly when it admits a quasiconformal extension to the upper half-plane whose complex dilatation mu makes |mu(z)|^2/y dxdy a strongly vanishing Carleson measure, meaning the Carleson-box ratios vanish as intervals shrink, grow, or drift to infinity. The purpose is to repair the usual VMO-Teichmüller space, which lacks conformal invariance under the Cayley transform; the new class forms a group, and its quotient is compatible with the unit-circle model. If the theorem is right, the obstacle that made earlier VMO-Teichmüller spaces behave badly disappears.

What carries the argument

The load-bearing object is the class CM_s(H) of strongly vanishing Carleson measures on the upper half-plane: a Carleson measure $\lambda$ belongs to it when $\lambda$(Q_I)/|I| tends to 0 simultaneously as |I| tends to 0, as |I| tends to infinity, and as the interval I+x drifts to infinity. Its companion is CMO(R), the closure of compactly supported continuous functions in the BMO norm, characterized in Lemma 3.1 by exactly those three vanishing conditions on mean oscillations. The argument also depends on the conformal invariance of these classes under the Cayley transform, proved in Theorems 3.4 and 3.7, and on a Beurling–Ahlfors-type extension whose complex dilatation is controlled through Littlewood–Paley estimates. In the reverse direction the machinery is conformal welding: the Schwarzian derivative S_g of the lower-half-plane welding map, together with an estimate reducing |N_g|^2 y to |S_g|^2 $y^{3}$, converts strong vanishing of the dilatation measure into log g' in CMOA(L).

What would settle it

Identify an explicit f in VMO(R) \ CMO(R), for example a sum of smooth bumps of height epsilon_k tending to 0 supported on intervals drifting to +infinity, and compute the complex dilatation of the Beurling–Ahlfors-type extension constructed in the proof. Theorem 1.1 predicts that the resulting measure |mu(z)|^2/y dxdy fails the x-to-infinity strong-vanishing condition, so the ratio $\lambda$(Q_{I+x})/|I| for a fixed interval I should stay away from zero as x tends to infinity; computing that ratio for a concrete such h would settle whether the characterization holds.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is a boundary-correspondence theorem: a sense-preserving homeomorphism h:R->R is strongly vanishing symmetric if and only if it can be extended to a quasiconformal mapping of the upper half-plane onto itself whose complex dilatation mu satisfies |mu(z)|^2/y dxdy in CM_s(H), the class of strongly vanishing Carleson measures. A second theorem states that h in SS_0(R) if and only if the conjugate gamma o h o $gamma^{{-1}}$ is strongly symmetric on the unit circle, where gamma(z)=(z-i)/(z+i) is the Cayley transform. The proof combines a variant of the Beurling–Ahlfors extension adapted to strong quasisymmetry for the forward direction with conformal welding and the Schwarzian derivative for the reverse direction. Corollary 4.3 then gives four equivalent criteria for a conformal welding to be strongly vanishing symmetric: membership of the boundary map in SS_0(R), strong vanishing of the dilatation measure, strong vanishing of |S_g|^2 |y|^3 on the lower half-plane, and log g' in CMOA(L). Together these statements establish that T_c(R)=SS_0(R)/~ is a conformally invariant VMO-Teichmüller space with a group structure.

Load-bearing premise

The reverse inclusion in the proof of Theorem 4.2 assumes, as a black box, an earlier theorem that a circle homeomorphism with vanishing Carleson dilatation gives a conformal welding map whose Schwarzian derivative is Carleson-vanishing at the boundary; that theorem is cited, not reproved.

