Large-size expansion for triangular Wilson loops in confining gauge theories

P.V. Pobylitsa

REVIEW 3 major objections 3 minor 50 references

Large-size expansion for triangular Wilson loops in confining gauge theories

T0 review · 3 major / 3 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The L-2 term in the large-size expansion of triangular Wilson loops is fixed by the three interior angles, the area, and the space-time dimension.

desk verdict A genuine new triangle result for f_2(C) in EST, but the analytic regularization's uniqueness is asserted rather than proven, so the final formula is a strong prediction with a known gap. read the letter →

arxiv 1908.01724 v2 pith:L3M3R5NB submitted 2019-08-05 hep-th hep-lat

P.V. Pobylitsa This is my paper · ORCID

classification hep-thhep-lat

keywords Wilsonloopseffectivestringtheorylarge-sizeexpansionSchwarz-Christoffelmappinganalyticalregularizationconfininggaugetheoriesarealawtwo-loopcorrection

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 computes the $L^{-2}$ correction in the large-size expansion of the logarithm of a Wilson loop for triangular contours in confining gauge theories, within effective string theory (EST). The claim is that this two-loop correction, $f_2(C)$, carries no new free parameters: for a triangle with interior angles $\theta_1,\theta_2,\theta_3$ and area $S$ it equals $(1/\sigma S)$ times the product $\prod_{k=1}^3(\pi/\theta_k-1)$, with a further factor of order $D-2$. The computation uses a new analytical regularization of the ultraviolet-divergent two-loop EST integral, built from the Schwarz--Christoffel mapping of the polygon. If correct, the result is a parameter-free prediction that lattice simulations of triangular Wilson loops can test, extending the previously known rectangle formula to the simplest non-rectangular contour.

What carries the argument

The machinery is an analytical regularization of divergent integrals of the type $$$M_P^{{(n)}}$(\{\gamma_k\},\{z_k\})=\int_{\mathbb C_+} $d^{2}$z\,|\operatorname{Im} z|^{\gamma_0}\prod_{k=1}^n |z-z_k|^{\gamma_k} P(z,z^*),$$ defined by analytic continuation from the region of converging exponents (equivalently, through the functions $\Pi_P^{(n)}$). These integrals carry the ultraviolet divergences of the two-loop correction after the first renormalization step, which subtracts the short-distance singularity of the Dirichlet Green function and replaces the bare string tension by the physical $\sigma$. The Schwarz--Christoffel mapping encodes the polygon's geometry; for a triangle the exponents are determined by $\beta_k=1-\theta_k/\pi$, and the integrals reduce to Euler $\beta$ and gamma functions, producing the product $\prod_{k=1}^3(\pi/\theta_k-1)$.

What would settle it

Measure the $\lambda^{-2}$ coefficient for triangular Wilson loops on the lattice in a confining gauge theory for several triangle shapes and compare the full angle dependence with formula (1.19); a mismatch in the product $\prod_k(\pi/\theta_k-1)$ would show the analytical regularization is not the physical one. A cheaper check is to renormalize the same SC integrals with a different regulator, such as dimensional regularization, and see whether eq. (1.19) is reproduced or replaced by a scheme-dependent value.

Watch

Extended reading notes

Core claim

The central result is formula (1.19): in $D$ space-time dimensions, with string tension $\sigma$, the coefficient $f_2(C)$ of the $\lambda^{-2}$ term in $\ln W(\lambda C)$ for a triangular contour with interior angles $\theta_1,\theta_2,\theta_3$ and area $S$ is $$f_2(C)=\frac{1}{\$\sigma$ S}\frac{D-2}{192}\left[\prod_{k=1}^{3}\left(\frac{\pi}{\theta_k}-1\right)\right]\left[1-\frac{D-2}{24}\prod_{k=1}^{3}\left(\frac{\pi}{\theta_k}-1\right)\right].$$ The author derives this by identifying the two-loop EST correction with a renormalized figure-eight Feynman diagram, rewriting it through the Schwarz--Christoffel map of the triangle, and applying analytical continuation to the divergent integrals over the upper half-plane. The final expression is symmetric in the three angles and regular for all physical triangles, which the paper takes as a consistency check of its renormalization scheme.

