Pith. sign in

REVIEW 3 major objections 3 minor 33 references

The paper claims that adding a non-dynamical torsion to the Taub-NUT connection removes the Misner-string singularity and turns the solution into a gravitational dyon carried by two spin-fluid beams.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

T0 review · deepseek-v4-flash

2026-08-01 01:49 UTC pith:RIKD5BVL

load-bearing objection A compact, plausible torsion-based reinterpretation of Taub-NUT Misner strings; the main soft spot is the distributional rigor of the curvature computation. the 3 major comments →

arxiv 2607.25654 v1 pith:RIKD5BVL submitted 2026-07-28 gr-qc hep-thmath-phmath.MP

Taub-NUT as gravitational dyon with torsion

classification gr-qc hep-thmath-phmath.MP MSC 83C5783C4083D05
keywords Taub-NUTMisner stringstorsiongravitational dyonKomar chargescosmic stringspin fluidS-duality
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper tries to resolve a long-standing problem in the Taub-NUT spacetime: the symmetry axis, the Misner strings, makes the Levi-Civita connection singular and produces an Einstein tensor containing squares of delta functions, which is not mathematically meaningful. The authors propose absorbing that singularity by introducing torsion into the connection, chosen exactly to cancel the singular part of the vielbein differential. With this move, the Einstein tensor becomes that of a cosmic-string-like source with negative variable tension, concentrated on two beams along the symmetry axis. The torsion then accounts for the singular Komar fluxes, the asymptotic mass and angular-momentum charges, and the chronology-violating region around the strings. If correct, the Taub-NUT metric is a soliton supported by two stationary spin-fluid beams, and its gravitational-dyon nature is realized through torsion.

Core claim

The central claim is that the singular part of the Levi-Civita connection on the Misner strings can be removed by a non-dynamical torsion two-form with component T^t_xy = 4π s n δ²(x). After this substitution, the previously ill-defined Einstein tensor with delta-squared terms collapses to a distributional source of cosmic-string type: ε = -p = λ_m/(2Σ) δ²(x), and the source is identified as two stationary spin-fluid beams with spin density s_xy = s n F/2 δ²(x). The same torsion produces the singular Komar fluxes along the strings, so torsion — not an external matter rod — is what generates the asymptotic Komar charges. The paper also derives a dual Einstein tensor of the same form, showing

What carries the argument

The machinery is the first Cartan equation de^a + ω^a_b ∧ e^b = T^a. Because the azimuthal coordinate has de^0 containing a singular term 4π n s F δ²(x) dx∧dy, the torsion is chosen with component T^t_xy = 4π s n δ²(x) to cancel it. A regular spin connection is constructed with this torsion; computing its curvature and Einstein tensor yields the cosmic-string source. The torsion also acts as the generator of the singular Komar field via k_a T^a, and its divergence translates into the dislocation vector describing a time translation around the strings.

Load-bearing premise

The argument rests on accepting the distributional identity ddφ = 2π δ²(x) dx∧dy and on fixing the torsion amplitude 4π s n by hand to cancel the singular term; if this distributional calculus is not legitimate or the cancellation is not unique, the claimed cosmic-string and spin-fluid source is not established.