Editorial extensions

If this is right

The boundary class SS_0(R) coincides with the class of homeomorphisms admitting quasiconformal extensions whose dilatation measure is strongly vanishing, so the two definitions are interchangeable.
The class SS_0(R) is a group, so the quotient T_c(R)=SS_0(R)/~ is a genuine Teichmüller-space model on the real line.
The quotient T_c(R) is compatible with the unit-circle VMO-Teichmüller space, since h in SS_0(R) exactly when gamma o h o gamma^{-1} in SS(S^1), removing the non-conformal-invariance obstruction of the usual VMO-Teichmüller space.
Corollary 4.3 provides four equivalent criteria for a conformal welding to be strongly vanishing symmetric, including a Schwarzian-derivative criterion and membership of the logarithmic derivative in CMOA(L).
The paper replaces the usual VMO-Teichmüller space, described as lacking Teichmüller-space properties, with a conformally invariant model.

Reading between the lines

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

Not pursued in the paper: because the three vanishing directions defining CM_s(H) correspond to the three ways a point of the upper half-plane approaches the boundary of the disk under the Cayley transform, one could transplant the definition to other simply connected domains and obtain conformally invariant CMO-type boundary classes whenever the domain has sufficiently regular boundary.
A natural testable conjecture, not stated in the paper, is that SS_0(R) is the closure of compactly supported diffeomorphisms of the line in the quasisymmetric topology, which would give a topological model of the connected component of the identity in the universal Teichmüller space.
The strong vanishing condition should force the Beurling–Ahlfors-type extension to be asymptotically conformal at all three boundary approaches, so one could verify Theorem 1.1 numerically by approximating h with dyadic piecewise-linear maps and checking the three Carleson limits on the computed dilatation.

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.

Desk Editor's Note

read the letter

I think this paper is on solid ground. It introduces a new subclass of the real-line universal Teichmüller space, the strongly vanishing symmetric homeomorphisms SS0(R), and proves the boundary correspondence: h is in SS0(R) iff it has a q.c. extension whose complex dilatation defines a strongly vanishing Carleson measure |μ|²/y dxdy. That gives a CMO-Teichmüller space Tc(R) with the conformal invariance and group structure that the usual VMO-Teichmüller space on R lacks. The main theorems (1.1, 1.2) are genuinely new, and the proofs are mostly self-contained, with real estimates in Section 4.

What the paper does well: it identifies a precise defect in the existing VMO-Teichmüller program—the failure of conformal invariance on R—and fixes it by enlarging the vanishing condition to three regimes (small intervals, large intervals, far-out intervals), which is the CMO condition. The equivalence with the circle model, Theorem 1.2, follows cleanly from the CMs(H) invariance. The authors also honestly note that part of Corollary 4.3 was already in Semmes's thesis.

The soft spots are not fatal. Propositions 3.10 and 3.11 are stated as 'repeating the discussion' from [17] rather than proved; they are plausible but should be spelled out or at least given precise references to the corresponding lemmas in [17]. More importantly, the reverse inclusion in Theorem 4.2 imports [19, Theorem 4.1] as a black box to get the Schwarzian measure condition from the dilatation condition on the circle. That cited result is load-bearing, and the authors do not reprove it. Given that it is published and the hypotheses match, this is acceptable in a research paper, but a referee should check that the application is legitimate. The self-citation is heavy, but it reflects that the authors built this program; I don't see inflated claims.

Net: the central claim holds up as far as I can verify. The paper is for specialists in Teichmüller spaces and BMO/VMO function theory. It deserves a serious referee and, with modest revision, publication. I'd send it to peer review and probably accept.

Referee Report

2 major / 7 minor

Summary. The paper introduces the class SS0(R) of strongly vanishing symmetric homeomorphisms of the real line, defined by the conditions h ∈ SQS(R) and log h′ ∈ CMO(R), together with the class CMs(H) of strongly vanishing Carleson measures on the upper half-plane, defined by three simultaneous vanishing conditions for Carleson squares as |I|→0, |I|→∞, and x→∞. The main result, Theorem 1.1 (restated as Theorem 4.2), asserts that h ∈ SS0(R) if and only if h admits a quasiconformal extension to H whose complex dilatation μ satisfies |μ(z)|²/y dxdy ∈ CMs(H). The proof proceeds in two directions: a forward direction using a variant of the Semmes extension, and a reverse direction using conformal welding, the Cayley transform, and earlier results from the authors' work on VMO-Teichmüller spaces. The paper also establishes conformal-invariance statements for CMO(R) and CMs(H) under the Cayley transform (Theorems 3.4 and 3.7), and uses these to define a CMO-Teichmüller space Tc(R) that is claimed to be conformally invariant and compatible with the unit-circle VMO-Teichmüller space (Theorem 1.2).