Load-bearing premise

The load-bearing premise is that the analytical continuation of the divergent Schwarz--Christoffel integrals is unique and reproduces the physical renormalization of the underlying gauge theory, a proof the paper postpones for arbitrary polygons; the triangle calculation in appendix H is explicitly formal.

Editorial extensions

If this is right

For triangular contours the $L^{-2}$ term is fully predicted once $\sigma$ and $D$ are known, so no additional effective-string coupling enters at this order.
The two-step renormalization, short-distance subtraction followed by analytical continuation of Schwarz--Christoffel integrals, gives a template for computing $f_2(C)$ for arbitrary polygonal contours.
Ratios of Wilson loops such as (1.13) now have an explicit $\lambda^{-2}$ prediction for triangle-based loop sets, making the term directly accessible to lattice Monte Carlo.
The symmetry of the final formula under permutations of the angles supports the paper's claim that the renormalized result is compatible with the SL(2,R) invariance of the SC parametrization.

Reading between the lines

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

Editorial inference: the factorization of $f_2$ into a product over vertices suggests that for general polygons the $L^{-2}$ coefficient may factor over vertices into functions of the interior angles, with the rectangle formula as a limiting case.
Editorial inference: because formula (1.19) stays regular for obtuse triangles, it makes a controlled prediction for re-entrant contours, where a naive perimeter expansion would be less trustworthy.
Editorial inference: a lattice measurement of the triangle $f_2$ would probe the assumption that the Nambu action dominates the effective string action up to order $L^{-2}$; a mismatch would signal that boundary terms contribute earlier than the paper assumes.

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.

Desk Editor's Note

read the letter

Dear colleague,

The one thing to know: this paper gives the first analytic expression for the 1/L^2 term of a triangular Wilson loop in effective string theory, but the calculation leans on an analytic regularization whose uniqueness is asserted, not proven, and the paper says so itself. That makes the result a well-motivated prediction, not an ironclad one.

The new content is real. The rectangle case was the only polygon where f_2(C) was known; the triangle case is new, and the final expression (1.19) depends on the three angles and area in a nontrivial way. The method is also a step forward: the Schwarz-Christoffel based analytic regularization is designed for arbitrary polygons, and the two-step renormalization programme (interior divergences first, then boundary and cusp divergences) is a reasonable framework. The derivation is long but careful; appendices G and H contain explicit integrals, and the permutation symmetry of the final expression is a useful internal check. The paper is also honest: section 6.2 explicitly postpones the general proof of path-independence and regularity to a separate work, and appendix H warns that it performs a formal calculation ignoring divergences of intermediate integrals.

The soft spot is exactly there. Equations (6.10)-(6.11) define the renormalized integrals by analytic continuation of M_P^{(n)}, and everything downstream, including (1.19), inherits the assumption that this continuation is unique and matches the physical renormalization of the EST/MGT. For triangles (n=2) the integrals are computed in closed form, so the continuation is explicit for the case actually used; that mitigates the concern. But the general statement in section 6.2 that the continuation is path-independent is not demonstrated, and the step (H.5), which passes from the full plane to the upper half-plane after symmetrization, is a place where a boundary term could hide. If the scheme is not unique, the numeric value in (1.19) is a regularization artifact. I don't think the paper proves otherwise; I also don't think the concern is fatal, because the author has shown enough explicit structure that a proof for n=2 is plausibly within reach.

The citation pattern looks solid: the known rectangle results, the Laplace determinant work of Aurell-Salomonson, and the EST lore are all used appropriately, with no fitted parameters or invented entities.

This paper is for EST theorists and lattice practitioners testing effective string predictions. It deserves a serious referee even with the open uniqueness point. If the referee can push the author to present a dedicated proof of path-independence for the n=2 case, or at least a compelling argument that any admissible continuation gives the same value, the paper would be much stronger. As it stands, treat (1.19) as a plausible prediction with a known gap, not a settled theorem.

My recommendation: send it out for peer review, with a referee instructed to focus on the regularization uniqueness. I'd cite it in my own work, with a caveat.