What would settle it

Evaluate the claimed Einstein tensor using a one-parameter family of smooth regularizations of δ², such as Gaussians of width ε, with the prescribed torsion component, and take ε→0; if the resulting distributional Einstein tensor depends on the regularization profile, the claimed source is not a well-defined geometric fact.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • The ill-defined Einstein tensor proportional to the square of a delta function is avoided: with the chosen torsion, all curvature contractions contain only linear delta functions, so the distributional geometry is under control.
  • The Misner strings acquire a concrete matter source: two cosmic-string-like beams with equation of state p = -ε and variable tension, plus spin density, rather than mysterious infinite-mass rods.
  • The singular Komar fluxes along the strings are generated by torsion, making torsion directly responsible for the asymptotic Komar mass and angular momentum; the total angular momentum remains zero through flux balance.
  • The dual Einstein tensor has the same structure, so the Taub-NUT solution becomes a gravitational dyon under S-duality, with the mass and NUT parameters rotating into each other.
  • The chronology-violating region around the strings is explained by a torsion-induced dislocation: transporting the azimuthal vector around the axis produces a time translation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the regularization is not unique, one can test the claim by computing the Einstein tensor with a family of smooth approximations to δ² and taking the zero-width limit; if the limit depends on the regularization profile, the cosmic-string source is an artifact of a specific convention rather than a geometric fact.
  • Because the torsion here is non-dynamical and tuned by hand, a natural next step is to couple the same torsion to a fully dynamical spin-density theory and check whether the spin-fluid beams actually solve the field equations, or whether they are merely a label for a fixed geometric input.
  • A similar distributional-torsion regularization may apply to other solutions with Misner strings or cosmic dislocations, especially rotating NUT metrics, where the force-line patterns already show asymmetry; extending the argument there would test whether the interpretation is stable.
  • If accepted, this reinterpretation removes Taub-NUT from the class of vacuum geometries in the distributional sense; that reclassification could affect uniqueness theorems and no-go statements that treat Taub-NUT as a vacuum solution.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 3 minor

Summary. The paper studies the Misner strings of the Taub-NUT metric from a Cartan-geometry viewpoint. Using the distributional identity d dφ = 2π δ²(x) dx∧dy, the authors first show that the Levi-Civita connection is singular on the strings and that its Einstein tensor contains ill-defined products of delta functions (Eq. 19). They then introduce a non-dynamical torsion two-form, Eq. (23), engineered to cancel the singular part of de⁰, and claim that the resulting torsionful connection has a regular Einstein tensor of cosmic-string type, Eq. (25), and a dual Einstein tensor Eq. (29). The source is identified with two stationary Weyssenhoff spin-fluid beams with spin density (34) and equation of state ε = −p (37). Finally, the torsion term is shown to reproduce the singular Komar field obtained previously in [8], leading to the conclusion that torsion generates the Komar fluxes and asymptotic charges.

Significance. If the mathematical steps were made rigorous, the paper would offer a concrete geometric interpretation of Taub-NUT Misner strings as torsion defects, connecting them to Einstein-Cartan spin fluids and to the earlier Komar-flux analysis of [8]. The construction is elegant and builds on a plausible physical idea. However, the central result currently depends on products of distributions that are not well-defined, and the torsion is chosen ad hoc rather than derived. The paper does not provide machine-checked proofs or independently testable predictions, so its significance is presently conditional on resolving the distributional issues.

major comments (3)
  1. [Torsionful connection, Eqs. (21)–(25)] The computation of the curvature 2-forms (24) is not a well-defined distributional calculation. The torsion T^0_23 contains δ²(x) (22), and the connection forms (21) contain terms singular at the poles; for example, ω^2_3 contains −v e³/(u√Σ) = −v dφ, whose exterior derivative has distributional support at θ=0,π. To compute ℛ^a_b one must form products such as δ²(x)·(1/u²), δ²(x)·∂δ²(x), and analogous terms, but no regularization (Colombeau algebra, mollification, or other) is specified and no proof of independence from the chosen delta sequence is given. Without this, Eq. (25), the Einstein tensor (25), and the source identification (37) are not established as distributions. The paper itself calls the corresponding δ² terms in Eq. (19) 'problematic'; the same standard must be applied to the torsionful computation.
  2. [Cartan equations and Recovering the Komar forms, Eqs. (18), (22)–(23), (43)] The torsion amplitude is fixed by the very singularity it is meant to cancel: Eq. (22)–(23) is chosen so that T^0_23 equals the singular part of de⁰ in Eq. (18). Consequently, the statement in Eq. (43) that torsion generates the singular Komar field is a restatement of the ansatz, not an independent result. The coefficient 4πsn is exactly the coefficient already present in Eq. (6) from [8]. To avoid circularity, the authors should clarify what the construction predicts beyond re-expressing the singular Komar flux in terms of torsion, and what observational or mathematical consequence would distinguish this description from the Levi-Civita one.
  3. [Torsionful connection, Eqs. (21), (31)–(37)] The connection (21) is not shown to be unique: the first Cartan equation fixes only the combination T^0_23 that cancels (18), leaving freedom in the remaining connection components. The paper does not demonstrate that the Einstein tensor (25), the dual tensor (29), or the spin-fluid interpretation (32)–(37) are independent of this choice. Moreover, the torsion is non-dynamical: it is inserted by hand, and the Weyssenhoff spin fluid is then read off from the algebraic relation (31). This makes the source an interpretation rather than a consequence of field equations. The authors should state this limitation explicitly and, if possible, prove invariance of the final source under the allowed connection freedom.
minor comments (3)
  1. [Introduction and conclusions] Minor typographical issues: 'pictires' in the caption of Fig. 2, and reference [10] is empty in the bibliography. These should be corrected.
  2. [Eq. (12)] The notation λ^s_J for 'fictitious spin densities' is introduced but the relation to the later spin density (34) is only given at infinity (35). The intermediate quantities would benefit from a short explanatory sentence.
  3. [Fig. 1 and 2] The figures are not referenced in the main text with enough detail. In particular, Fig. 2 is described only in the caption; the reader would benefit from a few sentences in the text explaining what the flux patterns show.