Significance. If the central theorem is correct, the paper settles a natural open problem identified in the recent VMO-Teichmüller space literature: the usual real-line VMO-Teichmüller space is not conformally invariant and does not have a group structure, while the new CMO-Teichmüller space, built on strongly vanishing symmetric homeomorphisms, is designed to repair both defects. The introduction of CMs(H) as a three-parameter vanishing Carleson measure class is a useful contribution, and the paper provides concrete characterizations of the new class through conformal welding and Schwarzian derivative conditions. The main proof is largely self-contained, with explicit estimates in Section 4 and a clean reduction to the unit-circle model via the Cayley transform. The paper does not include machine-checked proofs or code, but the analytical arguments are detailed and the main theorem gives falsifiable conditions (existence of a q.c. extension with a strongly vanishing Carleson measure) that are checkable in principle.

major comments (2)

[§4, estimates (4.9)–(4.13)] The proof of (4.2) contains a gap in constant tracking. In (4.9), the inequality labeled ≲ replaces the factor |yN_g(z)|^4 + |yN_{γ_ζ}(z)|^2 |yN_g(z)|^2 by a constant, but that constant depends on sup_{z∈K_{n1}} y|N_g(z)| and on n1 (through the bound on |yN_{γ_ζ}| for ζ outside K_{n0}). The subsequent explicit estimates (4.10)–(4.13) bound only the simplified integral η/[y((x−ξ)^2+(y+η)^2)] and show it is O(1/n1) or O(ε). Without tracking the implicit constant from (4.9), the conclusion J_{32}(ζ)<ε does not formally follow from the displayed inequalities, and the issue is load-bearing because (4.2) is the core of the reverse inclusion in Theorem 4.2. A simple repair is to argue by compactness: after fixing n1 so that the other terms in (4.6) are small, the integrand in J_{32} is a bounded function on K_{n1} times |γ'_ζ(z)|, and ∫_{K_{n1}} |γ'_ζ(z)| dm(z) → 0 as ζ→∞ in the three required regimes; one can then choose K_{n0} depending on ε, g, and n1. The manuscript should either implement this or track the constants explicitly.
[§3, Propositions 3.10 and 3.11] Propositions 3.10 and 3.11 are stated without proof, with the note that they follow by 'repeating the discussion' from Lemmas 7.1, 7.3 and Proposition 7.4 of [17]. These propositions are used in essential ways: Proposition 3.10 is needed to deduce S_g ∈ A_0(L) from |S_g|^2|y|^3 ∈ CMs(L) and to prove Corollary 3.12, and Proposition 3.11 is used in Corollary 4.3. Since the class CMs(H) involves three limiting conditions and is not verbatim the CM or CM0 case treated in [17], the manuscript should provide complete proofs or a precise statement of how each of the three vanishing regimes follows from the corresponding argument in [17]. As written, the central chain of implications in Theorem 4.2 is missing an explicitly justified technical input.

minor comments (7)