Referee Report

3 major / 3 minor

Summary. The paper computes the coefficient f2(C) of the 1/L^2 term in the large-size expansion of the logarithm of a Wilson loop for triangular contours in confining gauge theories, using effective string theory (EST). Starting from the two-loop correction to the Nambu action, the author reduces the problem to integrals of the finite part K_{\mu\nu} of the Green-function derivative, expresses K_{\mu\nu} through the Schwarz-Christoffel (SC) mapping of the polygon, and proposes a two-step renormalization: a local subtraction of the short-distance singularity followed by an analytic regularization of the remaining divergent SC integrals via meromorphic functions M_P^{(n)} / \Pi_P^{(n)}. For triangles the relevant n=2 integrals are computed explicitly in appendices G and H, leading to the closed formula (1.19), which depends only on the string tension, the triangle area, and the interior angles. The paper also reviews the one-loop terms f_ln(C) and f_0(C), including Laplace-determinant results and balance conditions for renormalization-invariant combinations of Wilson loops.

Significance. If correct, Eq. (1.19) is a parameter-free EST prediction that can be tested by lattice Monte Carlo simulations, and the proposed SC-based analytic regularization is a substantial technical step beyond the rectangle-specific calculations of refs. [26,27]. The derivation is detailed, and the explicit computations in appendices G and H, together with the final permutation symmetry of the triangle result, are valuable internal consistency checks. However, the scheme-independence of the analytic continuation, which is the load-bearing premise behind the renormalized integrals I_1^ren and I_2^ren, is not fully proved in the manuscript; the paper explicitly postpones the general proof and appendix H is announced as a formal calculation. The significance is therefore conditional on closing that gap.

major comments (3)

[Sec. 6.2 and App. H] The definition of I_1^ren and I_2^ren in Eqs. (6.10)-(6.11) relies on the analytic continuation of M_P^{(n)} being path-independent and regular at the physical values of the exponents. Section 6.2 explicitly postpones the proof for arbitrary n, and for the n=2 case relevant here it only states that the properties 'can be seen' from the explicit formulas, without providing the required uniqueness argument. Moreover, Appendix H derives \Pi_P^{(2)} by a formal calculation that 'ignores divergences of some intermediate integrals', in particular Eq. (H.5) identifies a full-plane integral with twice the upper-half-plane integral for objects that are divergent before regularization. If this step discards a boundary or surface term, or if a different admissible continuation reaches a different value, then Eqs. (7.6), (7.23), and hence the central result Eq. (1.19), would be regularization artifacts rather than physical EST predictions. A self-contained proof of uniqueness and regularity of the continuation for the n=2 functions actually used should be supplied.
[Secs. 5.4 and 7.4] The renormalization prescription is formulated after sending one SC vertex to infinity, so the value obtained could in principle depend on which vertex is chosen. Section 5.4 states that the result must be independent of this choice, and Section 7.4 notes only that the final expression is symmetric and that a general proof of SL(2,R) covariance is postponed to future work. For the triangular claim this is directly checkable: the author should verify that I_1^ren and I_2^ren computed with each of the three vertices at infinity give the same f_2(C), or prove the SL(2,R) covariance of the analytic continuation. The permutation symmetry of the final formula is a necessary but not sufficient check.
[Secs. 3.3 and 6.3] The two-step renormalization replacing G_{\mu\nu} by K_{\mu\nu} and \sigma_0 by \sigma, followed by analytic continuation to physical exponents, is assumed to reproduce the physical renormalization of the underlying gauge theory that leads to Eq. (2.22). The manuscript does not prove or even explicitly state this identification as an assumption; it only asserts that the divergent quantities are renormalized. Since the final prediction depends on this matching, the author should either justify that the combined subtraction plus analytic continuation is equivalent to the standard EST/MGT renormalization, or at minimum state this as an explicit assumption and discuss the possible finite counterterms that could alter Eq. (1.19).

minor comments (3)