Circularity Check

2 steps flagged

Torsion is hand-set to cancel the Misner-string singularity and is then called its generator; the Komar 'recovery' is an identity by construction.

specific steps
  1. fitted input called prediction [Cartan equations, Eqs. (18), (22)-(23)]
    "Thus, we have two options: either take the singular Levi-Civita connection ω^a_b = ω^a_b(e) with zero torsion, or choose a regular connection with non-zero torsion compensating for the singular term (18). ... The non-zero torsion component is: T^0_23 = 2s n v^2 F/Σ (∂^2_x V + ∂^2_y V) = 4π s n F/Σ δ^2(x)."

    The torsion two-form is not derived from Einstein-Cartan field equations or from an independent spin-fluid model; its amplitude 4π s n is exactly the coefficient of the singular term (18) that it is introduced to cancel. Therefore the 'regular' connection and the resulting Einstein tensor (25) are the output of a fit to the Levi-Civita singularity. The source in (37), ε = −p = λ_m/(2Σ) δ^2(x), is read off from this chosen T and simply transcribes the coefficient already inserted in (22). Calling this a prediction of a cosmic-string-like source is renaming the fitted singularity as a physical origin.

  2. self definitional [Recovering the Komar forms, Eq. (43)]
    "The second term in (40) reads: k_a T^a = −4π s n Δ/Σ δ^2(x) dx∧dy, which is the singular field κ_s obtained in [8]. Finally, we see that within the framework of Einstein-Cartan theory, it is torsion that generates the MS singular terms."

    The singular Komar field κ_s was previously computed in [8] from the distributional identity d dϕ = 2πδ^2(x). Here T was fixed in (22)-(23) to be exactly 4π s n δ^2(x), i.e. the same distribution. Substituting this chosen T into the Cartan identity (40) necessarily reproduces the known singular Komar term. Thus 'torsion generates Komar fluxes' is not a dynamical consequence of torsion; it is the definition of the torsion that was inserted to cancel the singular part of de^0. The recovery of κ_s is algebraic bookkeeping with the fitted input.

full rationale

The central construction reduces, at its core, to a fitted input being presented as an explanation. Equations (22)-(23) define a torsion whose nonzero component is set, by hand, to compensate the singular term (18) in the Cartan equation. Everything that follows is tied to that choice: the Einstein tensor (25) and the dual tensor (29) are functions of the same δ^2(x) singularities, and the spin-fluid source (37) is read off from the Einstein tensor rather than derived from independent matter dynamics. Likewise, the statement that torsion 'generates' the Misner-string Komar fluxes (43) is an identity because the torsion was chosen to equal the singular distribution that already produced those fluxes in the metric-based calculation of [8]. This is not a case where an external benchmark or independent, parameter-free computation is being confirmed; the torsion amplitude is not predicted from any equation of motion. I also note that [8] is a self-citation, but that alone would not be circular; the circularity is the construction of T to match the very singularity it is then said to explain. The open mathematical issue of products of distributions in the curvature calculation is a correctness risk, not by itself a circularity argument, but it reinforces that the 'improved' connection (21) has not been shown to have a regularization-independent Einstein tensor. Overall, the paper contains real algebraic content (e.g., the explicit connection forms and the duality rotation), so a moderate score of 6 is appropriate: one or more central 'predictions' reduce by construction, but the derivation is not entirely vacuous.