[§2, definition of f_y] In the proof of SS0(R) ⊆ ~SS0(R), the definition f_y(x) = y^{-1} f(y/x) should read f_y(x) = y^{-1} f(x/y), as used in the convolutions α_y * e^a immediately afterward.
[§3, proof of Theorem 3.4] The set equality {z : |γ(z)| → 1} = {z : Im z → 0} ∪ {z : Im z → ∞} ∪ {z : Re z → ∞} is missing the region Re z → −∞; the correct set also includes Re z → −∞ with Im z bounded, and the proof should mention that this additional region is handled symmetrically.
[§3, proof of Theorem 3.8] The displayed identity Im α(w) |α'(w)| = |α'(w)|^2 (1−|w|^2) is false for the Cayley map α(w)=i(1+w)/(1−w); the correct identity is Im α(w) = (1−|w|^2)|α'(w)|/2. The equivalence of Carleson measure conditions is unaffected by the constant factor, but the displayed equality should be corrected.
[§1, Theorem 1.2] Theorem 1.2 is stated in the introduction but never proved explicitly; it follows from Theorem 4.2 and Theorem 3.7, but the authors should add a one-sentence proof or a reference to the precise location where this implication is established.
[§4, use of [19, Theorem 4.1]] The reverse inclusion in Theorem 4.2 relies on [19, Theorem 4.1] to pass from the vanishing Carleson condition on the dilatation to the Schwarzian measure condition on D*. This is a published result and the hypotheses appear to match, so I do not regard the reliance as circular; however, the theorem is load-bearing and should be stated explicitly in Section 4 for self-containedness.
[§3, Lemma 3.1] The equivalence (2) ⇔ (3) in Lemma 3.1 is not fully justified: (3) ⇒ (2) follows by Hölder, but (2) ⇒ (3) is omitted. This is standard via the John–Nirenberg inequality for functions with vanishing mean oscillation, but the authors should either prove it or give a precise reference.
[Throughout] There are numerous typographical errors, including 'A /greaterorsimilarB' in the notation section, 'Propostion', 'extened', 'argurments', and 'qusisymmetric'; a careful proofreading is needed.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the CMO-Teichmüller boundary correspondence is proved by a self-contained construction plus legitimate use of an external circle-model theorem.

full rationale

I walked the derivation chain of Theorem 1.1/4.2. The forward inclusion SS0(R) ⊆ ~SS0(R) is a self-contained Semmes-type Beurling-Ahlfors extension with explicit estimates, and it does not invoke the target theorem. The reverse inclusion imports [19, Theorem 4.1] to pass from the unit-circle dilatation condition |μ_~ρ|²/(1−|w|²) ∈ CM0(D) to the Schwarzian condition |S_{gμ∘γ^{-1}}|²(|w|²−1)³ ∈ CM0(D*); this is a published theorem about the unit-circle VMO-Teichmüller space, not an equivalent restatement of the real-line CMO characterization being proved. The remaining estimates (4.2)–(4.13) reduce the problem to that cited input, after which the pull-back property of CMO gives log h′ ∈ CMO(R). The definitions of SS0(R) and CMs(H) are not defined in terms of the conclusion; Theorem 3.7 is a genuine conformal-invariance equivalence proved from earlier characterizations. The closing note about Semmes' thesis is an acknowledgment of priority, not evidence of circularity. No fitted parameter is renamed as a prediction, and no uniqueness claim is imported from the authors' own prior work to force the conclusion. The only load-bearing external input, [19, Theorem 4.1], is parameter-free and concerns unit-circle SS(S1), a hypothesis that does not include the target real-line CMO result; its use is thus legitimate independent support rather than circularity.

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

The paper introduces no numerical fitting, and the mathematical objects (SS0, CMs, CMO-Teichmüller space) are defined constructs, not empirical entities. The load is carried by cited theorems from the authors' and others' prior work, plus standard harmonic analysis facts.