[Eq. (G.7)] In the displayed expression for \Pi_1^{(2)} the prefactor contains (z_2)^{\gamma_1+\gamma_2-1-\alpha}, whereas the general formula (G.9) and the translation to arbitrary z_1,z_2 in (G.10) use |z_2|; please clarify that the intermediate formula assumes z_2>0 (as is allowed by the chosen normalization) or replace the non-invariant factor by |z_2|.
[References] Reference [31] contains a typographical artifact ('hep-th/10 08.1178' with an inserted space) and should be corrected to arXiv:1008.1178.
[Eq. (4.19)] The prefactor in Eq. (4.19), written in the extracted text as '- 2 \sigma (D-2)', is easy to misread as -2/[\sigma(D-2)]. From the subsequent reduction to Eq. (4.26) one infers that the intended factor is -2(D-2)/\sigma; please set the fraction in an unambiguous form.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the f2(C) prediction is a parameter-free EST computation; the postponed analytic-continuation proof is a scheme-dependence risk, not a circular reduction.

full rationale

The derivation chain is self-contained: expand the Nambu action to quartic order, perform the Gaussian/Wick contraction to the figure-eight diagram, apply the two-step renormalization, pass to the Schwarz-Christoffel representation, define the analytically continued integrals M_P^(n), evaluate them for triangles, and obtain eq. (1.19). No fitted parameter enters the final formula; the inputs are the string tension sigma, the spacetime dimension D, and the geometric data of the triangle, which appear with the correct scaling by dimensional analysis. In particular, I1^ren and I2^ren are computed directly from the SC mapping and the analytic continuation, not imposed to match the triangular Wilson loop result. The paper explicitly postpones a rigorous proof of path-independence and regularity of the analytic continuation for general polygons (Sec. 6.2) and notes that Appendix H uses formal manipulations that ignore intermediate divergences; this is an honest scheme-dependence or correctness risk, but it is not circularity because the paper does not define f2(C) to be the regularized expression by construction. The self-citation [38] is only a background lattice review and is not load-bearing, and the external result [39] is used only for the 1-loop Laplace-determinant appendix. The permutation symmetry of I1^ren and I2^ren is presented as an internal consistency check, not as an input. No step reduces, by construction or by self-citation, to its own inputs.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The computation relies on the EST framework, the Nambu action, and a new analytical regularization whose uniqueness for general polygons is asserted but not proven. No free parameters are fitted; inputs are the string tension, dimension, and triangle angles.

assumptions (3)

ad hoc to paper The analytical continuation of divergent integrals M_P^(n) to physical parameter values yields a unique, finite renormalized value.
Section 6.2 states this as an assertion for general n, with proof postponed to a separate work; used to define I_ren_1 and I_ren_2.
domain assumption The point-splitting subtraction in eq. (3.9) correctly renormalizes the string tension for arbitrary polygons.
Standard EST renormalization assumption; not derived in the paper.
domain assumption Effective string theory and the Nambu action approximation are valid for computing the L^-2 term of the large-size Wilson loop expansion.
Basic premise of the EST program, stated in the introduction and section 2.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Large-size expansion for triangular Wilson loops in confining gauge theories." pith.science (2026). https://pith.science/paper/L3M3R5NB

@misc{pith2026190801724,
  author       = {Pith},
  title        = {Pith review of: Large-size expansion for triangular Wilson loops in confining gauge theories},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L3M3R5NB}},
  note         = {Machine review of arXiv:1908.01724}
}

abstract

The asymptotic behavior of Wilson loops in the large-size limit ($L\rightarrow\infty$) in confining gauge theories with area law is controlled by effective string theory (EST). The $L^{-2}$ term of the large-size expansion for the logarithm of Wilson loop appears within EST as a two-loop correction. Ultraviolet divergences of this two-loop correction for polygonal contours can be renormalized using an analytical regularization constructed in terms of Schwarz-Christoffel mapping. In the case of triangular Wilson loops this method leads to a simple final expression for the $L^{-2}$ term.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 41 canonical work pages

[1]

Wilson, Phys

K. Wilson, Phys. Rev. D 10 (1974) 2445

work page 1974
[2]

A. M. Polyakov, Nucl. Phys. B 164 (1980) 171. 49

work page 1980
[3]

’t Hooft, Nucl

G. ’t Hooft, Nucl. Phys. B72 (1974) 461

work page 1974
[4]