Axiom & Free-Parameter Ledger

1 free parameters · 5 axioms · 2 invented entities

The central reinterpretation rests on distributional angular-coordinate identities, the choice of a non-dynamical torsion to cancel the singular part of the connection, and the Einstein-Cartan/Weyssenhoff dictionary used to turn that torsion into matter. The metric parameters m,n are inputs, not fitted. The only hand-set quantity is the torsion amplitude (Eq. 23), proportional to n. No new dynamical degrees of freedom are introduced, but the physical interpretation is not derived from a variational principle.

free parameters (1)
  • Torsion amplitude T^t_xy = 4π s n δ²(x) (Eq. 23)
    Chosen to cancel the singular part of de^0 in the first Cartan equation (Eq. 18). This is an ad hoc compensation rather than a value obtained from a dynamical theory.
axioms (5)
  • domain assumption Distributional identity d dφ = 2π δ²(x) dx∧dy
    Invoked at Eq. (3); underlies all singular terms and the construction of the torsion. The product of such distributions used at Eq. (19) is not standard.
  • standard math Cartan structure equations with torsion and the curvature definition (17), (24), (26)
    Framework of Riemann-Cartan geometry; assumed throughout without proof.
  • domain assumption Algebraic Einstein-Cartan relation between torsion and spin, and the Weyssenhoff fluid model (31), (33), (36)
    Used at Eqs. (31)-(37) to turn the non-dynamical torsion into a spin-fluid source; not derived in the paper and specific to this classical EC model.
  • ad hoc to paper Existence of a regular connection with torsion compensating the singular part of de^0
    The connection one-forms (21) are proposed by hand so that the torsion (22) cancels (18). No variational principle or uniqueness argument is given.
  • domain assumption SO(2) duality rotation of curvature with Hodge dual in EC theory
    Borrowed from Kol-Yau [23]; used to define the dual Einstein tensor (29) and the S-duality claim.
invented entities (2)
  • Non-dynamical torsion T^t_xy = 4π s n δ²(x) no independent evidence
    purpose: Compensates the singular part of the Levi-Civita connection on Misner strings and is said to generate the singular Komar fluxes.
    A field configuration chosen ad hoc; no independent observable outside the Taub-NUT interpretation; could be replaced by a different regularization.
  • Two stationary Weyssenhoff spin-fluid beams no independent evidence
    purpose: Provide the matter source of Taub-NUT in Einstein-Cartan theory, with ε=−p and spin density (34).
    Interpretive source constructed from the torsion; no independent detection or prediction outside the paper's framework.

pith-pipeline@v1.3.0-alltime-deepseek · 7803 in / 15472 out tokens · 159651 ms · 2026-08-01T01:49:01.704328+00:00 · methodology

0 comments
read the original abstract

We show that the Levi-Civita connection for the Taub-NUT metric is singular on Misner strings and generates an Einstein tensor containing the squares of the delta function. This can be improved by introducing torsion, in which case the Einstein tensor may be interpreted as that of a cosmic string with negative variable tension. Komar flows in Misner strings are generated by torsion, which is therefore responsible for the asymptotic Komar charges. The Taub-NUT metric then appears as a soliton supported by two spin-fluid beams.

Figures

Figures reproduced from arXiv: 2607.25654 by Dmitri Gal'tsov, Rostom Karsanov.