assumptions (6)

standard math Lemma 3.1: f ∈ BMO(R) lies in CMO(R) iff the three mean-oscillation limits vanish (Neri/Uchiyama).
Imported from literature; not proven here. Supports Proposition 3.2 and conformal invariance of CMO(R).
standard math Fefferman-Stein: f ∈ BMO(VMO) on R iff |∇P_f(z)|² y dm(z) is a (vanishing) Carleson measure.
Used in Theorem 3.8 to link CMO(R) to strongly vanishing Carleson measures.
standard math Semmes: h ∈ SQS(R) iff h extends to a q.c. map whose dilatation satisfies |μ|²/y ∈ CM(H).
Starting point for SS0 definition and extension construction in Section 4.
standard math SS(S1) is a group.
Used in Lemma 4.1 to show ~SS0(R) is a group; from Astala-Zinsmeister and Shen-Wei.
domain assumption [19, Theorem 4.1]: for VMO-Teichmüller welding on the circle, |S_{g_μ∘γ^{-1}}(w)|²(|w|²-1)³ ∈ CM0(D*).
Imported from authors' earlier paper; load-bearing for reverse inclusion of Theorem 4.2.
standard math Pull-back P_h preserves CMO(R) for h ∈ SQS(R).
Stated via density argument; used to prove SS0(R) is a group and the final log h' ∈ CMO(R).

how reviews work

0 comments

Cite this review

Pith. "Pith review of The boundary correspondence under quasiconformal mappings and VMO-Teichmuller space." pith.science (2026). https://pith.science/paper/OEEMEWWV

@misc{pith2026241116042,
  author       = {Pith},
  title        = {Pith review of: The boundary correspondence under quasiconformal mappings and VMO-Teichmuller space},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OEEMEWWV}},
  note         = {Machine review of arXiv:2411.16042}
}

read the original abstract

In this paper, we introduce a class of vanishing Carleson measures with conformal invariance and corresponding strongly vanishing symmetric homeomorphisms on the real line and prove that they can be mutually generated under quasiconformal mappings. This is motivated by constructing a nice VMO-Teichmuller space on the real line, which completely removes the obstacle of the usual VMO-Teichmuller space that lacks conformal invariance and is repeatedly encountered in the papers [17,22-24].

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 25 canonical work pages

[17]

Shen, Y., VMO-Teichm¨ uller space on the real line, Ann. Fenn. Math. , 2022, 47(1), 57–82

work page 2022
[1]

Ann., 1991, 289, 613–625

Astala, K., Zinsmeister, M., Teichm¨ uller spaces and BMOA, Math. Ann., 1991, 289, 613–625

work page 1991
[2]

Beurling, A., Ahlfors, L.V., The boundary correspondence under quasiconformal mappings, Acta Math., 1956, 96, 125–142

work page 1956
[3]

Bourdaud, G., Lanza de Cristoforis, M., Sickel, W., Functional calc ulus on BMO and related spaces, J. Funct. Anal. , 2002, 189(2), 515–538

work page 2002
[4]

Bourdaud G., Remarques sur certains sous-espaces de BM O(Rn) et de bmo(Rn), Ann. Inst. Fourier, 2002, 52(4), 1187–1218

work page 2002
[5]

Analy se Math

Carleson, L., On mappings, conformal at the boundary, J. Analy se Math. 1967, 52(4), 1–13

work page 1967
[6]

, 51, 241–250, 1974

Coifman, R., Feﬀerman, C., Weighted norm inequalities for maximal f unctions and singular integrals, Studia Math. , 51, 241–250, 1974

work page 1974
[7]

Coifman, R., Weiss, G., Extensions of Hardy spaces and their use in analysis, Bull. Amer. Math. Soc. , 1977, 83, 569–645

work page 1977

Show all 25 references

[8]

J., The theory of weights and the Dirichlet problems for elliptic equations, Ann

Feﬀerman, R., Kenig, C., Pipher. J., The theory of weights and the Dirichlet problems for elliptic equations, Ann. of Math. (2), 1991, 134, 65–124

1991
[9]

Gardiner, F.P., Sullivan, D., Symmetric structures on a closed curv e, Amer. J. Math. , 114, 1992, 683–736

1992
[10]

Garnett, J.B., Bounded Analytic Functions, Academic Press, Ne w York, 1981. 22

1981
[11]

Girela, D., Analytic functions of bounded mean oscillation, Complex function spaces , 2001, 4, 61–170