Veneziano, Nucl

G. Veneziano, Nucl. Phys. B 117 (1976) 519

work page 1976
[5]

Yu. M. Makeenko and A. A. Migdal, Nucl. Phys. B 188 (1981) 269

work page 1981
[6]

A. A. Migdal, Phys. Rept. 102 (1984) 199

work page 1984
[7]

Alvarez, Phys

O. Alvarez, Phys. Rev. D24 (1981) 440

work page 1981
[8]

J. F. Arvis, Phys. Lett. B 127 (1983) 106

work page 1983

Show all 50 references

[9]

Ambjørn and Y

J. Ambjørn and Y. Makeenko, Phys. Lett. B756 (2016) 142, arXiv:1601.00540 [hep-th]

2016 arXiv
[10]

Makeenko, arXiv:1206.0922 [hep-th]

Yu. Makeenko, arXiv:1206.0922 [hep-th]

arXiv
[11]

Yu. M. Makeenko and P. Olesen, Phys. Rev. D 80 (2009) 026002 , arXiv:0903.4114 [hep-th]

2009 arXiv
[12]

Yu. M. Makeenko, Phys. Part. Nucl. 45 (2014) 771, arXiv:1208 .1209 [hep- th]

2014
[13]

L¨ uscher, K

M. L¨ uscher, K. Symanzik and P. Weisz, Nucl. Phys. B 173 (1980 ) 365

1980
[14]

L¨ uscher, Nucl

M. L¨ uscher, Nucl. Phys. B 180 (1981) 317

1981
[15]

L¨ uscher, G

M. L¨ uscher, G. M¨ unster and P. Weisz, Nucl. Phys. B 180 (198 1) 1
[16]

L¨ uscher and P

M. L¨ uscher and P. Weisz, JHEP 0407 (2004) 014, arXiv:hep-th /0406205

2004
[17]

H. B. Meyer, JHEP 05 (2006) 066, arXiv:hep-th/0602281

2006 arXiv
[18]

Aharony and E

O. Aharony and E. Karzbrun, JHEP 0906 (2009) 012, arXiv:090 3.1927 [hep-th]

2009
[19]

Aharony and M

O. Aharony and M. Field, JHEP 1101 (2011) 065, arXiv:1008.2636 [hep-th]

2011 arXiv
[20]

Aharony and N

O. Aharony and N. Klinghoﬀer, JHEP 1012 (2010) 058, arXiv:100 8.2648 [hep-th]

2010
[21]

Aharony and M

O. Aharony and M. Dodelson, JHEP 1202 (2012) 008, arXiv:1111 .5758 [hep-th]

2012
[22]

Aharony, M

O. Aharony, M. Field and N. Klinghoﬀer, JHEP 1204 (2012) 048, arXiv:1111.5757 [hep-th]

2012 arXiv
[23]

Aharony and Z

O. Aharony and Z. Komargodski, JHEP 1305 (2013) 118, arXiv:1 302.6257 [hep-th]

2013
[24]

Bill´ o, M

M. Bill´ o, M. Caselle, F. Gliozzi, M. Meineri and R. Pellegrini, JHEP 05 (2012) 130, arXiv:1202.1984 [hep-th]. 50

2012 arXiv
[25]

Gliozzi and M

F. Gliozzi and M. Meineri, JHEP 1208 (2012) 056, arXiv:1207.2912 [hep- th]

2012 arXiv
[26]

Filk, Regularization procedure for string functionals, prepr int BONN- HE-81-16 (Bonn U.), Sep 1981

T. Filk, Regularization procedure for string functionals, prepr int BONN- HE-81-16 (Bonn U.), Sep 1981

1981
[27]

Dietz and T

K. Dietz and T. Filk, Phys. Rev. D 27 (1983) 2944

1983
[28]

Bill´ o, M

M. Bill´ o, M. Caselle, V. Verduci and M. Zago, PoS LATTICE2010 ( 2010) 273, arXiv:1012.3935 [hep-lat]

2010 arXiv
[29]

Bill´ o, M

M. Bill´ o, M. Caselle and R. Pellegrini, JHEP 1201 (2012) 104, arXiv:1107.4356 [hep-th]

2012 arXiv
[30]