Figure 1
Figure 1. Figure 1: FIG. 1. Schematic picture of the angular momentum Komar [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Exact FL patterns of the Komar two form [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

33 extracted references · 14 linked inside Pith

  1. [1]

    and the significance of Misner strings (MS) [2], there is still no consensus on the nature of these objects. Misner

  2. [2]

    Bonnor [4, 5] proposed considering the MS as weak singularities similar to cosmic strings, and Dowker

    proposed eliminating strings by imposing a periodic condition on time, but this only exacerbated the prob- lem [3]. Bonnor [4, 5] proposed considering the MS as weak singularities similar to cosmic strings, and Dowker

  3. [3]

    Taub- NUT (Newman, Unti, Tamburino) Metric and Incompat- ible Extensions,

    J. G. Miller, M. D. Kruskal and B. B. Godfrey, “Taub- NUT (Newman, Unti, Tamburino) Metric and Incompat- ible Extensions,” Phys. Rev. D4, 2945-2948 (1971)

  4. [4]

    Empty space generalization of the Schwarzschild metric,

    E. Newman, L. Tamburino and T. Unti, “Empty space generalization of the Schwarzschild metric,” J. Math. Phys.4, 915 (1963)

  5. [5]

    The Flatter regions of Newman, Unti and Tamburino’s generalized Schwarzschild space,

    C. W. Misner, “The Flatter regions of Newman, Unti and Tamburino’s generalized Schwarzschild space,” J. Math. Phys.4, 924-938 (1963)

  6. [6]

    polarized

    declared the Taub-NUT solution to be a gravitational dyon. But the question of what matter can give rise to Misner strings in Bonnor’s interpretation remains a sub- ject of debate. Calculation of the outgoing Komar fluxes across the cylinders surrounding the strings [7] seemed to indicate the presence of negative mass rods with in- finite angular momenta ...

  7. [7]

    A new interpretation of the NUT metric in general relativity,

    W. B. Bonnor, “A new interpretation of the NUT metric in general relativity,” Math. Proc. Cambridge Phil. Soc. 66, no.1, 145-151 (1969)

  8. [8]

    Physical interpretation of N.U.T

    Sackfield A. Physical interpretation of N.U.T. metric. Mathematical Proceedings of the Cambridge Philosoph- ical Society. 1971;70(1):89-94

  9. [9]

    The nut solution as a gravitational dyon,

    J. S. Dowker, “The nut solution as a gravitational dyon,” Gen. Rel. Grav.5, no.5, 603-613 (1974)

  10. [10]

    Physical interpretation of NUT solution,

    V. S. Manko and E. Ruiz, “Physical interpretation of NUT solution,” Class. Quant. Grav.22, 3555-3560 (2005) [arXiv:gr-qc/0505001 [gr-qc]]

  11. [11]

    Gravielectric and gravi- magnetic fluxes in nutty black holes,

    D. Gal’tsov and R. Karsanov, “Gravielectric and gravi- magnetic fluxes in nutty black holes,” Phys. Lett. B878, 140513 (2026) [arXiv:2605.09786 [gr-qc]]

  12. [12]

    Gravitational field strength and generalized Komar integral,

    W. Simon, “Gravitational field strength and generalized Komar integral,” Gen. Rel. Grav.17, 439 (1985). [10]

  13. [13]

    Interaction between a stationary electric charge and a stationary magnetic dipole,

    W. B. Bonnor, “Interaction between a stationary electric charge and a stationary magnetic dipole,” Class. Quant. Grav.18, 2853 (2001) [arXiv:gr-qc/0203046 [gr-qc]]

  14. [14]

    The interaction between a stationary electric charge and a magnetic dipole

    W. B. Bonnor, “The interaction between a stationary electric charge and a magnetic dipole”, Physics Letters A284(2001), no. 2-3, 81–84

  15. [15]

    Exact Space-Times in Einstein’s General Relativity,

    J. B. Griffiths and J. Podolsky, “Exact Space-Times in Einstein’s General Relativity,” Cambridge University Press, 2009, ISBN 978-1-139-48116-8

  16. [16]

    Conical singularities and torsion

    K. P. Tod, “Conical singularities and torsion”, Classical and Quantum Gravity,11, 1331–1339, 1994

  17. [17]

    Spinning strings as torsion line space-time defects,

    P. S. Letelier, “Spinning strings as torsion line space-time defects,” Class. Quant. Grav.12, 471-478 (1995)

  18. [18]

    Spinning strings and cosmic dislocations,

    D. V. Galtsov and P. S. Letelier, “Spinning strings and cosmic dislocations,” Phys. Rev. D47, 4273-4276 (1993)

  19. [19]