2001
[12]

Jones, P.W., Homeomorphisms of the line which preserve BMO, Ark. Mat. , 1983, 21, 229– 231

1983
[13]

Liu, T., The Cauchy integral on chord-arc curves and the relat ed research, Thesis (Ph.D.)- Soochow University, 2023

2023
[14]

, 1975, 53, 175– 189

Neri, U., Fractional integration on the space H 1 and its dual, Studia Math. , 1975, 53, 175– 189

1975
[15]

Louis, 1983, 1–133

Semmes, S., The Cauchy integral and related operators on smo oth curves, Thesis (Ph.D.)- Washington University in St. Louis, 1983, 1–133

1983
[16]

Semmes, S., Quasiconformal mappings and chord-arc curves, Tran. Amer. Math. Soc. , 306, 233–263, 1988

1988
[18]

Shen, Y., Tang, S., Wu, L., Weil-Petersson and little Teichm¨ uller sp aces on the real line. Ann. Fenn. Math. , 2018, 43(2), 935–943

2018
[19]

Math., 2013, 234, 129–148

Shen, Y., Wei, H., Universal Teichm¨ uller space and BMO, Adv. Math., 2013, 234, 129–148

2013
[20]

M., Singular Integrals and Diﬀerentiability Properties of functions, Princeton Univ

Stein, E. M., Singular Integrals and Diﬀerentiability Properties of functions, Princeton Univ. Press, Princeton, N.J., 1970

1970
[21]

Uchiyama, A., On the compactness of operators of Hankel typ e, Tohoku Math. J. , 1978, 30, 163–171

1978
[22]

Wei, H., Matsuzaki, K., Symmetric and strongly symmetric homeom orphisms on the real line with non-symmetric inversion, Anal. Math. Phys. , 2021, 11(2), 1–17

2021
[23]

Wei, H., Matsuzaki, K., Beurling-Ahlfors extension by heat kerne l, A∞ weights for VMO, and vanishing Carleson measures, Bull. Lond. Math. Soc. , 2021, 53(3), 723–739

2021
[24]

Wei, H., Matsuzaki, K., Strongly symmetric homeomorphisms on th e real line with uniform continuity, Indiana Univ. Math. J. , 2023, 72, 1553-1576

2023
[25]

Zhu, K., Operator Theory in Function Spaces, Amer. Math. Soc . Press, Providence, 2007. 23

2007

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.

[1] [17]

Shen, Y., VMO-Teichm¨ uller space on the real line, Ann. Fenn. Math. , 2022, 47(1), 57–82

work page 2022

[2] [1]

Ann., 1991, 289, 613–625

Astala, K., Zinsmeister, M., Teichm¨ uller spaces and BMOA, Math. Ann., 1991, 289, 613–625

work page 1991

[3] [2]

Beurling, A., Ahlfors, L.V., The boundary correspondence under quasiconformal mappings, Acta Math., 1956, 96, 125–142

work page 1956

[4] [3]

Bourdaud, G., Lanza de Cristoforis, M., Sickel, W., Functional calc ulus on BMO and related spaces, J. Funct. Anal. , 2002, 189(2), 515–538

work page 2002

[5] [4]

Bourdaud G., Remarques sur certains sous-espaces de BM O(Rn) et de bmo(Rn), Ann. Inst. Fourier, 2002, 52(4), 1187–1218

work page 2002

[6] [5]

Analy se Math

Carleson, L., On mappings, conformal at the boundary, J. Analy se Math. 1967, 52(4), 1–13

work page 1967

[7] [6]

, 51, 241–250, 1974

Coifman, R., Feﬀerman, C., Weighted norm inequalities for maximal f unctions and singular integrals, Studia Math. , 51, 241–250, 1974

work page 1974

[8] [7]

Coifman, R., Weiss, G., Extensions of Hardy spaces and their use in analysis, Bull. Amer. Math. Soc. , 1977, 83, 569–645

work page 1977