Gliozzi, M

F. Gliozzi, M. Pepe and U.-J. Wiese, JHEP 1011 (2010) 053, arXiv:1006.2252 [hep-lat]

2010 arXiv
[31]

H. B. Meyer, Phys. Rev. D 82 (2010) 106001, arXiv:hep-th/10 08.1178

2010
[32]

B. B. Brandt, M. Meineri, Int. J. Mod. Phys. A31 (2016) 16430 01, arXiv:1603.06969 [hep-th]

2016 arXiv
[33]

Athenodorou, B

A. Athenodorou, B. Bringoltz and M. Teper, JHEP 1102 (2011) 030, arXiv:1007.4720 [hep-lat]

2011 arXiv
[34]

Athenodorou, B

A. Athenodorou, B. Bringoltz and M. Teper, JHEP 1105 (2011) 042, arXiv:1103.5854 [hep-lat]

2011 arXiv
[35]

Athenodorou and M

A. Athenodorou and M. Teper, JHEP 1306 (2013) 053, arXiv:13 03.5946 [hep-lat]

2013
[36]

Athenodorou and M

A. Athenodorou and M. Teper, JHEP 1610 (2016) 093, arXiv:16 02.07634 [hep-lat]

2016
[37]

Athenodorou and M

A. Athenodorou and M. Teper, JHEP 1702 (2017) 015 , arXiv:16 09.03873 [hep-lat]

2017
[38]

P. V. Pobylitsa, arXiv:1609.05869 [hep-lat]

arXiv
[39]

Aurell and P

E. Aurell and P. Salomonson, Commun. Math. Phys. 165 (1994) 233

1994
[40]

Pleijel, Arkiv f¨ or Matematik, 2 (1954) 553

A. Pleijel, Arkiv f¨ or Matematik, 2 (1954) 553

1954
[41]

Kac, Amer

M. Kac, Amer. Math. Monthly 73 (1966) 1

1966
[42]

Stewartson and R

K. Stewartson and R. T. Waechter, Proc. Cambridge Phil. Soc. 69 (1971) 353

1971
[43]

Balian and C

R. Balian and C. Bloch, Ann. of Phys. 64 (1971) 271

1971
[44]

P. B. Guilkey, The index theorem and the heat equation (Publish o r Perish, Boston, 1974). 51

1974
[45]

H. P. McKean and I. M. Singer, J. Diﬀ. Geometry 1 (1967) 43

1967
[46]

A. M. Polyakov, Phys. Lett. B 103 (1981) 207

1981
[47]

Durhuus, H

B. Durhuus, H. B. Nielsen, P. Olesen and J. L. Petersen, Nucl. P hys. B 196 (1982) 498

1982
[48]

Durhuus, P

B. Durhuus, P. Olesen and J. L. Petersen, Nucl. Phys. B 201 (1 982) 176
[49]

Alvarez, Nucl

O. Alvarez, Nucl. Phys. B 216 (1983) 125

1983
[50]

Belavin and V

A. Belavin and V. Belavin, JHEP 1002:010, 2010, arXive:0911.4597 [hep- th]. 52

2010 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Wilson, Phys

K. Wilson, Phys. Rev. D 10 (1974) 2445

work page 1974

[2] [2]

A. M. Polyakov, Nucl. Phys. B 164 (1980) 171. 49

work page 1980

[3] [3]

’t Hooft, Nucl

G. ’t Hooft, Nucl. Phys. B72 (1974) 461

work page 1974

[4] [4]

Veneziano, Nucl

G. Veneziano, Nucl. Phys. B 117 (1976) 519

work page 1976

[5] [5]

Yu. M. Makeenko and A. A. Migdal, Nucl. Phys. B 188 (1981) 269

work page 1981

[6] [6]

A. A. Migdal, Phys. Rept. 102 (1984) 199

work page 1984

[7] [7]

Alvarez, Phys

O. Alvarez, Phys. Rev. D24 (1981) 440

work page 1981

[8] [8]

J. F. Arvis, Phys. Lett. B 127 (1983) 106

work page 1983