    Spinning cosmic strings: A General class of solutions,

    N. Ozdemir, “Spinning cosmic strings: A General class of solutions,” Int. J. Mod. Phys. A20, 2821-2832 (2005) [arXiv:gr-qc/0508017 [gr-qc]]

  20. [20]

    Duality in linearized gravity,

    M. Henneaux and C. Teitelboim, “Duality in linearized gravity,” Phys. Rev. D71, 024018 (2005) [arXiv:gr- qc/0408101 [gr-qc]]

  21. [21]

    Monopoles for gravitation and for higher spin fields,

    C. W. Bunster, S. Cnockaert, M. Henneaux and R. Por- tugues, “Monopoles for gravitation and for higher spin fields,” Phys. Rev. D73, 105014 (2006) [arXiv:hep- th/0601222 [hep-th]]

  22. [22]

    Supersymme- try and Gravitational Duality,

    R. Argurio, F. Dehouck and L. Houart, “Supersymme- try and Gravitational Duality,” Phys. Rev. D79, 125001 (2009) [arXiv:0810.4999 [hep-th]]

  23. [23]

    Gravitational duality and rotating solutions,

    R. Argurio and F. Dehouck, “Gravitational duality and rotating solutions,” Phys. Rev. D81, 064010 (2010) [arXiv:0909.0542 [hep-th]]

  24. [24]

    Dual Komar Mass, Torsion and Riemann- Cartan Manifolds,

    U. Kol, “Dual Komar Mass, Torsion and Riemann- Cartan Manifolds,” [arXiv:2010.07887 [hep-th]]

  25. [25]

    Duality in Gauge Theory, Gravity and String Theory,

    U. Kol and S. T. Yau, “Duality in Gauge Theory, Gravity and String Theory,” [arXiv:2311.07934 [hep-th]]

  26. [26]

    Conical singularity in spacetimes with NUT is observer dependent,

    I. Kol´ aˇ r, P. Krtouˇ s and M. Ossowski, “Conical singularity in spacetimes with NUT is observer dependent,” Phys. Rev. D112, no.10, 104021 (2025) doi:10.1103/w2ct-wjh5 [arXiv:2507.21238 [gr-qc]]

  27. [27]

    Gravitational multi-NUT solitons, Komar masses and charges,

    G. Bossard, H. Nicolai and K. S. Stelle, “Gravitational multi-NUT solitons, Komar masses and charges,” Gen. Rel. Grav.41, 1367-1379 (2009) [arXiv:0809.5218 [hep- th]]

  28. [28]

    Stationary and axisymmetric solutions of higher-dimensional general relativity,

    T. Harmark, “Stationary and axisymmetric solutions of higher-dimensional general relativity,” Phys. Rev. D70, 124002 (2004) [arXiv:hep-th/0408141 [hep-th]]

  29. [29]

    On the Smarr formulas for electrovac spacetimes with line singularities,

    G. Cl´ ement and D. Gal’tsov, “On the Smarr formulas for electrovac spacetimes with line singularities,” Phys. Lett. B802, 135270 (2020) [arXiv:1908.10617 [gr-qc]]

  30. [30]

    f(T) teleparallel gravity and cosmology,

    Y. F. Cai, S. Capozziello, M. De Laurentis and E. N. Sari- dakis, “f(T) teleparallel gravity and cosmology,” Rept. Prog. Phys.79, no.10, 106901 (2016) [arXiv:1511.07586 [gr-qc]]

  31. [31]

    Rotating traversable wormholes in Einstein-Maxwell theory,

    G. Cl´ ement and D. Gal’tsov, “Rotating traversable wormholes in Einstein-Maxwell theory,” Phys. Lett. B 838, 137677 (2023) [arXiv:2210.08913 [gr-qc]]

  32. [32]

    Relativistic dynamics of spin-fluids and spin-particles moving with the velocity of light,

    J. Weyssenhoff and A. Rabbe, “Relativistic dynamics of spin-fluids and spin-particles moving with the velocity of light,” Acta Phys. Polon.97, (1947)

  33. [33]

    Y. N. Obukhov and V. A. Korotkii, Class. Quant. Grav. 4, 1633-1657 (1987)